With the French 55-45% “no” vote to the European constitution, one of the great experiments in post WWII history has come to an end, the idea that economic integration could overcome a past defined by political division. Much of the vote was driven by the fear of losing characteristic national political entitlements (benefitting some, disadvantaging others) that would not have survived in the long run under a more centralized EU government, such as a relatively greater degree of protection from the free flow of labor and capital. Thus, the perceived failure is twofold. The economy failed to deliver, and the government failed to protect the economy by putting the voters’ parochial interests first. Historically, the simultaneous lack of faith in both the economic process and government by a significant part of the voters precipitated political shifts from the center to the fringes. The coming weeks and months will show if that pattern reemerges, not only in France, but also in other European countries. It appears that advocating for a unified Europe requires more than an amorphous appeal to “a stronger Europe,” whatever that is supposed to mean, or a “more diverse Europe,” which may appeal to some, including me, but not to others. Advocating for a unified Europe requires that voters in the wealthy (slow growth) countries understand what’s in it for them if the poorer (high growth) countries join. From within a framework of redistribution, the answer is: very little. And that framework of redistribution is pervasive, not just in France, but also in Germany, the Netherlands, Italy, etc., which makes the rejection of the European constitution almost rational, assuming that the framework of redistribution will remain in place. Europe may have arrived at a juncture where that assumption must be challenged, where further political integration is dependent upon a political rediscovery of non-political values, such as individual freedom (not Rousseau’s), free markets, and open societies, three topics that most successful national political campaigns have avoided in the past.
Monthly Archive for May, 2005
Commenting on Larry Solum’s Legal Theory Lexicon entry on Formalism and Instrumentalism, drhfk recommends recasting the debate as the distinction between first and second order oberservation. This distinction derives from systems theory (particularly Luhmann), and rests on the concept that a social system (such as law) is marked by a fundamental dichotomy (such as “lawful/unlawful”) and that this fundamental dichotomy is applied by first-level observers (lawyers, judges, etc). To find out what law is, you would observe what lawyers and judges etc. are doing: how they apply the fundamental dichotomy and when. That is second-level observation.
I agree that formalism vs. instrumentalism can be recast as first-level vs. second-level observation. Deciding a case, by asking what the law is, is first-level observation; asking what law is, what the law should be, and what effect it has in society, is second-level observation. But it gets fuzzy quickly, and I think the distinction cannot take us as far as drhfk envisions:
In practice, every lawyer oscillates between first and second order observation. However, it remains a requirement of the legal system that every decision be issued and every claim be made on the basis of the code of the legal system: lawful/unlawful. And that requires us to assume the internal perspective of a first order observer. A plaintiff may, in fact, win because his or her victory is efficient, just, or politic. But a court could never say so, or else its ruling would be extra-legal communication. The plaintiff wins because (extra-legal) considerations of efficiency, justice, or national security (visible to the judge as a second order observer) have convinced the court (as a first order observer of the law) that, based on statute or precedent, the plaintiff has a right to the remedy.
What does it mean that the court could never say so? Is it “cannot” or is it “must not”?
To see what “cannot” looks like, imagine a game of chess. One of the players moves his bishop forward on the vertical column and takes his opponent’s piece. Bishops only move on the diagonal in chess. Moving the bishop like a castle is not only a wrong move, it is no move at all: The rules of chess define not only correct/incorrect moves, but also moves/non-moves. Consistently moving the bishop like a castle is meaningless. The game is over, or it is not chess.
Is that how law works? Are there impermissible moves in a judge’s opinion that render the argument extra-legal, such as handing one side the victory based on extra-legal (instrumentalist) considerations and saying so? Or is any argument, simply by virtue of having been written by a judge in his opinion in a formal legal proceeding, a legal argument? The latter is the case from a systems-theory standpoint, but I am not sure that this helps much in explaining the differences between formalism and instrumentalism.
First notice here a subtle stylistic difference between code-based legal systems and the common law. Code-based legal systems tend to incorporate instrumentalist arguments into their legal considerations, by slapping the label “law” on them, if you will. Instrumentalist arguments are viewed as being legal, but of a different order. For example, legal argument may be both de lege lata or de lege ferenda (”law as it has been laid down” or “law as it should be made”). The thrust is to consider everything a judge does legal.
