19. Making Law

The legal systems we have looked at demonstrate some of the ways in which rules come into existence and change over time. This chapter is on the implications of each. As we will see, there is no simple one to one relation between legal systems and ways of creating law. The current U.S. system is arguably created and changed in at least three different ways, Jewish and Muslim law in four.

Divine Inspiration

At least two of the legal systems we have looked at are viewed by believers as based on divine commands—Jewish law on the Torah, written and oral, Muslim law on the Koran and hadith. This raises a number of problems.

One is the risk of legal inconsistency growing out of inconsistent versions of the authoritative text. In the case of the Koran, an official text was established under Abd al-Malik (65/685—86/705618) and all variant versions destroyed. A similar process may have occurred with the written Torah at an earlier date.619 The problem remained with the oral Torah, instructions believed to have been given by God to Moses on Mount Sinai and passed down from him in a chain of oral transmission. The risk that different scholars would hold different views on its content was dealt with by the doctrine, based on the written Torah, that in case of uncertainty one should go with the opinion of the majority, interpreted as the majority of the Great Sanhedrin.

The Great Sanhedrin gave its last binding decision in 358 A.D. Its role was eventually taken over by the Geonim, the heads of the two Babylonian academies, which were at the time the leading bodies of Jewish scholarship. They received queries on legal and religious questions and sent back responsa, answers accepted as authoritative. Their authority weakened in the tenth century and ended, with the Geonim themselves, in the eleventh. Thereafter there was no single authority accepted throughout the Jewish world, although the views of a leading scholar might be accepted across many communities. Within any single community the communal authorities determined the law, if possible with the aid of one or more religious scholars. Since most disputes were intracommunal, uniformity within a community was usually sufficient.

A second problem is the risk that someone will claim legal authority by persuading others that he has his own pipeline to God. The story of the oven of Akhnai, discussed in chapter 4, can be read as a rejection of all such claims. Rabbi Eliezer provides the strongest possible proof of divine support, a string of miracles followed by a voice from heaven. The sages remain unconvinced, holding that the law is determined by the view of the majority. Taken literally, the story is puzzling; one might expect that the sages, after hearing God’s opinion on the question, would change their votes. But it makes sense as a forceful rejection of the claims of purported miracle workers. God has transferred responsibility for determining the law to the human authorities.

The solution for Sunni law was the doctrine that Mohammed was the final prophet, hence there could be no further Koranic revelation. There were occasional attempts by later figures to claim prophethood, including at least one case of claimed authority for a new Koran, but they were rejected by the bulk of the Islamic community. Shia Muslims believed that their imams were, like Mohammed, divinely inspired, their practice like his evidence of God’s will. That source of law ended for the Twelver Shia, the largest group, in the tenth century when the final Imam went into occlusion to reappear only in the final days.

While the creation of an authoritative text of the Koran eliminated one source of disagreement, a second remained. Islamic law was also based on the Sunna of Mohammed, the practice of the Prophet as recorded in hadith, traditions. Different scholars could, did, and do disagree as to which were genuine, which doubtful. Eventually several authoritative collections were created, collections of hadith widely believed to be well supported, reducing but not eliminating that source of disagreement.


God never answers all the questions that matter. Divinely inspired law, like legislated law, requires interpretation; there is no sharp boundary between interpretation that fills gaps or clarifies ambiguity in religious law and interpretation that makes new law. In the Jewish case, interpretation occurred first by vote of the Greater Sanhedrin, later as decisions and treatises by scholars whose authority was based on their reputation. The opportunity for further interpretation was kept open by the doctrine of hilkheta ke-vatra’ei, “the law is in accordance with the views of the later authorities.” Eventually the Bet Yosef and Shulhan Arukh of Joseph Caro, along with associated commentary, came to be accepted across most of the diaspora.

Islamic scholars differed on the rules for interpreting Koran and Hadith, hence on their legal conclusions. The process, for Sunni Muslims, was conducted through the orthodox schools of law, eventually reduced to four. Different Muslims might choose to accept the interpretation of different schools, although often one school was dominant in a region. Within a school there might be disagreement on some points among different authorities, leading to rules within each school for choosing among the interpretations based on the relative status of different scholars. All four schools accepted the doctrine of consensus, according to which an interpretation which was at some time accepted by all legal scholars is thereafter considered certain, on the basis of hadiths reporting Mohammed as having said, in one or another form, that his people would never be agreed on an error.

