The Rational Legislator Revisited. Bounded Rationality and Legisprudence

Abstract

In this chapter the classical view of the rationality of the legislator is challenged. Although it operates as an assumption in legal interpretation, it is argued that legislators are not omniscient or that they are more rational than other agents. Like any other agent or decision maker, they have limited skills and computational capacities, and act under time constraints. These limitations affect the way legislation is created. The thesis of this chapter is that legislators are boundedly rational agents who make at best “satisficing” norms and not optimal ones. To legislate rationally involves taking into account these limits of rationality.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1.1 The Familiar View and Its Discomfort

The familiar view or the view that lawyers commonly hold on the law reflects a standard belief of what law is and how it operates. I will begin by articulating some focal aspects of the familiar view in order to highlight the problem which legisprudence as a theory of rational legislation deals with. Following that, I will challenge the familiar view on the basis that it pays insufficient attention to legislative law making from a theoretical perspective, after which I will explore what “rational legislation” seems to involve and what type of questions arise when dealing with that problem.

One aspect of the familiar view is that the object of legal science or legal dogmatics is the law, both the law in the books and the law in action. Legal science describes, systematizes, and explains the law as it is. It follows that the propositions of legal science are propositions de lege lata, that is, existing positive law or what counts as the set of binding legal norms of a legal system. The familiar view is not concerned with what or how the law could or should be. On this view, considerations de lege ferenda fall out the scope of legal science.

The distinction between law de lege lata and de lege ferenda reflects another aspect of the familiar view. This points to the separation between law and morals on the one hand and between law and politics on the other. It comprises part of the familiar view in that legal science in order to be scientific is to be objective. Description, systematization, and explanation of the law can only be scientific if it avoids making evaluative statements or propositions de lege ferenda. As far as the distinction points to the separation between law and morals, legal theory is said to be positivistic while its pointing to the separation between law and politics makes it legalistic. Under the familiar view both positivism and legalism characterize the law and contribute to the delineation of both the object and the method of legal science. They contribute to the determination of both the nature and the way of scientifically investigating the law.

Yet another feature of the familiar view is that the law is a system. In this respect lawyers refer to the law as the legal system or the legal order as a systematic set of legally valid norms. Valid legal norms are part of a system, and any system consists of valid norms. The distinguishing characteristic of being legally valid is a necessary and a sufficient criterion for a norm to belong to the system. The validity of a legal norm is usually defined—at least in most civil law systems—as its being created in accordance with the hierarchically higher norms of the system.

This reveals a further facet of the familiar view: the belief that legally valid norms are unquestionably legitimate. Legal validity derives from conformity with hierarchically higher norms. These norms confer the power to issue new norms, and as long as the power holder acts in accordance with these power-conferring norms the ensuing norms are valid. Power-conferring norms confer upon—as well as limit the competence of—the power holder which makes his power a legal power. Any norm-creating act within the competence of the power holder results in a norm that is both valid and legitimate. More specifically, norms created on the basis of hierarchically higher norms are formally valid and internally legitimate, while the legal system composed of these norms is itself externally legitimate.

This points to an additional feature of the familiar view being that the legal system itself is believed to be legitimate because and in so far as it is formally legitimated. The ultimate basis of the system’s legitimacy is consent expressed in terms of the consent of the governed that is confirmed on a regular basis by democratic elections. The social contract establishes the institutions of a political society that finds its foundation hence its legitimacy in the initial agreement. The establishment of the institutions of political society reflects a distributional organization of power which is commonly articulated as the separation of powers. This articulation of the distribution of powers involves a hierarchical order of the institutions of the legal system according to which the legislator creates law that the executive implements, while the judiciary has the duty to apply both the legislative and the executive norms. This is supported by the idea of the rule of law that holds that power can only be exercised on the basis of the law.

The familiar view culminates in the idea of legal science that describes, systematizes and explains the law as it is, thus making true propositions about it, and in so doing sustains—if not corroborates—the familiar view.

Current legal theory as the theory of legal science or the meta-theory of law has focused most of its attention to date on the position of the judge. Viewed from this angle, legal theory has developed various methodologies of law application. These methodologies essentially focus on judicial interpretation of the law, taking into account the judiciary’s institutional subordination to the legislature. The core assumption of judicial interpretation is that the law is rational, and that its rationality is to be preserved throughout its subsequent application. This assumption underpins the actions of the judges as well as legal scholars.

Following on from this assumption underlying the familiar view the legislator is assumed to be rational. This involves a regressive assumption: from the assumption of the rationality of the product the familiar view regressively assumes the rationality of its maker. This regressive assumption of the rationality of the legislator is kept intact despite the deep-rooted conviction that legislative lawmaking has its origin in politics. Politics, in turn, is commonly described as a “power game”, “Realpolitik”, or even logrolling and horse trading, which ultimately leads to sofa compromises. Despite this not particularly elevated way of making the law, it is held to be rational and so is its maker.

In this contribution I will challenge some aspects of what I have described as the “familiar view”. More specifically, I will explore the assumption of the rationality of the legislator underlying judicial activity and the corroborating support it receives from legal science. In doing so I will adopt a meta-theoretical point of view throwing a different light on the position of the judge from which a critical assessment can be carried out of the legislator’s assumed rationality. Following on from the fruitfulness of this exploration I will then further investigate the possibility of critically theorizing the legislator’s rationality using what in earlier work, I have called a “legisprudential” theory of law.

A legisprudential theory of law in contradistinction to the widespread jurisprudential approach no longer takes for granted the central position of the judge as the main legal agent. It aims instead to enlarge the spectrum of legal theory as the meta-theory of legal science by including the legislator as a legal agent thus downplaying his mainly political role in lawmaking. If my exploration and diagnosis turns out to be fruitful, it will provide us with a different way of theorizing legislative law making. As a note of comfort to adepts of the familiar view who may feel alarmed by this pronouncement, I should add that my approach amounts to a way of putting the emphasis differently as opposed to a complete revolution. It is more an invitation to look at the law in a different way than a call to radically change it.

This new way of looking at the law is triggered by an often felt discomfort with legislative law. Under the familiar view law creates order. It is the type of bureaucratic law that Weber described as the fertile soil for capitalism that emerged from one of the variants of Protestantism. The order thus created secures the economic interests that make it worthwhile to take the risk of capital investments. It does not however only protect economic interests by providing a stable environment. In creating order it also secures other interests by stabilizing legitimate social expectations. From this broader perspective law creates social stability in that it provides standards of behavior in the form of rules that serve at the same time as a standard of criticism for deviant behavior. In doing so the law aims to create certainty through guiding human behavior.

The pendant of the social order created by the law is the set of norms, in the form of rules, that are referred to as the legal order that legal science describes, further systematizes and explains. According to positivist legal theory the criterion for membership of a norm of the legal system is its validity. A general version of this criterion is “being ordered by the sovereign”, and can be further constrained by more specific criteria such as “in accordance with the existing norms of the legal system”. It can—and most of the time is—supplemented by a theory of the sources of law including some form of hierarchy among them.