The common law has resisted this move, and has done so long before realism. Equity is an example: It is conceived as a separate normative system, and the premise of equity is that an outcome at (common) law is wrong or otherwise unsatisfactory and requires fixing in a manner the (common) law cannot provide. There were other such separate “legal” systems: ecclesiastical law, the law of the manorial and mayoral courts, the law merchant. In modern times, realism has insisted not only that a formalist legal argument may hide instrumentalist considerations (and other factors), but also that doctrinal and instrumentalist considerations cannot be distinguished. The radical realist position is that there is no law, that the judge should make reference only to instrumentalist considerations, and that the legal form of arguments is at best harmless and at worst serves to obscure the real meaning of the decision behind a cloud of transcendental nonsense.
Here is an example of how internal and external perspectives play out in real life. In Planned Parenthood v. Casey, 505 U.S. 833 (1992), the Supreme Court faced the question whether to overrule Roe v. Wade. In considering stare decisis, the Court states that “a decision to overrule should rest on some special reason over and above the belief that a prior case was wrongly decided.” What special reason could this be but an external, instrumentalist reason? Any internal reason would merely make the prior case wrongly decided. The Court goes on to consider its own prestige and reputation:
In two circumstances, however, the Court would almost certainly fail to receive the benefit of the doubt in overruling prior cases. There is, first, a point beyond which frequent overruling would overtax the country’s belief in the Court’s good faith. Despite the variety of reasons that may inform and justify a decision to overrule, we cannot forget that such a decision is usually perceived (and perceived correctly) as, at the least, a statement that a prior decision was wrong. There is a limit to the amount of error that can plausibly be imputed to prior Courts. If that limit should be exceeded, disturbance of prior rulings would be taken as evidence that justifiable reexamination of principle had given way to drives for particular results in the short term. The legitimacy of the Court would fade with the frequency of its vacillation.
The defendant in Casey loses because the Supreme Court doesn’t want to be seen to vacillate, even though from the internal perspective Roe v. Wade may have been wrongly decided. The interesting point here is that the reference to external considertations is open, and that there is no attempt made to reconcile the internal, formalist and the external, instrumental points of view. In fact, the Supreme Court’s argument from its own role and reputation is “legal” only in the abstract sense that it is part of a formal decision issued by a court of law in a formal proceeding. Therefore, I think there is more to the relation of realism and formalism than can be reduced to the formula that “[a] plaintiff may, in fact, win because his or her victory is efficient, just, or politic. But a court could never say so, or else its ruling would be extra-legal communication.”
Following up on a post by Larry Solum and other posts in this blog (for example here and here), I would like to explore what I consider to be a sensible theory of legal formalism.
In order to understand legal formalism, it is helpful to begin with a rough sketch of the core features of legal realism. Legal realism is the descriptive, non-normative study of what determines a judicial decision. Often, the goal of legal realism is explicitly stated as uncovering the causal relationship between inputs (facts and law) and output (decision). Legal norms, according to the realists, are not particularly determinative inputs. Judges are fact driven to a much greater extent than they are norm driven. And there are a number of good reasons for this peculiar weakness of norms as determinants of judicial behavior. Laws are indeterminate, that is, the judge doesn’t get an unequivocal normative input. “Every contract in restraint of trade is illegal,” now what exactly is that supposed to mean? The indeterminacy thesis comes in (at least) two variants. One takes the internal, first-order perspective into account and argues that there is no “one right answer,” which is a form of hermeneutic indeterminacy. Different canons of construction will lead to equally plausible yet contradictory claims, either in all cases, so the more radical versions of the indeterminacy thesis, or at least in “hard cases,” so the more moderate versions. (The same applies, of course, to choosing among the canons themselves.) The other variant, true to the naturalistic roots of realism, assumes the position of a second order observer, for example that of a sociologist, who is observing the judge, who thinks (or rather, is trapped in the delusion) that he or she is “finding” the right answer to a dispute by applying facts to norms. The sociologist is able to see the real determinants of the judge’s behavior, which are either rooted in the judge’s individual psychology or in his or her socialization and features of the social universe at large. Critical legal studies, the heirs of traditional legal realism, take the latter position as their point of departure.