Both Jewish and Islamic law, in their original forms, permitted a husband to divorce his wife but did not permit a wife to divorce her husband. Jewish law eventually changed to give the wife the de facto, although not de jure, possibility of divorce: If she could persuade the court that she had adequate cause, the court could compel the husband to divorce her. Islamic law permits a husband to delegate the power to divorce him to his wife as part of the marriage contract or a later agreement. It is also possible for a wife to request a divorce in exchange for compensation. Details vary across the schools.


The mechanism for producing law most familiar to moderns is legislation. In Imperial China the authority was the Emperor, in Athens the assembly, in Jewish law communal authorities. Amish legislation, change in the content of the Ordnung of a congregation, is proposed by the bishop but requires the congregation’s unanimous acceptance.

The oddest case of those we have looked at may be the Cheyenne. In at least one case, rules adding to customary law by declaration of a single soldier society following its resolution of a dispute were seen as binding in the future on all Cheyenne. A somewhat similar pattern among the Romani, specifically the Vlach Rom in America, was the holding of a diwano or kris to agree on a modification of Romania.

While legislation can be used to resolve ambiguities in the law, that did not always suffice, not even in Chinese law, designed as a complete mapping from offense to punishment. Someone has to decide how to deal with cases that do not exactly fit the rules. The boundary between resolving ambiguity and creating law is rarely a sharp one, as illustrated by the expansion, in U.S. constitutional law, of congressional authority to regulate interstate commerce into something close to a blank check to regulate all economic activity.


Another approach to creating law is binding precedent. A court’s decision is binding on future courts below its level, determining law not only for the case being interpreted but for future cases.

Precedent has several advantages over legislation. Since it grows out of the resolution of a large number of cases, it generates a more detailed set of legal rules than is likely to be produced by either a ruler or a legislature. Because it originates with real cases, it maps more closely to real issues than rules created in the abstract. And because the rules it creates change as new cases are resolved, with courts able to overrule their own past precedents and higher courts free to overrule the precedents of lower, it creates a legal system that adjusts to changing circumstances.

Binding precedent has disadvantages as well. Since it is developed simultaneously by multiple judges, it may produce inconsistent rulings. Until and unless the inconsistency is resolved by a higher court, litigants find that the outcome of litigation depends on the choice of what court to litigate in. In systems where it is in the self-interest of judges to have cases come to them, the result is to bias legal rules in favor of whichever party controls the choice of forum. Daniel Klerman has argued that English common law had a pro-plaintiff slant until the nineteenth century because judges received fees for each case and it was the plaintiff who decided which of several alternative courts the case went to. Statutes in 1799 and 1825 eliminated fees as a source of judicial income. The result, according to Klerman, was to shift common-law rules in a pro-defendant direction.

Something that some will see as a bug and some as a feature is the slow response to change of a system of precedent. Judges reach a high level late in their careers and remain there until forced to retire by old age or death. Old judges may base their decisions on the views and values of their youth. Legislators, even aging legislators, are under pressure to support positions popular with the current electorate.

Customary Law and Non-Authoritative Precedent

For a different approach, consider Somali law as described by Van Notten or pre-Islamic Bedouin law as described by Schacht. Judges are arbitrators, their verdicts opinions about the law not decisions creating it. That multiple arbitrators see the customary law as implying a particular resolution to some legal issue is evidence that it does imply that resolution, especially if the verdicts of those arbitrators were viewed as just and if those who produced them continue to be chosen as arbitrators. Past decisions influence future decisions as evidence but not as binding precedent.

For a modern example of the same logic, consider Annie Lee Turner et al. v. Big Lake Oil Company et al. The Supreme Court of Texas had to determine what compensation, if any, Big Lake Oil Company owed Annie Lee Turner and her neighbors for damage done to their property by the escape of polluted water from Big Lake’s storage pools. A key issue was whether Big Lake was liable only if it was negligent.

The Texas judges looked first not to the statute law of Texas but to a case decided in another country almost a century earlier: Rylands v. Fletcher, decided by a committee of the British House of Lords in 1868. They spent a considerable part of their written opinion explaining why they reached a different result.