It is here that the discomfort with legislative law comes to the fore. Legal science of the positivistic brand is concerned with the existence of valid law and who can make and apply it. The description and systematization of this product of the legal system is the object of legal science. Apart from who can do what it expresses little concern for how law is made. In most western legal systems an explosive growth of norms can be observed. One need only take note of the exponentially growing number of pages of the official publication bulletins of legislation to verify this observation. Something similar to road traffic is in evidence here. The more cars there are on the roads, the more collisions occur. Traffic no longer runs smoothly, but generates jams, and instead of facilitating mobility it tends to have the opposite effect. The same happens when legal systems start producing norms with an ever increasing rapidity: the risk of collisions between norms grows proportionally. As a consequence, the set of norms assembled in the legal system turns out to be no longer systematic but rather tends towards disorder. This decreasing systematicity of the legal order tends to leave in its wake a weakening of the securing ordering of social interaction which seems to result in the opposite effect to legal certainty. Other defects often complained of are that the norms contradict each other or are in conflict with the constitution, while many norms are poorly drafted. Some norms are of dubious procedural origin, lack a significant rationale, or embody a faulty means-end hypothesis.

An important side-effect of this is that while law aims at offering guidance, it also tends to generate new causes for conflicts rather than preventing them. This phenomenon is referred to as “adversarial legalism” of which divorce law is a sad example. Indeed, it seems that sometimes law creates conflicts, rather than solving them, and so it tends to deviate from guiding behavior by facilitating social interaction.

Another aspect of the discomfort with legislative law underlying the exponential growth of the legal system is that legislative law is produced at an ever quicker pace. This often results in weak preparation and poor drafting of normative texts. Grandmothers use to say that good work needs time. Legislators are likely to dismiss this advice and tend to produce vague texts that are difficult to apply or very detailed ones that quickly become redundant. On this point as well, both officials and citizens express their discomfort while still maintaining belief in the rationality of the legislator.

1.2 Rationality

The modern Western metaphysical tradition conceives Reason as a faculty with which humans are endowed. All humans—ontologically speaking—have it as a differentia specifica, and the correct use of this faculty results in rational, true knowledge and action. Reason, in other words, is the faculty the correct use of which results in true knowledge and moral action.

A post-metaphysical view, for its part, is critical of this direct access to reality, including human nature. To say that Reason is a faculty we have, amounts to ascribing to ourselves a quality. This ascription may be taken as an assumption which is close to the metaphysical approach. Under the post-­metaphysical view, it amounts to a belief, which we may truly hold, yet the truth of the proposition in which we ascribe rationality to ourselves crucially depends of course on the correct use of reason. This is problematic since we may truly believe that we are rational, yet the justification for the belief crucially depends on its justification hence the use of reason. We may truly believe that we are rational or that we have the faculty of reason, but for other than rational reasons (e.g. because God created us as beings endowed with reason). Hence the rational justification of the belief, however truly we hold it, presupposes again that we have the faculty of reason or that we are rational beings.

In a similar vein, assuming that we have the faculty of reason is one thing. It is quite another to include in that assumption the idea that we therefore know how to use it. The assumption that we have it, and if we use it correctly it will lead to true knowledge and action seems to include the idea that we also have the manual for how to use it correctly. Apart from the circular self-­ascription of rationality this observation suggests the self-referentiality of the concept of rationality. Rationality thus conceived contains its own criteria which is a stipulative definition.

If we are no longer content with relying on “human nature” presupposing a direct access to reality, nor on a stipulatively defined self-referential notion of rationality—or self-evident reason including its own criteria—nor self-­confidently though circularly describing ourselves as having it, we must look for different grounds for dealing with rationality in a meaningful way, seeing it as an aptitude rather than as a faculty.

When leaving aside the aspiration of a direct access reality in terms of “human nature” or conceiving rationality in terms of “faculty”, we are less at risk of circularity in ascribing to ourselves a self-referential quality. Reason as a faculty which we are assumed to have (or that we ascribe to ourselves), so it seems, is not always “in act” (even, ironically, in the act of ascribing it to ourselves), which does not of course exclude the possibility of the belief that we are rational, in that we are capable of rationally justifying this belief.

One way of seeing this is that this capacity is not a capacity of isolated individuals as the Western tradition following Descartes suggests. To be a capable subject is not identical to saying that we are simply endowed with the capacity for rational knowledge and action waiting for the appropriate time, place and circumstances to be activated. Seen from this angle, we have a “concept of the self” distilled from a direct access to reality or stipulatively defined. I have argued elsewhere for the view that it is “time, place and circumstances” that activate the capacity.¹ Time, place and circumstances as the context of a subject are constitutive of this capacity. This suggests that rationality is not only agent-related but is also context dependent. It depends on a context of interaction (or participation) with others since it is only through interaction with others that the capacity develops.

Rationality is not a concept that belongs to the concept of the self, whatever this means. As a capacity for rational knowledge and action it belongs to our self-conception. If we were to suppose for a moment that “rationality” was a concept, we would be back on the track of “searching for its true meaning”. That being the case, we can say that in the absence of a true meaning, agents can truly hold different conceptions of this concept. Different conceptions of rationality may conflict with each other, which leads Descartes to suggest that none of the dissenters are right. Dissent about the true meaning of rationality is not a sign of the absence of rationality though, it is only an indication of dissent/absence of agreement.

It is not because conflicts about the meaning of rationality arise that we are all wrong about its point. Another way of looking at it than the Cartesian solution is to conceive of rationality as an agent-related capacity for rational knowledge and action which includes the capacity for solving conflicts (theoretical in this case). In this respect we may fail to come to an agreement, but this is not a sign of irrationality. It is rather a sign that our capacities are limited without being inexistent or meaningless.

Different conceptions of rationality may appear in different contexts. Rationality in an economic context is different from rationality in a political or legal context. Referring to “rationality as capacity”, this suggests that the point of economics is to solve problems of scarcity by using rational capacities (calculating, anticipating, etc.). But “rationality as capacity” is not exhausted in the tabulation of different conceptions or rationality. These different conceptions are connected to each other, something which I will not touch on here.

The disadvantage of rationality according to what I have broadly referred to as the Western tradition is that it leads to an abstraction of the real world. It is common sense to say that, even if we are rational beings, not every one of us, and not even us ourselves, always behave that way. It is intuitively correct to say that some people are more rational than others, in that they act more rationally in certain circumstances than others. And there is no shame in saying that we are essentially fallible.

To conceive of rationality as an agent-related capacity to acquire knowledge and to act in a rational way is one way of downplaying the core idea of the intellectualist tradition conceiving rationality as the divine part within us. The embedded subject being capable of coping with complexity—both theoretical and practical—to paraphrase Herbert Simon’s felicitous expression. This approach may be less “strong”, but it also seems less “wrong”. It may to a significant extent temper our illusionary ambition in the quest for the Holy Grail of unqualified truth, disillusioning ourselves of the belief that we ultimately reach it, without however falling victim to the opposite mindset of vulgar pragmatism or even cynicism.

The view of rationality as the capacity to cope with complexity has been formulated by Simon as a critique of homo economicus, who can be seen as an heir of the intellectualist tradition, a twin of Descartes’ homo rationalis. Homo economicus as the rational agent in neo-classical economics is supposed to always act in such a way as to maximize his wealth. Homo economicus settles for nothing less than the best. On Simon’s view, the rationality assumption of neoclassical economics, for different reasons, is too strong. Simon’s view amounts to substituting a weaker version of rationality for the comprehensive model of rational choice connected to substantive rationality which neoclassical economics relies on. Simon calls this weaker version of rationality “bounded rationality”, which I will come to shortly.