Now what about formalism? Formalism claims that judicial decisions are norm driven, not fact driven. In my view, this claim is correct, it is almost correct by definition, but only if formalism is understood in a very specific and somewhat technical manner. Legal systems are self-referential and self-replicating systems of communication. (Some writers have referred to such systems as autopoietic, but that term carries too much unnecessary baggage for my taste.) In other words, the elements of a legal system produce the elements of a legal system that produce the elements of a legal system, ad infinitum. In this sense, legal systems are closed. They contain nothing and consist of nothing but legal communication. Everything else, including judges, lawyers, courthouses, jails, minds, brains, the world at large, is emphatically not part of the legal system. That is not to say that the legal system could exist without its environment, of course it couldn’t. But the legal system and its environment do not merge. Explaining a legal decision as caused by its environment is similar to explaining thoughts (or other mental states) as caused by the brain. While the mind is structurally coupled to the brain, both operate independently and at no point do their operations merge. Mental operations connect to mental operations connect to mental operations and neuro-physiological operations connect to neuro-physiological operations connect to neuro-physiological operations. But at no time does a thought connect to a neuro-physiological operation and vice versa. The mind cannot access the brain, and the brain cannot access the mind. In the same manner is communication independent from mental operations. Communications are not thoughts and thoughts are not communications. They are both structurally coupled, because without minds there would be no communication, but they do not merge. The legal system is a communicative system. Its elements are legal communications, that is, communications using lawful/unlawful as their operative distinction. Only legal communications connect to legal communications. And that’s what results in the autonomy of the legal system. The legal system is autonomous in its decision as to what (in its environment, seen from from the vantage point of the legal system) is legally relevant, that is, the decision of legal relevancy can only be made on the basis of (prior) legal communications. For example, economics is not about to “overcome” law (as Posner claims). Rather, the law selectively assimilates economics on its own terms. And that leads to the conclusion that the closure of the legal system is a prerequisite for its openness. That is not a paradoxical statement. Consider a cell, which is operationally closed but energetically open. Whatever molecule enters a cell and thus becomes part of the cell metabolism enters only because the cell selectively chooses to “let it in.” In the process, the molecule assumes an additional role as an element of the cell, which is now defined by a self-referential and self-replicating biological system. Of course, we could track the molecule from the outside, but that would tell us little more about the operations of the cell metabolism than observing the judge tells us about the law. Legal formalism is correct in insisting on the autonomy of the legal system. The legal system draws its own boundaries, and it draws it according to legal criteria and legal criteria only. Legal formalism is incorrect when it confuses autonomy with autarky. Legal systems do not exist in a vacuum, they depend on their environment, and they selectively incorporate it. Legal realism is correct in reminding us of the significance of that environment to the legal system. Legal realism is incorrect to the extent that it confuses different modes of operation, for example, legal communication with preferences, interests, or states of mind.
Larry Solum posted an entry on Formalism and Instrumentalism in his excellent Legal Theory Lexicon. Here is a taste:
The core idea of formalism is that the law (constitutions, statutes, regulations, and precedent) provide rules and that these rules can, do, and should provide a public standard for what is lawful (or not). … [In contrast:] … [M]ost instrumentalists would agree on the idea that legal rules should be interpreted in light of their purposes. When applying the letter of the law would undermine its purpose, then the rule should be interpreted so that it does not apply. And likewise, if the spirit of the law would be served by its application, then judges should give the rule an expansive interpretation.
In my opinion, the debate between formalism and instrumentalism can fruitfully be recast as the distinction between first and second order observations. The first order observer must by definition assume the normative relevance of the law. From an internal point of view, questions of indeterminacy will appear as issues of construction or interpretation and will be dealt with the tools provided by and immanent to the law. The second order observer, in contrast, has no built-in normative commitment. He or she looks at the law as text, literature, economic phenomenon, sediment of a power struggle, social institution, etc. In most instances, observing the first order observer (and the fact of his or normative commitment) is also a significant part of second order observation. (Strictly speaking, a second order observation is defined as observing another observer. For example, a sociologist observing a judge. But second order observation, at least in my view, is not limited to only observing observers.) In practice, every lawyer oscillates between first and second order observation. However, it remains a requirement of the legal system that every decision be issued and every claim be made on the basis of the code of the legal system: lawful/unlawful. And that requires us to assume the internal perspective of a first order observer. A plaintiff may, in fact, win because his or her victory is efficient, just, or politic. But a court could never say so, or else its ruling would be extra-legal communication. The plaintiff wins because (extra-legal) considerations of efficiency, justice, or national security (visible to the judge as a second order observer) have convinced the court (as a first order observer of the law) that, based on statute or precedent, the plaintiff has a right to the remedy.
