1

Expressive Claims about Law

Legal scholars and political actors make broad claims about the expressive dimension of statutes, judicial opinions, prosecutorial decisions, jury verdicts, and criminal sentencing decisions. Expressive theories span topics as diverse as rape shield laws, property law, patents, the regulation of CEO compensation and corporate directors’ duties to creditors, the legal concepts of the appearance of impropriety and standing to sue, cyber-stalking regulation, the fourth amendment exclusionary rule, international law, and bank regulation.1
The nature of the claims varies greatly. Commentators say that legal change will “send a message” of some sort, for example, that legalizing medicinal marijuana will convey to teenagers the harmlessness of smoking pot, that strengthening drunk driving or domestic violence laws will articulate the grave harm those behaviors cause, and that a jury verdict in favor of a rich but undeserving defendant will communicate a message of unequal justice. Some legal scholars claim that people comply with certain laws, such as seat belt mandates and smoking restrictions, because of the expressive (or symbolic or educative) effect of these rules.2 Or that the law can work by changing the “social meaning” of a behavior, as where the historic introduction of a law against dueling created a new meaning to the decision to refuse a duel: not that one was a coward without honor, but that one felt an honorable duty to obey the law.3
Legal scholars also offer to explain the existence of certain laws by their symbolism. For example, there may be little or no behavioral effect from the legislative repeal of constitutionally unenforceable segregation or sodomy laws, from laws recognizing English as the “official” language, or from local ordinances declaring a nuclear-free zone, but voters may demand and support such laws for the values they are understood to express. Even when a law has some behavioral effect, the politics of its enactment may be dominated by its symbolic importance. Pro-life voters may like what a law against “partial-birth abortion” expresses even if the law, by permitting other abortion procedures, has no effect on the total number of abortions.4 Commentators explain legal rules of market inalienability such as the ban on selling human organs, sexual services, or electoral votes by the public’s desire to express the incommensurability or pricelessness of certain values.5
There are also many normative claims about legal expression. Some constitutional theorists claim that the Equal Protection Clause of the Fourteenth Amendment should be read to forbid laws that express the inferiority or subordination of a racial group or sex, because such laws create “expressive harms” regardless of whether there are further consequences from the law.6 The Supreme Court arguably took an expressive stance when it ruled that the clause prohibits racial gerrymandering that creates bizarrely shaped electoral districts because it sends a “message” that “members of the same racial group . . . think alike, share the same political interests, and will prefer the same candidates at the polls,” and when it ruled that the clause prohibits public single-sex higher education and gender-based jury selection because they reinforce gender stereotypes.7
Not surprisingly, the meaning of the term “expressive” is not constant across the scores of articles discussing all these expressive claims. What might at first appear to be a single legal literature about the expressive theory of law is really a set of distinct literatures employing the same term. There is no grand unified “expressive theory” that encompasses all of these literatures. This observation is not a criticism of any of the expressive literatures. Instead, the word “expressive” has enough flexibility to cover an array of inquiries. One can believe in the value of each inquiry without thinking that the inquiries are essentially the same. Thus, I have no interest in trying to limit the use of the word “expressive” to the use I make of it and no quarrel with legal scholars describing their theories as such even though they differ from the ones that concern me. (Indeed, if the reader is unable to tolerate my using the term expressive in this book as I do, please feel free to imagine a different term, perhaps “communicative” or “educative.”).
The contrary assumption—that there is a single field of “expressive law”—has generated confusion. Without a clear understanding of the distinct projects using the term “expressive,” there is a tendency to lump together articles that address different topics. To avoid further confusion, I am at pains to observe that this book concerns, at least primarily, only some of the expressive law literatures. To explain, I must distinguish the various expressive literatures, which I do by offering a simple typology.