Let us first see how we can get a better hold on the rational legislator, and thus explore the context in which he shows up most frequently, that is, in the practice of judicial interpretation or construction of legislative texts. Starting from judge-centered legal theory according to which the judge is the central agent in the legal system, the best way, I would suggest, to get a hold on the problem of the rationality of legislative law and the legislator is to look at it through judicial eyes.

The argument will follow two lines. The first amounts to a clarification of some characteristics of judicial interpretation. Its main characteristic seems to be that the legislator is held to be rational, which is a variant of the principle of charity at work in judicial interpretation. My examination aims to identify a number of aspects of the assumption of the rationality of the legislator which are connected to the familiar view.

The second line of the argument, connected to the first, amounts to a reinterpretation of the principle of charity based on shedding a different light on the rationality of the legislator. Since legislators are human agents, there is no reason to suppose that they have the faculty of rationality or that they are rational. Rather, we assume that they have such faculty or that they are rational. My thesis is that the legislator is capable of rationality and acts according to “time, place and circumstances” or in a social context. Following on from this idea, I will qualify rationality as “bounded rationality”. The combination of the two lines of argument will be helpful in further elaborating the contours of a legisprudential theory of law as a theory of rational law making.

Throughout the act of interpretation the author of the interpretandum is presumed to be rational in that he intends his work to be intelligible. When it comes to the interpretation of, say, a contract, an executive norm, or a statute, the interpreter assumes that the authors of the documents in question are rational agents. Upon this assumption he tries to make as much sense as possible of the text before him. He tries, that is, to find out what the text means assuming that its authors have an intelligible end or purpose which they want to achieve. In doing so, he reads paradoxical or contradictory contract clauses and legislative and executive rulings in a way that shows them to be the will of a rational agent rather than construing them as absurd, unintelligible, or contradictory.

For the purpose of this essay, I will limit the scope of inquiry to legislative acts. Their interpretation follows a recognizable path: if contradictions or inconsistencies appear within legislation both judicial and doctrinal interpretation consists of removing these using canonical principles like lex posterior, lex specialis, or lex superior. Interpretation using these principles aims at safeguarding the rational connection between the norms of the system by attributing the legislation to a rationally acting legislator. The purpose of this is the preservation of the consistency or the systematic nature of the legal order.

This is only one aspect of how legal interpretation proceeds in order to keep the system rational: the interpreter assumes that the author is a rational agent and acts accordingly. The assumption of the rationality of the legislator portrays him as an agent who issues “consciously made rational rules” to use Weber’s wording. According to Weber domination on the basis of valid legal norms is legitimate. Legal norms are valid based upon their being issued according to the prevailing secondary norms of the legal system. His position, in short, means that legality involves legitimacy. Consciously made rational norms legitimize domination.² This includes the suggestion that legitimacy depends on the rationality of norms, driven by the assumption of the rationality of their author.

The assumption of the rationality of the legislator, as we have already seen, allows the interpreter to erase a number of contradictions in or between legislative norms, as well as a number of impertinent declarations of the legislator. It also allows the reader to interpret an unconstitutional norm as if it were in conformity with the Constitution. Such an assumption can, furthermore, steer the adaptation of legislative rules to newly emerging circumstances based on the assumption that a rational legislator would have adapted his ruling had he been aware of such circumstances. These are but a few of the possibilities of the ideas that are included within the assumption of the rationality of the legislator. This assumption is brought into play as a hermeneutic tool in order to make as much sense as possible of legislative law. As a hermeneutic tool it also has a curative function in that the qualities which we ascribe to legislative norms are often lacking: some norms do contradict each other, some of them contradict the constitution, and many of them are poorly drafted. Other norms may be of dubious procedural origin, lack a significant rationale, or embody a faulty means-end hypothesis. This was referred to above as the discomfort with legislative law.

These deficiencies are patched up through interpretation by assuming that the legislator is, after all, a rational agent. What the interpreter can do, yet the only thing he can legitimately do, is making the best of the situation by assuming that the author is a rational agent. In his attempt to optimize his interpretation he anticipates the sense that the interpretandum has. This presupposition is an anticipation of the perfect unity of sense (“Vorgriff der Vollkommenheit”), as Gadamer puts it. To attempt to make sense of a text, typically a statutory text, presupposes that it makes sense and this presupposition includes a preconception of its perfection.³ Optimization of interpretation relies on the hypothesis of the maximal coherence of the interpretandum on the assumption that its author is rational. This is what in legal interpretation the “rationality of the legislator” amounts to.

1.3 The Principle of Charity

The idea that interpretation involves an anticipated recognition of the rational character of the interpretandum and its author has a long history. The “anticipation of perfection” as Gadamer calls it, can be found in Christian Thomasius’ work where it is suggested that it is the proper attitude of a benigna interpretatio. G F Meier qualifies this attitude as aequitas hermeneutica.⁴ In more recent times, Neil Wilson⁵ has coined the expression “principle of charity” to point to this fundamental premise underlying interpretation that was further elaborated by Davidson and Quine.

Quine has argued that translation should preserve logical laws. That is to say, we should translate a speaker’s utterances in such a way as to avoid construing them as contradictory or absurd. We presume, in other words, that the speaker of a foreign language follows the laws of logic. The same applies to a foreign speaker’s utterances in a language with which we are familiar. Quine points out that the speaker’s silliness, beyond a certain point, is less likely than bad translation. The same idea applies a fortiori when we have the same native tongue as the speaker.

In a similar vein, Davidson’s is the idea that interpretation should preserve the content of the interpretandum on the assumption that it is intelligible based upon the author’s rationality. Both the preservation of the content and the laws of logic through interpretation reflect a charitable attitude towards the author of the interpretandum.⁶

This is an important aspect of judicial interpretation. Judicial application of norms requires understanding of their text. Understanding involves interpretation, since no text is self-interpreting. In line with Gadamer’s suggestion, the anticipation of perfection provides a clarification of the “understanding of understanding”. Therefore it seems that the principle of charity can be meaningfully implemented in judicial interpretation preceding the application of legislative law. Under the principle of charity, the judge considers the author of legislative norms to be rational in assuming that he did not wish to contradict himself, that he did intend to avoid absurd results following on from his ruling, that he was aware of the preferences of those subjected to his rules, etc.

All of this sounds very idealistic not to say counterfactual or counterintuitive. Our experience with today’s legal systems is sometimes deceptive and contrary to what the familiar view promises. Legislators behave in ways that are often viewed as being anything but rational, a quality that we keep assigning to them against our better knowledge. Laws can be unconstitutional, vague, poorly drafted, ill-adapted to the end which they aim at realizing, patently inequitable, or “running behind the facts”. Logical problems are, so it seems, but the easy part of the game while most legal systems seem ill-equipped to tackle the other parts.

The assumption of the rationality of the legislator in legal interpretation lies at the very basis of legal interpretation as a “prejudice” or a condition for the understanding of legislative texts. It is, one can say, a foundational premise of the legal order that is effectuated throughout judicial interpretation of legislative law. All methods of interpretation as a matter of fact rely on that foundational premise. In order to get a better hold on the discomfort with the law in connection with the familiar view, we must work our way through this foundational premise and find out what use a more finely-tuned interpretation of it can offer.