Heeding my own call for a re-examination of consequentialist theories of punishment, and responding to Bloomfield’s posts here and here, it might make sense to examine how the radical dichotomy of consequentialist and retributivist accounts came about. In the Aristotelian world order, which dominated European political thought into the 17th Century, humans were “political animals” and as such they ontologically belonged to a social order, an order within which each had his or her predefined place. Society was the human biotope. It was unthinkable as against nature for humans to live without or outside of society. Those who did, were either wolves or gods. (Not surprisingly, the Werewolf, or man-wolf, has its origins in Aristotelian substantive social ontology.) Against this backdrop, the critical question of legitimacy was one of justice, that is, whether an existing social order was good or bad. Hence Aristotle’s taxonomy of political systems: tyranny, aristocracy, democracy, oligarchy, ochlocracy. Whether there should be society at all, was beyond the frame of reference of Aristotelian political thought. The substantive social ontology, in which injustice was being out of place had wide-ranging implications for a theory of punishment. Most significantly, there was a qualitative, ontological difference between the criminal and everyone else. The criminal was wicked by nature. The crime merely revealed that nature for everyone to see. And so, the through his crime, the criminal literally expelled himself from society. Crime, and punishment as a reaction to the crime, therefore did not, strictly speaking, take place within society; rather, the crime was evidence that the borders of society had been drawn too liberally, mistakenly including the criminal. The response was an exercise in downsizing society, that is re-drawing the borders of society under exclusion of the criminal. Once properly excluded, the criminal could be dealt with in an instrumental manner like any other natural (that is, extra societal) object. He could be killed, mutilated, maimed, brandished, locked up, sent into exile, etc.
The collapse of Aristotelian substantive ontology in the late 17th Century (brought about by Galileo, Newton, Leibnitz, and others), also led to a collapse of the corresponding social ontology and its theory of punishment. Thomas Hobbes was the first to transpose the revolution in the physical realm to the social world. As the physical world was reduced from a rich, qualitative ontology to a monistic concept of matter in motion, its human inhabitants were stripped of their various (inherently unequal) essences, which pre-defined their proper position within society. Hobbes re-conceptualized humans as trivial machines, that is, complex, causally determined automata, programmed to ensure their individual survival by rational means. One of the many implication of this seismic shift in the concept of the subject was a sudden urgency in the search for new sources of political legitimacy. Such new sources could no longer be found within the existing social order. And so, in the mid-17th Century, Hobbes posed the radical question, from which all modern political thought embarks: “Why have a government at all? Why not anarchy?” This question was only intelligible against the backdrop of the new, radically simplified social ontology, in which the individual, for the first time in Western intellectual history, logically and genetically preceded the state. The modern state, as Hobbes saw it, is composed of individuals, and it exists to serve one specific, shared objective of each of such individuals: the establishment and maintenance of peace. Beyond that, no shared objectives could be presumed. Difference replaced unity as the starting point for political thought. With that, the semantics of political legitimacy changed from a qualitative question (”Is the government good or bad?”) to an instrumental one (”¬Ä¬úIs the government effective in preserving the peace?”). Seen from the vantage point of the modern individual, the legitimacy of the government came to hinge on a comparison of costs and benefits. As long as the benefits of having a government outweigh the costs, including opportunity costs, it is rational to have a government. Once that bottom line changes, the government has lost its legitimacy. The costs of government are measured in the restrictions that it imposes on the exercise of any one individual’s freedom of action. Therefore, at the heart of each individual’s interaction with the state is an exchange. The individual gives up some measure of freedom of action (for example, not to steal, having to pay taxes) in exchange for a greater measure of freedom of action (for example, through the existence of secure property rights and the absence of violence). The qualitative summum bonum of old gave way to the quantitative maximum bonum. Freedom (that is, negative freedom or freedom from), replaced justice as the central criterion of legitimacy.
As the new monistic ontology was unable to account for qualitative differences between individuals, the exclusion of the criminal on ontological grounds was no longer plausible. The criminal did no longer appear as qualitatively different from the law-abiding citizen. This reconstitution of the individual as a rational cost-benefit calculator opened the door to consequentialist accounts, such as Beccaria’s On Crimes and Punishment. It also paved the way for theories of punishment and political legitimacy built on the idea of a social contract. (Negative freedom as the source of legitimacy is the common root of consequentialism and social contract theory.) Social contract theories, while radically different from political Aristotelianism in their philosophical foundations, quickly recreated the effects of exclusion on ontological grounds by expelling those in breach of the social contract from society. (Hence, one of the recurring problems of social contract theories is that, taken to their logical extreme, there is only one crime, the breach of the contract, and one punishment, the termination of the contract.) Therefore, little changed in practice. Criminals were still excluded and treated accordingly (that is, brutally). In other words, the first seismic shift from the Aristotelian world to that of the modern economic man, did not yet bring about the sharp divide between consequentialist and deontological accounts. It took another shift of similar magnitude and consequence to arrive at the sharp division that characterizes and haunts modern ethical theory until this day: the discovery of universal rationalism. (More on that in a later post.)