Four Types of Expressive Claims about Law

I count four categories of expressive claims in the legal literature: (1) that law influences beliefs, emotions, or behavior by what it expresses, an expressive theory of law’s effects; (2) that expressive politics determine the content of law, an expressive-politics theory of law; (3) that the normative status of law depends on its meaning, a normative theory of expressive law; and (4) that the normative status of the private behavior the law regulates depends on its meaning, a normative theory of expressive conduct.8
Only the first of these categories (the one in italics in Figure 1.1) is the main subject of this book. The coordination and information theories are expressive theories of law’s effects (category 1). Although I emphasize behavioral consequences throughout, there are other consequences. In my behavioral theories, for example, legal expression first changes beliefs, which in turn changes behavior. Yet belief change by itself is an expressive consequence. Indeed, perhaps the most obvious expressive effect is the emotional reaction to the beliefs the law inspires. If one subjectively feels respected by the law, that gain is an expressive consequence. If one feels disrespected, that loss is an expressive harm. I shall occasionally refer separately to these reactive emotions, but the main event here is behavior.
Now let’s consider the other three categories. A prominent example of an expressive-politics theory of law (category 2) is Joseph Gusfield’s explanation of American Prohibition.9 A simple causal theory might say that Americans adopted Prohibition because they thought it would alleviate the social and medical problems of alcoholism. More pessimistically, one might posit that the law was merely “therapeutic” in the sense that people demanded “something be done” about the problems and there were no good alternatives. Gusfield claims instead that in the early twentieth century, alcohol consumption came to be associated with new, mostly urban and Catholic, immigrant groups, so the prohibition of alcohol worked to express disapproval of those groups. The law’s advocates—descendants from earlier waves of immigration, mostly rural Protestants—sought this disapproving expression, valuing it as a symbol of their higher status. Thus, on Gusfield’s account, law is a symbol over which political groups struggle; it is the meaning the competing groups attach to the legal symbol that explains what law is adopted.10 He explains: “We have always understood the desire to defend fortune. We should also understand there is the desire to defend respect.”11
Figure 1.1:    Categories of Expressive Claims About Law
Figure 1.1 Categories of Expressive Claims About Law
The work of Dan Kahan and Don Braman generalizes this result. Their “cultural cognition” theory suggests that symbolic effects are pervasive in the regulation of risk.12 Because social groups view regulation as a political test of their cultural values, the enactment of handgun regulation or energy conservation standards, the ban on extra large sizes of sugary drinks, or the mandate of an HPV vaccine all depend on what those laws mean to the dominant political groups. There being symbolic competition among social groups, the members of one group will favor the laws they perceive as expressing their social standing and oppose laws the they perceive as expressing the social standing of other groups. Thus, Kahan and Braman, like Gusfield, treat law as the dependent variable and use legal meaning as the independent variable that explains why a particular law exists. By contrast, my (category 1) expressive theories treat behavior as the dependent variable and use legal expression as the independent variable that explains why a particular behavior (especially compliance) exists.
The causal claims do not constitute normative judgments; it might be good or bad that law influences behavior expressively and good or bad that the political struggles over meaning influence the enactment of law. Of course, for a consequentialist, it is easy to argue that expressive effects are good when they generate more compliance with legal rules that promote welfare, or that expressive political struggles are sterile and wasteful. Yet the positive claims being made do not strictly entail any normative commitment. By contrast, an example of a normative theory of expressive law (category 3) is the work of Richard Pildes and Elizabeth Anderson, who propose to evaluate law by whether it expresses appropriate equal respect for individuals, regardless of consequences.13 This view favors interpreting the Equal Protection Clause of the Fourteenth Amendment as condemning laws that express superiority or inferiority of a class of individuals and favors interpreting the Establishment Clause of the First Amendment as condemning state action that expresses an endorsement of religion or some particular religion.14
Another example of a category 3 theory is Joel Feinberg’s famous article about punishment.15 Feinberg defines punishment as requiring hard treatment (that an authority imposes on an individual on account of her rule violation) that carries with it a certain expression, namely, “the expression of attitudes of resentment and indignation, and of judgments of disapproval and reprobation.” Given that understanding, Feinberg makes a normative claim, which is that proportionality of punishment should mean that the level of expressive condemnation, rather than the harsh treatment used to express condemnation, is proportionate to the wrong. Thus, one can normatively evaluate a legal scheme for punishment based on how proportionately it expresses the wrong being punished.
Although my category 1 theories are positive, not normative, Chapter 6 discusses Feinberg’s account of punishment and Chapter 8 more generally explores the normative implications of the coordination and information theories. I do not there offer an expressive normative theory; I instead use a non-expressive normative theory—a concern for consequences—to argue for certain legal policies based on their expressive effects. I wind up addressing some of the same subject matter Pildes and Anderson cover, discussing topics like the establishment of religion.
Finally, there are normative theories of expressive conduct (category 4), which ascertain the moral status of private conduct by examining what that conduct expresses. These theories relate to law because they may justify the legal regulation of private conduct by the conduct’s expressively immoral character. An example is Deborah Hellman’s theory of private discrimination.16 She claims that the discrimination is wrongful if, but only if, it demeans another by expressing her inferiority. The wrong of discrimination thus depends on what the discriminatory act expresses. The theory recommends that law should ban discrimination that, by virtue of its expression, is wrongful. Similarly, Dan Kahan argues that a bias crime is worse than the same crime without the bias because it expresses a more wrongful message, showing greater contempt for the dignity and humanity of the victim.17 The greater expressive wrong justifies greater expressive condemnation, in the form of harsher punishment. Again, my category 1 claims are different because they are causal, not normative.
Despite the distinctions I draw, there are important connections between my primary topic—law’s expressive effects—and other expressive claims about law. These connections should make the theory here of broad interest.
First, an expressive effects theory helps to explain why the stakes in expressive politics are so high. Christians and atheists may struggle politically over laws about a moment of silence in public school just because they prefer a symbolic victory over their opponents. Supporters and opponents of abortion rights may conflict over a new law they perceive as symbolizing the wrongness of abortion, even though they do not believe it will prevent any abortions. But if we understand that these symbolic laws also have expressive effects on behavior, the stakes are higher. As we shall see, government endorsement of religion may not only irritate some people, by reminding them of their outsider status, but also strengthen social norms of public religiosity. Even if abortion regulations contained enough exceptions that they fail to directly lower access to abortion, they might strengthen anti-abortion social norms. Stronger norms mean stronger informal sanctions against the counter-normative behavior, which can change behavior. Thus, the expressive power of law explains why symbolic struggle is so fraught, why it is so rarely ever merely symbolic.18
Similarly, expressive effects interact with the normative theories of expressive law, or so I argue in Chapter 8. The basic claim is that normative objections are stronger when they involve behavioral effects. For example, the normative theorist may be correct to condemn a law because it implicitly endorses, say, gender-based stereotypes, regardless of consequences. Yet the objection to the law is distinct and more powerful if the expression will also have the consequence of strengthening norms of inequality and producing more discriminatory behavior.

The Limits of Expressive Claims

As noted, the primary goal of this book is to demonstrate that law influences behavior expressively, that legal compliance is not explained entirely by deterrence and legitimacy. This thesis is addressed to social scientists who work with these dominant theories. In particular, to convince economists who focus on deterrence, I build the theory of this book with simple rational choice tools. Yet I also wish to address a different audience for the purpose of making a different ...