Adopting a rudimentary scheme of interpretation, three different contexts can be identified, namely, a semantic, a syntactic, and a pragmatic context. Most methods of legal interpretation can be classified according to these contexts. So, literal interpretation best fits the semantic context, while logical, systematic, and grammatical interpretation falls within the syntactic context. Teleological interpretation, finally, comes within the pragmatic context. What connects the different methods of interpretation in relation to their different contexts is their aim of making sense of normative texts. Making as much sense as possible in these different contexts in other words is what legal interpretation amounts to. Upon the underlying assumption of the rationality of the legislator, the different types of interpretation articulate different—though not separate—aspects of this assumption.

It is generally agreed, that the principle of charity operates as a constraint on interpretation. Charitable interpretation requires that we consider the object of interpretation to be the work of a rational agent. We attribute rationality to the agent and in doing so we constrain our interpretation of the interpretandum. One way of seeing this assumption is to put it in terms of a presumptio juris et de jure which is irrefutable, by which I suggest a possible reinterpretation of the principle of charity as a presumption.⁷

The reinterpretation of the principle of charity which I propose places it between two extremes. One extreme states: “do not assume a priori that the legislator is irrational”; the other dictates: “never interpret the legislator as irrational”. The former is the weakest reading of the principle of charity. While it says that the legislator’s irrationality must not be a priori assumed, it does, however, leave room for doing so. The latter is the most stringent reading in that it suggests that Reason is simply incarnated in the legislator and is straightforwardly reflected in the norms which he issues. On this reading of the legislator he is held to be omniscient. This is the reading of the principle of charity as a presumptio juris et de jure articulated in the foregoing paragraphs and that underlies the familiar view.

The weakest reading, for its part, is an attractive interpretation for our purposes, though it requires some qualification. Taken as it stands, it does not inform us when (not) to consider an act of the legislator to be irrational; it only tells us not to assume this a priori.

The required qualification amounts to an “in-between” reading of the principle of charity in the following manner: “Do not judge the legislator to be irrational unless you have an empirically justified account of what he is doing when he violates normative standards”.⁸ This in-between reading of the principle of charity is attractive in that it allows one to substitute empirically justified criticism of the rationality of the legislator for an a priori assumption. In doing so, the rationality of legislative law is prima facie presumed and allows a gradual qualification in the light of empirical data. It opens up the possibility of an empirical assessment of the legislator’s rationality that may be assessed on the basis of concrete experience with legislative law. On this in-between reading of the principle of charity the legislator’s rationality can be turned, so it seems, from a presumptio juris et de jure into a presumptio juris tantum. The in-between reading of the principle of charity then invites us to consider the legislator as a rational agent although his performance is not always optimal without for that reason being irrational.

Before continuing our discussion, we should briefly pause to pick up on a remark made above. I stated that the aim of a legisprudential theory of law amounts to a reflection on making legislative law more rational and more rationally. This points to a distinction between what I propose to call legislation in the passive and in the active sense. Legislation in the passive sense is a product. It is used in expressions like “environmental legislation” or “tax legislation”. Legislation in the active sense on the contrary refers to the process of legislation. While problems with the rationality of legislation in the passive sense only appear when it comes to interpretation and application, problems of the rationality of legislation in the active sense appear at an earlier stage, that is, throughout the process of legislative law making. A legisprudential theory of law that no longer takes for granted the central position of the judge, but also considers the legislator as a legal agent (although he is also a political agent) focuses on the process of legislation or legislation in the active sense. From this perspective, the process of legislation seems to be the appropriate context for the exploration of the rationality of the legislator.

I hasten to add that when using the notion of “legislative process” I do not have in mind something like the descriptive approach adopted by some political scientists in their study of legislative behavior. These political scientists usually tend to concentrate on the informal processes of law making—like lobbying, compromise, and the like—rather than on more formal or rational aspects of the legislative process. Within the aspirational context of legisprudence, I rather have in mind the process of legislation as legislative action, that is, behavior for a reason. This comes down to asking ourselves how, that is, for what reasons, a legislator acts throughout the process of legislation.

These reasons refer to choices which the legislator makes. These choices express his “will”, so it is the process of justification of legislative law making that I am interested in. In this respect the scope of legisprudence is broader than discovering the will of the legislator throughout judicial or doctrinal interpretation. The point of my concern, in short, is practical reason in legislation. It does not so much deal with who makes which choices than with how they are made; and it does not so much deal with what norms are made than with how they are made. What can we—rationally—expect from a rational legislator in this respect?

In trying to answer this question it is helpful to turn to the in-between reading of the principle of charity that I have just referred to. The principle as was stated, constrains interpretation. The product of legislation is what judges interpret, yet the process of legislation is what legislators are engaged in. To avoid considering the legislator a priori irrational amounts to a prima facie presumption that he intends to be rational in the process of legislation even if the product in question is not always optimal. In following the line of the in-between reading of the principle of charity I propose to further explore the legislator’s rationality in the next section.

1.4 Bounded Rationality

We should recall the most stringent reading of the principle of charity according to which we a priori assume that the legislator is a rational agent. In a similar way to the familiar view in law, neoclassical economic theory assigns a standard set of characteristics to decision makers that are covered by a substantive version of rationality involving a projective element. Decision makers, so neoclassical economic theory assumes, work with well-­defined problems and have a full array of alternatives to consider. Furthermore, they are assumed to have full baseline information as well as full information about the consequences of each alternative and the values and preferences of those affected by their decisions. Finally they are assumed to have adequate time, skill, and resources to make their decisions. It is upon these abstract assumptions that such decisions are optimal or rational. This strongly reminds us of some of the characteristics of the rational legislator portrayed above.

The real conditions under which decision makers act look quite different though. Most of the time, decision makers consider problems as “clear cut” or given while a more realistic view is that problems are “constructed”. Problems are indeed not self-evident which means that there is risk of failing to understand them. In addition, problems are often poorly or ill-defined due, among other things, to a lack of information about alternatives, incomplete information about the baseline or the background to the problem, the consequences of supposed alternatives or the range and content of values, preferences, and interests. Choices which we call “rational” are often based on incommensurate or ill-integrated goals.

Furthermore, decision makers are supposed to act deliberately in their search for solutions without being affected by emotions. The influence of the environment on decision makers is, most of the time, left out of view, just like the uncertainty under which they make their decisions. Their search for information is often incomplete, selective, and non-optimal while the marginal cost of relevant information may become prohibitively high, since the scarce resource is not information but attention.⁹ All of this, so it seems, does not come as a surprise. The surprise rather comes in the persistence of the belief in the decision makers’ omniscience involving the belief in the optimality of their decisions.

The more down to earth approach to rationality just outlined may profoundly shake our self-conception as rational agents. It provides us with a more realistic approach though as to how decision makers behave. Simon has coined the term “bounded rationality” for it, thus articulating the limitations of human rationality.

The main limitations to human rationality according to Simon are time, skills, and resources.¹⁰ Decision makers act in a context of time pressure. Their computational skills and attention are limited due to their restricted brain capacity. More people and more computers for example may help to overcome this restriction, though this does not in principle tackle it. In addition, resources, mainly information and time are scarce and the marginal cost of additional information needed to reach optimal decisions may be prohibitively high. Decision makers do not have a full and comprehensive view of the values of those subjected to their decisions, and they may indeed meet with unforeseen circumstances that affect their decisions.