[tags]Hobbes, Aristotle, punishment, anarchy[/tags]
Along the ceiling of the long walkway connecting the Port Authority and the Times Square subway stations runs a poem: “Overslept, so tired. If late, get fired. Why bother? Why the pain? Just go home. Do it again.”
Who put this exquisitely depressing statement of alienation on M.T.A. property? Given the official looks of the signs, could it have been the M.T.A. itself, which would raise a fundamental question indeed: Are complex organizations capapble of irony? Apparently they are! Here is an excerpt from the FAQChest’s Railroad edition, quoted from Scribbling.net:
The sign is not meant as a cruel joke but rather to relay an artist’s empathy for the commuting public. The sign was commissioned 10 years ago through the M.T.A.’s Art for Transit program as a temporary work. The artist, Norman B. Colp, said the work was inspired by the old Burma Shave advertisements that were seen along highways from 1927 to 1963. The signs were spaced a few hundred yards apart and were to be read one line after the other as you traveled along. The piece in the walkway combines the Burma Shave model with Mr. Colp’s experience as a commuter, one who recalls that he “made it to work barely on time.
More of this, and I will never complain about fare hikes again!
I have previously developed the thought the experience of the Frontier and the duel in particular, show that in America law and society as cultural phenomena are understood in terms of inclusion and exclusion. I called this the edge of the law, refering to the sense that there is an area outside the law and society that individuals can enter: membership in society, if you will, is a matter of choice.
Taking this to the level of the state, there is a difference here bewteen America and the older European states. The history of the state in Europe is bound up with the history of the Church. When the Roman Emperor Constantine converted to Christianity on his deathbed in 337, it meant the union of the Roman Empire and the Church. The existence and presence of the state became part of Church doctrine, just as the Roman state became the model for the later nation states and republics. The Catholic belief, based on a reception of Aristotle, is that man is a political being (zoon politicon) and that the organization of humans into society is natural. The state is god-given as the secular government, just as the Church is god-given as the spiritual government of human affairs. For Luther, the existence of the state is also god-given, but in a slightly different sense: Because temporal existence is imperfect (natura corrupta) the state is necessary to govern imperfect beings in their imperfect existence. The European monarchs ruled “by the grace of God.” The point is: society is regarded as pre-existent and not as subject to the individual’s choice. Enlightenment emphasized the individual right to leave a particular state, to emigrate; but there was no concept that the individual could leave all society, to exist outside a state.
In America, the experience has been different. Based on the belief that humans have inalienable rights (rights that don’t presuppose the existence of a state against which they are claimed), the state appeared as the voluntary association of individuals that surrendered part of their sovereignty in the social contract. If you consider the origins of the American colonies, it is remarkable that they were organized as legal corporations – this is true both of the “business” colonies set up to exploit the natural wealth of the New World, and the “religious” colonies (such as Massachusetts Bay). Citizens were shareholders who associated of their own free will. The American myth of foundation is voluntary association, both of enlightened individuals, and original colonies as members states.
These are broad brush strokes, of course, and in America there have been other colors mixed into the picture: Lincoln’s words at Gettysburg “our fathers brought forth … a new nation” suggest a society bound together by a shared destiny that members cannot quit. But despite the Civil War, and despite the New Deal, there is a strong cultural sense of primacy of the individual’s choice. One such choice is crime. Society means a shared normative space, and crime is the choice to opt out. The outlaw has stepped outside the bounds of the law. Since there were individuals before society existed, individuals can exist outside society, as outlaws in the Wild West and as “predators” (that is, non-humans) in today’s metaphorical space. The death penalty, which is socially acceptable in America but not in Europe, is the extreme from of exclusion from society. The criminal, who has chosen to opt out of society, removing him permanently from society. Both crime and punishment are about the edge of the law and the edge of society.
Drhfk writes:
I am not sure that framing the justification debate in terms of rights against the state for protection is particularly helpful. Those rights, I would argue, are derivative of the state’s right to punish criminals. And whether such rights are morally defensible is the issue. … I do agree with Bloomfield’s observation that the choice of a particular theory of punishment as the “official” theory would commit a state to certain features of a criminal justice system. For example, if retributivism were the official theory, then plea bargains, prosecutorial discretion, indeterminate sentencing, and any form of coercive crime prevention would be hard to justify.
The question lurking in the back of my mind is: Will retributivism prove too much? If we imagine a state that justifies punishment purely on retributive grounds, would such a state be able to consistently not punish some instances of crime? Prosecutorial discretion can be explained in terms of consequentialism as an efficient use of resources as measured by the ends of punishment, that is, deterrence, which can be achieved by punishing less than 100% of all crimes. If the crime itself, on the other hand, is sufficient grounds for punishment, can there be retributivist reason not to punish?