The law lords who decided Rylands v. Fletcher were both a court deciding the outcome of a particular case and a body of experts interpreting the common law of England. In that second role the conclusion they reached was relevant to, although not determinative of, the deliberations of a Texas court deciding a similar case based on the same underlying system of law. An English judge has authority over an American case in precisely the same sense in which an English scholar has authority over a dispute in his field carried on by American scholars; he is an expert whose opinion is relevant to deciding a disputed question of fact. Seen from this point of view, the common law functions as a system of binding precedent within a single national legal system and a system of informative precedent more broadly.
For another mixed system, consider Amish law. The final decision that determines the content of a congregation’s ordnung is legislation, proposed by the bishop and accepted by the unanimous consent of the congregation. But the process that changes the ordnung looks more like the process by which customary law is changed. Someone does something arguably in violation of the ordnung as currently interpreted. If others object, the congregation will probably find that what he did is forbidden. He may have to publicly confess to a misdeed. But if few or none object and others imitate him, the view of what is permitted may change and the changed view be incorporated in the next revision of the Ordnung.

Works such as the Mishnah Torah of Maimonides claimed not to create law but to discover and describe it. Halakhic authorities were offering their interpretation of an existing body of law that claimed divine origin. Their opinions were not binding as precedent—that would be inconsistent with both the rule that the law was according to the opinion of the later authorities and the actual practice of legal scholars. One of the criticisms of Maimonides’ work was that, because to each legal question he gave only one answer, he failed to provide a judge with the opinions and arguments of past scholars needed to reach his own conclusion.

We know where legislated law comes from, where judge-made law in a system of binding precedent comes from, where law based on divine revelation comes from. It is less clear who creates customary law or how. As Justice Stewart said of obscenity, you know it when you see it.

Consider a form of customary law with which we are all familiar. While getting dressed before going to class yesterday I went in search of a clean pair of pants, the previous pair having just gone into the washing machine. Hanging in my closet was a worn pair of blue jeans. I gave no serious thought to wearing them, although they would probably have been more comfortable, perhaps also warmer, than the slacks I ended up with.

The reason is that doing so would have violated customary norms of behavior for professors teaching courses. Norms are a form of customary law. Nobody legislated those rules, nobody legislates their changes, yet they exist. The mechanism of change is much the same as among the Amish. A particularly brave and self-assured professor chooses to push the boundaries, to appear before his class in shorts and flip flops—I have a real example in mind. If students and colleagues look at him oddly, if he notes that at the next faculty meeting his opinions carry less weight than in the past, if his wife mentions that a colleague’s wife asked her if her husband was going through some sort of stage, he may abandon his norm-violating behavior. If nobody seems particularly shocked, if colleagues react, if at all, with friendly amusement, he continues it. After a while—a week, a month, perhaps a year—another professor shows up in shorts, possibly after discovering at the last minute that all his pants either are in the wash or should be. The norm changes.

Social norms are a form of customary law with which all of us have first-hand experience. They may be the original source of all customary law. Individuals behave in a certain way, constrained by the expectations of others. When a dispute arises, an arbitrator rules on the basis of what behavior is or is not accepted in that society. Norms, enforced for the most part by social pressure, evolve into customary law enforceable by violence or the threat of violence.

Mix and Match

As some of these examples show, real systems often mix multiple approaches to making law. The ordnung of an Amish congregation can be seen as legislation by unanimous assent, but the process by which it changes looks more like customary law. The Anglo-American legal system consists of both legislation and common law, the latter a system of precedent, but with decisions largely based on judges’ perception of customary law. Somali law was mostly customary but in part Islamic law, seen as based on divine revelation. Islamic law was based on revelation, but revelation interpreted by mujtahids, religious scholars, and supplemented by legislation such as the Ottoman kanun.

Flexibility: Solution and Problem

All of the systems I have described are to some degree flexible, whether by legislation or interpretation. That provides a potential solution to two problems—mistakes discovered in the original law code and changes in circumstances relevant to what the legal rules should be. But flexibility also raises additional problems. One is that it makes law less predictable. The more flexible the legal system, the less certain someone making decisions today can be as to what legal rules will apply to him tomorrow. Another is that those who expect to gain or lose by a change have an incentive to spend resources supporting or opposing it. In the modern U.S. system, those expenditures take the form of lobbying for and against legislation, litigating in the hope of establishing a favorable precedent, and efforts to influence the appointment of judges, especially Supreme Court justices. In Islamic law, where legal rules were in part based on hadith, “in nothing do we see learned men more prone to untruth than the fabrication of traditions.”