Under the conditions of bounded rationality, decision makers in the real world strive for rationality in that they behave intendedly rational, as Simon puts it, but only boundedly or limitedly so.¹¹ Therefore they settle for less than the best in that they should content themselves with satisficing solutions instead of optimal ones. Satisficing solutions are reasonable solutions, in that they strive to achieve a balance.

What does the rational legislator look like once we dress him up as an agent who behaves intendedly rational while not always performing in such a way?

1.4.1 A Boundedly Rational Legislator as a Legal Agent

The combination of the in-between interpretation of the principle of charity and the idea of bounded rationality results in what was stated in the foregoing section and leads us to the thesis of the legislator as a boundedly rational legal agent. This amounts to considering him to be a rational agent although his performance is not always optimal. Yet, optimal solutions are not part of the real world; at best, they exist on paper. Optimality therefore only seems to be useful as a regulative ideal for evaluating the legislator’s rationality, not as a binary yardstick. Optimality from this perspective is conceived of as the outer end of a spectrum.

Following upon the adaptation of the standard of rationality to real world conditions, decision makers’ dealing with its complexity in an empirically justifiable manner is what characterizes the rationality of their actions. It is because the agents are boundedly rational that they have to make choices. If they were really omniscient, the very idea of a choice would never come to mind. The rationality of his choices then depends on how the agent copes with the complexity of social reality. How can action, that is, legislative decision making, by a boundedly rational agent like the legislator be framed so that it can be qualified as “rational”?

Boundedly rational legislators should not simply issue their rules to the world. What classical judge oriented theories of law seem to overlook is that legislation, in order to be rational, must not only be issued according to the rules of competence and procedure of the legal system. They must also “make sense in the world”. Making sense of normative texts as a matter of coherence at the level of interpretation as I have expounded above, has a counterpart at the level of legislative law making by the legislator. Once the assumption of rationality of the legislator is abandoned or at least tailored to the real world, it falls on the legislator to show how he makes sense of the complexity of the world.

This question triggers a variety of new questions. For the purposes of this essay, I will limit myself to the following: how does the boundedly rational legislator cope with the complexity of facts? Apart from the different facets of rationality that were briefly referred to above, this question focuses on what can broadly be characterized as the epistemic aspect of the legislator’s rationality. His cognitive openness to social reality while being normatively bounded to the Constitution calls for socially sensitive decision making that is apt to counter the discomfort felt with the law.

The boundedly rational legislator was characterized as an agent empowered by the Constitution to make law, yet with limited time, skills, and resources for doing so. Upon this limitation he can at best produce “satisficing” rules, that is, the best rules possible (1) rebus sic stantibus, (2) all things considered (3) now. This formulation includes the suggestion of the importance of the time dimension of the process of legislation.

1.4.2 Contingency

The epistemic aspect of the legislator’s rationality brings to the fore the time dimension of law. The legislator’s time perspective is replete with contingency since the future is to a large extent unpredictable. The legislator knows only probabilistically the state of affairs of the social world. Contingency points in two directions viz. a synchronic and a diachronic direction. Synchronic contingency refers to facts at the time of the preparation and promulgation of the legislative norm. Diachronic contingency for its part is slightly more complex in that it refers to facts and their change over time, including effects or facts resulting from legislative norms. Taking into account the effects of legislative law then points to the pragmatic dimension of the legislator’s rationality.¹²

Synchronic and diachronic contingency affect the rationality of law in that a legislator’s dealing with facts is part and parcel of rationally making law. Both affect legislative fact finding since the legislator has only probabilistic knowledge of facts due to the contingency of the world on the one hand and his bounded rationality on the other.

The distinction between synchronic and diachronic contingency in connection with the process of legislating allows us to distinguish the latter in the pre-legislative and the post-legislative phase. Synchronic contingency affects the legislator’s dealing with complexity in the pre-legislative phase. His framing of legislative norms is preceded by a process of fact finding in view of the problem which his norms aim at regulating. The formulation of a problem therefore must rely on an adequate description of relevant facts the clustering of which results in the situation that is held to be undesirable. The regulation of the undesirable situation_u purports to transform it into a more desirable situation_d upon connecting legal consequences to the occurrence of situation_u.

This connection is not causal in that situation_d will of and by itself occur as the result of the legal consequences attached to situation_u. In order to make this happen, a prognosis of the consequences is already in place. Prognosis of the consequences however again suffers from contingency since they are only probabilistic. Prognosis of the consequences, due to the probabilistic character of the connection between situation_u and situation_d requires a twofold critical appraisal of it. It first requires a comparison with alternative means to obtain situation_d. In addition, it requires a reasonable prospection of future circumstances, that is to say, a prospection of changing facts as well as facts as the result of the legislative norm or its effects.¹³

Due to diachronic contingency, what was considered rational or satisficing all things considered at one moment in time can become less rational or satisficing at a later moment. This can be clarified as follows. In addition to the other aspects of bounded rationality, a boundedly rational legislator acts among things under time constraint. This affects his initial fact finding as a matter of synchronic contingency. In addition to that, even if it may have looked “optimal” at the beginning, it may turn out to have been—or become—only “satisficing” after a lapse of time. Put differently, even if a legislator more or less successfully anticipates the future, he cannot predict with certainty what will happen after he has promulgated his norm. Time in other words is not at his disposal.

In this respect one may recall the familiar view that was outlined at the beginning of this essay. It characteristically makes abstraction of—synchronic as well as diachronic—contingency in assuming the legislator’s rationality. Facts are not the primary concern under the familiar view, far less time. One should recall in this respect the rationality assumption in neoliberal economics. On this assumption agents are rational and act in perfect markets with full transparency, no transaction costs, and no time constraints. Boundedly rational agents for their part, on Simon’s approach, behave intendedly rational, but only boundedly so.

They act in social contexts that are contingent upon the flow of time. As a consequence, the intention to behave rationally and behave only boundedly so, that is, without performing in a perfectly rational way, is not the result of some “lack” due to laziness, negligence, or bad will, but is due to the inherent contingency of social reality in which decisions are made and implemented.

If then, at the moment of its promulgation, a legislative norm is held to be rational, it is not because it is bestowed with rationality from a “one shot” a-temporal perspective. The latter reflects the product approach according to which law is omnitemporally rational upon the incarnation of Reason in its omniscient author. The process approach to legislation for its part which I am advocating here suggests that legislators act in a context that is inherently contingent and complex. The rationality of legislation then depends on how it is made, that is, on how its author copes with the complexity of the context in making it. Put differently, the rationality of legislative norms depends on how a boundedly rational legislator makes sense of complex social reality of which contingency is an inherent aspect. Legislation as a process faces the contingency of social reality as one of the aspects of complexity which a rational legislator has to cope with in a rational manner. This way of putting it amounts to stating that on the process view of legislation that I am advocating here, a rational legislator is required to keep track of his norms over time. Without exhausting the matter, and preferring focus to detail, I confine the argument here to the effects of legislative norms as a particular type of facts from a diachronic perspective. The point of my argument is to problematize the factual dimension of legislative norm making through the lens of its effects.