Pure retributivism can be seen to lead to a duty to punish, because a purely retributive system would find no reason to justify selectively not punishing crime. (If the state has a duty to punish, it is only a small step for the citizen/victim to demand that the state fulfil its duty, but that is not a central point here.) A duty to punish is not widely accepted today, and I think it is an indication that retributivism is seen as taking the theory of punishment only so far, even by its supporters.
Perhaps the right approach is to bifurcate the justification of punishment: Punishment needs to be justified vis-a-vis society and vis-a-vis the criminal. Society demands a justification for punishment, because it has an interest in limited state abridgment of liberty and because unjustified punishment may affect all of us, even the innocent. To justify punishment in social terms it is necessary (and sufficient) to demonstrate that the threat of punishment and the act of punishment prevent crime (or in a lesser form of the argument, lend the citizens a sense of security and peace despite the presence of crime). This justification does not work for the criminal: The crime committed can no longer be prevented. Future crime can be prevented just as easily by punishing an innocent citizen and proclaiming him or her to be guilty as by punishing the guilty citizen and proclaiming him or her to be guilty. “Why me?” is the question the criminal asks to demand a justification of punishment, and the answer commonly given is retributivist: “Because you are guilty of the crime.” In this way, the “principle of guilt” justifies and limits the state in meting out punishment. The criminal’s guilt justifies that he or she is singled out for the purposes of crime prevention in an consequentialist calculus. The requirement of punishing only the guilty limits the state in using the same consequentialist calculus to punish those who may or may not be guilty but whose punishment will effeciently deter others.
While I greatly appreciate Bloomfield’s addition of the bears that shoot laser beams out of their eyes to the cast of characters that appear in the discourse about the justification of punishment (such as the ubiquitous small-town Sheriff who, in order to prevent a murderous riot, sacrifices an innocent person to the mob), I am not sure that framing the justification debate in terms of rights against the state for protection is particularly helpful. Those rights, I would argue, are derivative of the state’s right to punish criminals. And whether such rights are morally defensible is the issue. Moreover, even if state punishment can be justified (on consequentialist and/or retributive grounds), one could still argue that third parties have no right to compel state action. Some countries grant such rights vis-?… -vis the executive branch more liberally than others. Other states even permit or have seriously considered permitting suits against the legislative branch for legislative inaction. But the considerations for and against granting such rights include a broader set of questions than “just” whether punishment is morally justifiable or even morally required.
That said, I do agree with Bloomfield’s observation that the choice of a particular theory of punishment as the “official” theory would commit a state to certain features of a criminal justice system. For example, if retributivism were the official theory, then plea bargains, prosecutorial discretion, indeterminate sentencing, and any form of coercive crime prevention would be hard to justify. The latter, in particular, would probably require a different justification altogether. The conceptual separation between (backward-looking, i.e., retributive) punishment and (forward looking) prevention arguably led to the bifurcated police system in some states such as Germany, where the police acts in two capacities, each subject to very different rules and constraints: criminal investigation (as part of the criminal justice system), and threat prevention (as part of the administrative system). In contrast, if the official theory was of consequentialist nature, we would expect the line between investigation and prevention to be blurred, as is the case in the U.S. We would also expect an opportunistic calibration and re-calibration of punishments, as well as a significant discrepancy in the announced punishments and the real time served. Moreover, prosecutorial discretion would be of paramount importance, as would be a practice of plea bargaining. However, the critical question that started this thread remains: Can punishment be justified on purely consequentialist grounds? Whether I am justified in defending myself against homicidal, cybernetically enhanced bears or whether I have a right to demand that the state protect me from this long overlooked yet endemic threat, is a different question.
In 1956 the Supreme Court concluded that DuPont, accused by the government of having monopolized the market for cellophane, did not have sufficient market power, because, in a properly defined product market, it was unable to raise prices profitably, that is, so many customers would have stopped buying the firm’s products (or would have bought less) that the higher prices extracted from the loyal customers would not have made up for the loss. The discussion was subject to almost immediate criticism, as lawyers and economists discovered an error that became known as the “Cellophane fallacy.” Here is the court’s central premise, stated positively: “Only a firm with market power can raise prices profitably.” Nothing wrong with that. But the negative implication, on which much of the opinion rests, does not follow: “If a firm cannot raise prices profitably, it doesn’t have market power.” The reason for the disconnect is that even a monopolist has no control over the reservation price of its buyers. At some point every customer is fed up and walks. So inability to raise prices any further only shows that the firm is already pricing optimally, irrespective of whether the firm is one of many competitors or a monopolist. The “raising prices profitably” test tells us something about whether the firm can exercise additional market power. It tells us nothing about whether the starting point for the exercise of additional market power is no market power or monopoly.