Making Good Law

I have so far ignored the most important question of all: To what extent can we expect one or another of these mechanisms to produce good law, law that promotes the welfare of those to whom it applies?

If divinely inspired law is the work of a benevolent deity, the result should be good law, but I have my doubts. Similarly if legislated law is created by a wise and benevolent ruler, but I know of no reason to expect that either. Believers in democracy might argue that legislated law will be designed in the interest of the voters since otherwise the legislators will be voted out of power, but I find that claim equally unconvincing. Distinguishing good law from bad is not easy. Individual voters, knowing that their vote has little effect on political outcomes, devote little effort to gathering the information needed to vote wisely. Politicians rarely help out by labeling themselves as bad guys or their bills as bad law. The result, as I have argued elsewhere, is a system that frequently produces bad law, a theoretical conclusion for which I find a great deal of empirical support.

Consider the case of trade restrictions. Economists have known for about two hundred years that a country that imposes tariffs on imports will under most circumstances make its inhabitants worse off by doing so. For that entire period, most countries, including most democracies, have imposed tariffs. While there are theoretical arguments that under certain special circumstances the general result does not hold, the tariffs actually imposed do not fit the pattern those arguments imply; protection is typically provided not to infant industries but to senile ones.

What about judge-made law? Richard Posner has argued that common law tends to be economically efficient, supporting that argument with extensive economic analysis of the efficiency of common-law rules. In my view, neither the evidence nor the theoretical arguments offered by him and others is sufficient to establish the truth of his claim; interested readers will find the question discussed at length in my Law’s Order.

Which leaves us with customary law.

Consider a norm which, if adopted by a group of individuals for interactions among its members, makes them better off—say a norm of honest dealing. Once one group has adopted it and others have observed the results, we would expect them to either join that group or imitate it. Thus desirable norms could spread through a society. Once established, they could become the basis out of which customary law develops.

There is an important limitation to this argument. In Order Without Law, Robert Ellickson described how nineteenth-century whalers developed an efficient set of norms with regard to issues such as what happened when one ship harpooned a whale and another eventually killed it, norms that efficiently changed as the whalers shifted from hunting one species of whale to another.

The reason they shifted from one species to another was that one species after another was being hunted to near-extinction. That suggests that they would have been better off if they had also had a norm that restricted the number of whales killed to a number that would not seriously reduce the population. No such norm existed.

A norm of restricted whaling benefits all whalers if they all follow it, but the benefit from my restraint goes to other whalers whether or not they act similarly. The right rule for all is to restrict. The right rule for any individual is to free ride on the restrictions of others. The problem is strictly analogous to the familiar case of the prisoner’s dilemma, more generally market failure, situations where individual rationality does not lead to group rationality.

It follows that customary law based on social norms ought to do a good job of generating the sort of rules that benefit the members of a group whose members adopt them but not the sort of rules which make us all better off if we all follow them.

Competition in Law

After I have signed a contract with you, I would prefer law that favors me in any resulting dispute. But before we have signed the contract, we have a common interest in law that maximizes our combined benefit, increases the size of the pie to be divided between us. That suggests that a legal system in which individuals get to choose in advance the judge, arbitrator, or court that will interpret their contract will tend to produce good law, at least for the contracting parties. That was to some degree the case in the traditional Islamic system, where parties could choose to set up their contract with a court following the school of law whose view of contract law they preferred. It is to some extent the case for modern U.S. corporate law, which is in part a definition of the terms of a contract among the stockholders of a corporation, since the original creators of a corporation can choose what state to charter it in. And it is true of the non-state law created by mechanisms for binding arbitration.

The same principle applies on a smaller scale as well. A contract is itself a miniature legal system, a set of rules agreed on by the contracting parties for controlling their interaction. It is in the interest of the contracting parties to choose rules that maximize their total gain from the interaction.
The articles of the pirate ships described in Chapter 7 provide another example of good law—good for the pirates if not their victims—generated by the same mechanism.

Добавить комментарий

Ваш адрес email не будет опубликован. Обязательные поля помечены *