1.4.3 Diachronic Contingency: Effects of Norms

We should recall that the purpose of the essay is to make plausible that the legislator is no longer covered by the assumption of his rationality. Let us also recall in this respect that on the in-between reading of the principle of charity empirically justified criticism can be substituted for the a priori assumption of the rationality of the legislator.

In this respect an epistemically rational legislator is not simply assumed, but is on the contrary required to be aware of social reality. His awareness of social reality and his responsiveness to the problems he detects will however be affected by his bounded rationality.¹⁴ Boundedly rational agents strive for rationality in that they behave intendedly rationally but only boundedly or limitedly so, because they have, generally speaking, only limited time, skills, and resources. What does this mean?

Assume that the legislator has issued a satisficing norm at time t ₀. Both bounded and epistemic rationality can be considered incentives to follow up the norm once it is issued. As a boundedly rational agent, the legislator is to be aware of the mere satisficing character of his norm. This awareness is reinforced by his epistemic rationality. As an epistemically rational agent he has to show how he has effectuated his awareness of social reality over time, that is to say, at t ₀ as well as at t ₁, t ₂, …, t _n .

This focus on the effects of a norm is a way of “keeping track” of it. Effects of a norm are multifaceted and vary from desired or undesired, desirable or undesirable, positive or negative, short term or long term, symbolic or concrete, intentional or unintentional and various combinations thereof.¹⁵ For the purposes of this essay, I propose to categorize the factual effects of a legislative norm along three lines, that is, “efficacy”, “effectivity” or “effectiveness”, and “efficiency”.

Efficacy points to the fact that a legislative norm achieves the purpose that the legislator had in mind when issuing it, in that the state of affairs he aimed at realizing has been realized. Effectivity—or effectiveness—of a legislative norm for its part refers to its being followed and applied by legal agents and the judiciary respectively. There seems to be some confusion as to this notion since a norm that is effective from the perspective of the legal agents need not be actually enforced—or “applied”—by the judiciary and so would turn out not to be effective from that perspective. The confusion seems to stem from the idea that effectiveness only affects mandatory norms—or “primary norms of obligation” as Hart calls them.¹⁶ There seems to be good grounds however for holding that permissive norms—“secondary norm of private power” in Hartian terminology—can qualify as effective or ineffective. This would be the case where there is a norm permitting individuals to make a will and no one makes one. Efficiency finally points to a cost-benefit relation between a legislative norm and its effects.

Some interrelationships can be detected between these three categories of effects. Without a claim to exhaustiveness, it is easy to see that a legislative norm can be effective without being efficacious. In that case, it is followed and/or applied but without realizing the state of affairs the legislator was aiming at achieving. Conversely, an ineffective legislative norm can hardly be said to be efficacious. If it is not followed and/or applied, it will most probably not produce the state of affairs that the legislator had in mind. A norm can also be effective and efficacious, without being a formally valid legislative norm, as is the case with customary law (or rules of positive morality).

Two formally valid legislative norms may turn out to be “incompossible”, that is, when they mutually annul each other’s effects. In that case they can barely be said to be efficacious. When they are applied simultaneously, there is no effect at all. If one of them is applied and/or followed at the expense of the other, the former may be said to be effective, while the other is not effective and consequentially not efficacious.

Finally, an inefficient norm may require an unreasonably high enforcement cost that jeopardises its effectivity as well as its efficacy.

In addition to the foregoing observations, it must be mentioned that none of these effects is a matter of optimality but rather a matter of degree. Optimality is a dispositive concept in that a legislative norm is or is not optimal. Put differently, optimality is not context sensitive. When we broaden the context of legislative norm making to social reality, it makes sense to say that it is “more or less” efficacious, effective, or efficient. This is, as it seems, not only a matter of common sense. It likewise fits the qualification of the legislator as a boundedly rational legal agent as well as the contingency thesis concerning social reality.

The issue of the effects of legislative norms articulates the legislator’s cognitive openness to social reality as one of the core aspects of his epistemic rationality that is conditioned by his bounded rationality. Effects of legislative norms are the proper theme of social science investigations. Social scientists describe the law from an external point of view while including in their description the internal point of view of the agent, c.q. the legislator. Legislators as legal agents for their part adopt an internal point of view towards the norms of the constitution while being cognitively open to social reality as social scientists describe it.

An epistemically rational legislator must take into account the actual state of affairs in social sciences since the facts he deals with are not “brute facts”. They are not yet institutional facts either, since it is the normative order instituted by legislation that confers this status on them. The legislator’s dealing with social facts is therefore mediated by what social scientists have to say about them. The theoretical framework of this interaction represents what I have described elsewhere as the “reversed hermeneutical point of view”.¹⁷

An important aspect of the interaction between legislator and social scientist is legislative evaluation. As the justification of judicial decisions allows their control, so does legislative evaluation open up the possibility of a rationality control. It is to a large extent an empirical undertaking resulting in conclusions concerning the effects of legislative norms, that is to say, their degree of effectivity, efficacy, and efficiency.¹⁸ It is through legislative evaluation that the epistemic rationality of the legislator comes to the fore. Legislative evaluation provides the ruler with reliable knowledge as to whether or not the implementation of his norms has taken place as planned, whether the target group has behaved as predicted or ordered, whether the outcome indicators move in the “right” direction, and whether these changes can be plausibly connected to the legislative norm.¹⁹

Legislative evaluation offers an empirical assessment of the legislator’s epistemic rationality and allows for an empirically justified criticism of it. It shows in other words how the legislator has dealt with the complexity of the social world in order to make sense of it.

The last stage of my argument will focus on the impact of the legislator’s dealing with facts on the validity of legislative norms.

1.5 Legal Validity

Legal validity is a multifaceted concept. Generally speaking, a distinction can be made between formal, factual, and axiological validity. Theories of legal validity can be arranged according to the weight which they attach to one criterion of validity. Natural law theories stress the importance of a substantive criterion of validity, while realist theories most of the time adopt a factual criterion. Formalist theories, finally, focus on a formal criterion of validity, mainly conformity with other—higher—norms of the legal system.

It seems that while all theories of validity favour one criterion, none of them is limited to a unique criterion. Natural law theories—typically Aquinas’ version—while adopting the substantive criterion of justice, include a formal criterion as well in that only “those who are in charge of the community” can make laws.²⁰ Realist theories for their part—typically following Holmes dictum that “the prophecies of what the courts will do in fact, and nothing more pretentious, are what I mean by law”²¹—adopt a factual or material criterion of validity, this being the actual behavior of the judiciary. Yet, the legal organization of judiciary precedes the latter’s normative activity. As a consequence, despite legal realism’s supposedly general rule skepticism in favor of judicial decisions, the norms organizing the judiciary are independent of the latter, and can be qualified as formal. A similar argument can be made for Scandinavian realism, which I will not focus on here.

Formal theories, finally,—typically Kelsen’s Pure Theory—endorse a formal criterion of validity in that a norm is valid if and only if it corresponds to a higher norm. Kelsen however includes an additional factual criterion as well since a norm’s validity is conditioned by its “Wirksamkeit”. Since legisprudence adopts the social sources thesis relying on a positivistic concept of law, Kelsen’s approach to validity seems to include a fruitful suggestion for our exploration of the rationality of legislation.