There might be a similar mistake with respect to consumer choice. Consumer choice is one of the bedrock principles of modern economics. In fact, the entire economic process has been portrayed as a transformation of relatively few raw materials into an ever expanding universe of goods that are ready for consumption. At each step along the way from planting trees in the nursery to buying a pencil in the store, value is added. And every supplier along the way coordinates its activities based on the consumer’s willingness to pay. The consumer is the source of all original price signals. Everything else is, in some fashion, derived demand. And price signals depend on the consumer’s subjective preferences. Thus, the idea of consumer sovereignty. The “wealth of a nation,” as we understand it today, lies not in the value of the gold stored in Fort Knox but in the quality, quantity, and variety of goods that consumers who happen to live here are able to enjoy.
But what about the universe of the consumer preferences? Consumers seek to maximize their utility, that is, they choose (by and large) reasonable means to achieve their ends. Thus, looking at me, economic theory would assume that I am, in fact, maximizing my utility. There is a not-so-subtle, legitimizing aspect to this theory, as I am, presently, in a maximalist state. Using the present state as a starting point, antitrust law would ask whether a merger between two competitors would significantly reduce my universe of choices, and if so (and if there are others that are similarly affected) chances are the the government would not permit the merger to proceed. But what about the starting point? In what way do my choices truly express my preferences, or the degree to which they are satisfied, rather than merely reveal which of the options available to me I choose to exercise? Put differently, I can only express my preferences within a pre-defined set of choices, and I have little to no opportunity to help define that set. For example, I have read that the average, specialized record store (apparently a dying species) has about 60,000 titles in stock. A typical big box retailer, in comparison, has 5,000, all purchased centrally. Big box retailers command by far the largest share of record sales. Similarly, local radio stations used to play different kinds of music, today’s near-monoculture pushes the exact same mix everywhere. Unless I seek out speciality stores, download what I need, or fork over $120 for satellite radio, I can only express my preferences from within a set of overall unappealing choices. Rational actor that I am, I settle for the least unappealing. A similar case could be made for the auto industry, where it requires significant effort to find low consumption, low emission cars. Here, too, my choices will be made and my utility maximized, within a set of rather unappealing options. And the list goes on. Of course, I understand and can anticipate the reply that my plight merely exemplifies that, after things went awry in the Garden of Eden, goods are limited and wants are not. Everybody has to choose from within the set of available options, and at no point in time have options been more varied and, significantly, available to a greater number of consumers than they are today. Moreover, my choices of the least unappealing alternatives will, if others share my preferences, provide incentives for producers to supply more of what I want, and to expand or shift the set of available options in a direction that will provide greater value to me, given my set of preferences.
Which brings me to an important point about the goals of antitrust policy. Which industry structure is more likely to be responsive to expand or shift the available sets of choices in response to diversified consumer demand? A concentrated industry or a collection of small firms in open markets? While I am not at all certain of the answer, it seems at least plausible that in markets for differentiated products that are not characterized by high fixed costs (in comparison to other markets, not to variable cost) or other significant scale economies, smaller entities have greater incentives to respond quickly to diverse customer preferences. The welfare gains from the expansion of the set of available choices may well be greater than achieving lower costs for more or less standardized and therefore incomplete substitute products. So maybe antitrust policy should embrace different standards: an efficiency standard for some markets, and a policy aimed at maintaining low concentration levels for others. As always, axiomatic reasoning only goes so far in these matters, and empirical evidence rules. That said, there is certainly nothing in the law or in the legislative history of the antitrust statutes to confine antitrust policy to operating within existing sets of options or markets. Rather, it may be an even more critical role of antitrust policy to ensure that conditions prevail on a broader scale, within which sets of options and markets can evolve more freely. Protecting the conditions of the possibility of market evolution is no call to return to the days of U.S. v. Alcoa, Brown Show v. U.S., U.S. v. Philadelphia National Bank, or U.S. v. Von’s Grocery. Rather, it is a call for a re-examination of the efficiency paradigm as the sole justification of the antirust laws. Sometimes, (productive) inefficiencies may be a price worth paying for significantly increased total welfare. But is it then still correct to speak of inefficiencies?