Kelsen articulates his concept of validity by distinguishing between the subjective and the objective meaning of an act. The act issuing a norm is an act of will and has the subjective meaning of a norm. The agent expresses his will that other agents behave as he wants them to behave. The act of will has the objective meaning of a norm if and only if another norm authorises the norm issuer to act in that way.²²

He stresses the fact that no other elements stemming from psychological, sociological, ethical, or political theory²³ interfere with his concept of legal validity, thus guaranteeing the purity of law and its “ought”-existence. The validity of “lower” norms is guaranteed by their being founded upon existing higher norms, the validity of the constitution of the legal order being derived from the legal system’s basic norm. As a transcendental-logical proposition the latter is a presupposed, or thought norm,²⁴ a credo involving the belief that the legal order is valid and that it must be obeyed and applied for that reason.

In one of its versions the basic norm reads: one ought to obey the prescriptions of the historically first constitution.²⁵ Since the basic norm is not a created norm, it must be considered as the constitution in the logical sense. A legal system then consists of all the norms issued on the basis of higher norms belonging to the same system, the unity of which is safeguarded by the basic norm.

If things are clear under Kelsen’s explanation of the formal validity of legal norms, they may seem to be obscured by his supplementing formal validity with the additional criterion of Wirksamkeit. While he contends that norms are valid or exist upon their being created on the basis of a higher norm, a valid norm that is not usually applied or obeyed loses its validity.²⁶ In other words, formally valid norms that are not followed by legal subjects or applied by the judiciary are no longer considered valid. Although formal validity and Wirksamkeit are analytically distinct concepts, Kelsen points to an essential relation between them, in that “a coercive order presenting itself as the law is regarded as valid only if it is by and large effective”.²⁷

Wirksamkeit as a condition for the validity of single legal norms and for the legal system as a whole is a condition for the validity of the constitution as well. A constitution that is not generally obeyed or applied loses its validity. Upon this conclusion, the basic norm reads as the reason for the validity of a legal order, that is, the reason why one ought to comply with an actually established, by and large effective, constitution, and therefore with the by and large effective norms, actually created in conformity with the constitution.²⁸

What makes Kelsen’s concept of validity interesting for our purpose is that formal validity is said to operate as the main criterion for a norm’s existence, while effectivity operates as a condition for a norm’s—continued—validity or existence. This idea of Kelsen’s calls for further exploration. In the foregoing pages “effectiveness”, “efficacy”, and “efficiency” were identified as the factual effects of a norm. Generally speaking, they can be classified under the heading of Wirksamkeit.²⁹

Kelsen’s view suggesting that Wirksamkeit conditions a norm’s—continued—validity or that the effects of a norm affect its validity, becomes more interesting once we compare it with the idea of the boundedly rational legislator.

As for the Wirksamkeit of a norm, a number of aspects come to the fore. Legislative norms can be poorly drafted. When this is the case, legal agents and the judiciary may meet with difficulties in understanding what pattern of behaviour the norm requires so that its effectiveness risks being undermined. Apart from being poorly drafted, norms can also become obsolete when they are no longer adapted to changed technical, social, political, and economic circumstances. This phenomenon is generally recognized as desuetude. It seems that Kelsen’s concept of validity points to these aspects, in so far as they contribute to the Unwirksamkeit of a norm. In connecting a norm’s Wirksamkeit to its existence, Kelsen only appears to commit adultery with social sciences.

His awareness of the value of a formally valid norm becoming unwirksam suggests that a norm’s existence or validity includes a temporal dimension that covers the manifestation of its effects over time. This idea becomes more significant once we get used to the idea of the boundedly rational legislator. His limitedness in time, computational skills, etc. are unknown to Kelsen but it seems to fit his idea of a norm’s validity. It draws attention to the fact that norms can become unwirksam because they are not preceded by a sufficient analysis of social reality, because they are poorly drafted, insufficiently monitored, or not corrected in time. Other aspects are that legal norms interact with each other, and produce unforeseen—even unforeseeable—effects at the time of the norm’s promulgation.

Kelsen’s idea of validity seems to include some promises of a legisprudential concept of validity when we bear in mind that the legislator is a boundedly rational legal agent. It should be recalled once again that a boundedly rational decision maker in the real world strives towards rationality in that he behaves intendedly rationally, as Simon puts it, but only boundedly or limitedly so. No rational agent can, in any meaningful sense, claim to be omniscient. This includes the suggestion that to be rational involves the awareness of the limitations or bounded character of one’s rational capacities.

Taking into account the distinction between the product and the process approaches to legislation, rational law making can be understood as “making rational law” on the one hand and “rationally making law” on the other. Making the law more rational in other words points to the larger perspective of making it more rationally. Within this larger perspective, a boundedly rational legislator who is rationally making law aims to make effective, efficient, and efficacious law. His knowledge about the future effects of his norms however, is only probabilistic in that he does not know in advance whether and to what degree his norms will be effective, efficient, and efficacious. His intention to make rational—effective, efficacious, and efficient—law may be frustrated by the—negative—effects of his norms or by unpredictable changes of circumstances in the real world.

The empirical investigation of the effects of legislative law may then allow one to describe and explain the possible discrepancies between the effects aimed for and the real or empirical effects that occur in the world. This then shows that the rationality of law is not only a matter of its optimization through interpretation. It is also a matter of rationally making it. Optimization of legislation is not identical to making it optimal. It simply means making it more satisficing or as satisficing as possible.

Kelsen’s thesis on the validity of law now shows its fruitfulness for legisprudence in the following way. The thesis holds that formally valid norms, that is, norms formally validated at the moment of their promulgation, cannot be a priori assumed to be effective, efficacious, or efficient. Since ineffectiveness, inefficacy, or inefficiency can negatively affect a norm’s validity, the process of legislation does not stop at the moment of a norm’s promulgation. In order to make rational law, the law must be made rationally throughout the process of legislation. The process of legislation therefore extends to the norm’s entire existence and requires keeping track of its effects. A norm, according to Kelsen, exists if and only if it is valid. It follows that the validity of a norm is not a matter of all or nothing in that validity would be bestowed on a norm at the moment of its promulgation.

It is here that the pragmatic dimension of the legislator’s epistemic rationality comes into play. Since the legislator’s duty is to make valid norms his legislative activity is not limited to acting according to the norms of the legal system empowering him to do so. This is a necessary requirement for a norm to be valid at all. In addition to that however, legislating rationally requires him to take care of the Wirksamkeit of his norms as well. Put differently, formally valid norms require constant reconfirmation or validation, and possibly correction, in order to meet a minimal degree of Wirksamkeit for a norm to remain valid. Norms falling below a certain degree of Wirksamkeit can no longer be said to be valid. On the legislator’s duty to create valid law, this duty includes upholding the validity of the laws in question in terms of their Wirksamkeit.

1.6 Legal Validity and Rationality Review

The foregoing considerations suggest that the evaluation of legislation, describing and explaining its effects, must be part and parcel of legislation as a rational process. The analytical part of our exploration has come to an end here. In the last section of this essay, I will briefly explore how this analytical part can be utilized in practice. Due to my own bounded rationality, this part will be short and mainly focus on a sample of decisions of the German Bundesverfassungsgericht (BVG) that seems to adopt the process view of legislation.

If our investigations have established that it is plausible that the legislator is tasked with keeping track of his norms upon—but not necessarily limited to—legislative evaluation, courts should not simply defer the issue to the legislator, taking him at his word. This would bring us back to the logic of the familiar view in assuming that the legislator is rational and has acted rationally. Neither can a court strike down a norm merely on its own evaluative findings de novo or interfere with the policy choices made by the legislator. This would put the court on the substantive review track, and risk the court being confronted with a political question and interfere with the separation of powers. Does this mean that courts have no legitimate grip on the rational making of law?

As I have argued throughout this essay, according to the process approach to legislation proper to legisprudence, it is up to the legislator to show how he has come to his decision. This includes, among other things, him showing how he has proceeded in fact finding, how his choices have been made and how he has effectuated the prognosis of changing circumstances as well as the effects of his norms. Call this the pre-legislative stage. After the promulgation of his norms in what can be called the post-legislative stage, this requires that he shows how he deals with the effects of its norms which includes how he has taken into account changing circumstances after the norm’s promulgation. This also means that he must show how legislation has been evaluated as a way of keeping track of norms, and how he deals with the results of such evaluation.

The way in which a court has a legitimate grip on the rational making of law amounts, then, to a rationality review which is different from a mere procedural review as well as a substantive review. The type of rationality review which courts can legitimately proceed with represents a marginal control of the rationality of the process of legislation. The marginal rationality review is critical of the court’s de novo findings as well as of a mere deference of legislative fact finding to the legislator. Its specificity is that it consists of a meta-evaluation of the factual evidentiary evaluations provided by the legislator throughout the legislative process. A marginal control of the rationality of legislation focuses less on norms as a product than on the process of which norms are the product. This process, as I have suggested, includes the whole life of the norm, that is, the pre-legislative as well as the post-legislative stage.

It is generally agreed that rationality review of legislative action scrutinizes the rational relationship between a legitimate state purpose and the means used for its accomplishment. So this rationality test comes down to scrutinizing whether the purpose of a norm is legitimate and whether the means to achieve it are rationally connected to it. Rationality review thus conceived amounts to a test of non-arbitrariness of a norm. It is not conclusive though as regards its rationality once this is conceived as the rationality of the legislative process of which it is a product. While arbitrary norms cannot be considered rational, non-arbitrary norms are not rational for that reason alone. If the legislator is assumed to be rational, this simple version of the rationality test will suffice. Since I have been endeavoring to show that the legislator is a boundedly rational agent, a more robust version of rationality review is required.

Marginal rationality review is such a test in that it requires that the use of power must be justified in showing how it has been exercised. This model of review leaves intact the discretionary power of the legislator to make choices. Marginal rationality review does not lead to a substitution of a Court’s choices for the legislator’s. It severs the choices and the power to make them from how they are made. This comes down to justifying how the power to make choices is exercised. The fact of the matter is that legislative norms suffer from a rationality gap as long as it is not justified how legislative decisions are made. The justification of how legislative choices are made is what rationally making legislative law amounts to. The assessment of the justification that fills the rationality gap is what marginal rationality review comes down to.

The following sample of cases of the German Bundesverfassungsgericht (BGV) shows the Court’s particular sensitivity to this problem. In adopting, it seems, a process approach to legislation, the BGV interprets art. 20, 3 of the Constitution³⁰ in such a way that the legislator must pay attention to the cleaning up or the modernizing of obsolete norms.³¹ As a matter of fact, the Court adopts the view that legislative norms are time related. Norms therefore can become unconstitutional over time³² or “move into the direction of unconstitutionality”, yet legislators must be given time to adapt a norm before declaring it unconstitutional.³³

Since initial legislative fact finding is the prerogative of the legislator according to the BGV, the Court respects this freedom. Yet, factual presuppositions that are patently incorrect may result in the unconstitutionality of a norm.³⁴ The legislator’s independent sphere of decision-making is according to the BVG transgressed when his factual considerations are so patently inadequate or incorrect that they cannot reasonably support the legislative norm.³⁵ The BVG has decided for example that, since the legislator has a duty to protect unborn life, it falls within his competence to collect and assess the relevant data affecting the issue. Reliance on mere statistical data on the number of abortions, the number of births, etc. was insufficient according to the BVG since reliance on such data precludes an adequate evaluation of the effects of the norm.³⁶ Initial legislative fact finding may turn out to be insufficient, however the legislator must be given time to collect relevant data regarding the effects of his norms. When empirical material contradicts the initial assessments and the legislator has failed to react adequately, the BVG has regularly held that the norm has become unconstitutional.³⁷

Another issue concerning facts are the prognoses made by the legislator. In this respect too, the BVG assesses the evidence produced by the legislator.³⁸ On a number of occasions the Court’s assessment of the legislator’s prognosis is based on a more intensive control of the norm’s content.³⁹ The legislator’s prognoses must be justifiable⁴⁰ which may require a correction of the norm.⁴¹ An apparently correct prognosis by the legislator may later turn out to be entirely or partly wrong which affects its constitutionality,⁴² thus requiring its adaptation according to art 20, 3 of the Constitution.⁴³ The Court has therefore decided that the legislator must systematically take into account and assess the available sources of knowledge in order to evaluate as accurately as possible the conceivable effects of his norms in view of correcting them when essential changes occur.⁴⁴

So for example the BVG has proceeded to carry out an evaluation of the legislator’s prognosis in saying that the effects of a tax regulation could not yet be fully assessed at the moment of the promulgation of the norm.⁴⁵ Yet changing facts may require the legislator to come up with additional support to uphold a norm.⁴⁶ A norm that was constitutional at the time of its promulgation may become unconstitutional due to changing facts.⁴⁷ When, for example, interest rates are increased only every 2 years and not more frequently due to changing facts, the BVG has not held the norm to be unconstitutional since it leads to inefficacy.⁴⁸ In a similar vein the BVG has decided that a change in the population makes an electoral circumscription unconstitutional without, however invalidating it. The legislator was instead tasked with updating the norm.⁴⁹

In this respect, the legislator must keep abreast of changing circumstances in the social reality, and check whether his original ruling can be upheld in view of such changes.⁵⁰ The Court allows the legislator a period of time during which he can collect relevant data concerning the effects of his norms⁵¹, after which the legislator must show that the proposed goal of the norm can be reached as was initially planned.⁵²

1.7 Conclusion

In this essay I started from the familiar view, which includes a specific approach of the legislator’s rationality. He was compared to an omniscient agent, and I have qualified this in terms of a boundedly rational legislator, referring to the principle of charity. A reinterpretation of the principle of charity opened the way to an alternative interpretation of the legislator as a rational agent. Boundedly rational agents lack time, information and computational skills, and therefore their decisions cannot be “optimal” but at best “satisficing”. This occurs, so I have argued, in the field of legislation. When legislators are no longer assumed unboundedly rational, their legislative work can be taken into account from a different angle. They are presumed rational, and no longer assumed to be rational. Legislators act under conditions of contingency, they have no privileged access to reality nor to its future changes. The articulation of the bounded rationality of legislator throughout this essay opens a new avenue on the legislator’s duties to fulfill throughout the process of legislation. More specifically, it was argued that they should keep track of their norms. They do so by assessing the effects of their norms in social reality, following upon their initial duty of fact finding at the moment of the promulgation of legislative norms. This thesis was illustrated by a sample of case law from the German Bundesverfassungsgericht that seems to adopt the main lines of the legisprudential theory that I have elaborated in earlier work.