[tags]economics, rational choice, welfare[/tags]
In every debate about the goals of antitrust policy, or, more broadly, economic and regulatory policy, proponents of diametrically opposed positions use the same words to (favorably) describe the ends of their policies. Consumer welfare, for example, is touted as a goal by both consumer advocates (who favor regulation) and economic conservatives (who abhor it). Efficiency is another positive, for which each camp claims to have the monopoly. And what about perfect competition and its assumption of many small firms? Populists use the model to defend an ideal of a decentralized economy, yet perfect competition is the bedrock of the neo-classical revolution and Chicago school economics, that led to near-monopoly mergers and did all but away with restraints on vertical mergers, checks on tying, and monopolization offenses such as predatory pricing. Without taking sides in the debate (for now, at least), let’s try to disentangle some of these concepts.
Efficiency is the guiding principle of modern economic policy. Allocative efficiency is realized, when all mutually beneficial trades have been made, and all goods are in the hands of those who value them most. Productive efficiency ensures that whatever is being made is produced in the most efficient manner, that is, no change in the mix of inputs would result in increased output, given the current technological constraints. A firm produces efficiently, if it moves along the production possibilities frontier. The greater allocative and productive efficiency, the greater the economic pie, that is, the quantity, quality, and variety of goods produced. Maximizing allocative and production efficiency therefore maximizes total welfare. How that welfare is being distributed between (and among) suppliers and consumers is a different matter. Consumer welfare, the stated goal of U.S. antitrust policy, appears to take sides in the debate about distribution, as consumer welfare implies (i) the creation of the greatest possible pie (efficiency) and (ii) giving the entire pie to the consumer (distribution). The means to achive both goals is perfect competition. Under perfect competition, marginal revenue equals marginal utility equals marginal cost equals price for all products. Allocative efficiency and maximum consumer welfare are the direct results. Against this backdrop, market power, the ability of a firm to charge prices in excess of marginal costs, is a twofold evil. First, supra-competitive prices as a result of market power are inefficient, because now there are units whose marginal utility exceed marginal cost that are not being made and cannot travel to those who value them most. Profitable stuff that doesn’t get made as a result of market power constitutes the deadweight loss. (By the way, pointing to the resources that are being saved as a result of producing less does not mitigate the problem, because those resources are now being put to less valued uses.) Second, charging supra-competitive prices also violates the distributive ideal of consumer welfare, as it channels the monopoly profits into the pockets of the suppliers and their shareholders. (Note that inherent in the ideal of perfect competition is the justification for antitrust regulation, as deviations from the ideal require governmental action to restore competition.) While perfect competition may be the only way to maximize both, efficiency and consumer welfare, efficiency without consumer welfare is not conditioned upon competition. A monopolist with the ability to perfectly price discriminate would achieve the same output as an industry under conditions of perfect competition. For every unit sold, marginal revenue would equal marginal utility would equal price (which would now be different for each unit), even though marginal revenue would exceed marginal cost. And even though the delta between marginal revenue and marginal cost that used to be the consumer’s surplus would now be the supplier’s surplus, there would be no deadweight loss. From a total welfare point of view, both models are equally efficient. As to increased production efficiencies, the monopolist could point to economies of scale and scope. (Whether a monopolist could sustain those advantages over time is another question.)
Contrasting consumer welfare with corporate welfare (pun intended) is instructive, because it serves as a litmus test for advocates of consumer welfare. Most proponents of “consumer welfare,‚Äù in reality, only embrace the efficiency prong but not the distributive ideal. One cannot consistently advocate for consumer welfare and, at the same time, support policies that ensure significant corporate profits. However, there is nothing inconsistent about a policy of increasing efficiency, which may benefit both, consumers and firms; this is simply to say that we should grow the pie so that there is more for everyone to eat. Thus, we are left with a bit of a quandary. With respect to the use of the term consumer welfare, both consumer advocates and economic conservatives alike take somewhat of a pick and choose approach. Consumer advocates focus on the distributive ideal and the “many small firms‚Äù element of the perfect competition model, but otherwise don’t necessarily see efficiency as the overriding economic policy goal. Economic conservatives focus on the efficiency prong, usually placing greater weight on the productive than on the allocative efficiencies, and discount the distributive ideal. A further complicating factor is that there is more to efficiencies than the static concepts of allocation and production. It is hard to deny that total welfare has been primarily increased through the discovery of new technologies. So it may well be more beneficial for society, at least in the long run (which is getting increasingly shorter, another implication of Moore’s Law) to trade some allocative and productive efficiency for greater corporate profits, if such profits and the additional capital that they attract are being re-invested in research and development.
Against this backdrop, it seems that the discussion about the goals of antitrust and economic policy would benefit if we abandoned the “consumer welfare‚Äù or “consumer benefit‚Äù rhetoric for more explicit statements about how we want to make the pie (that is, what kind of efficiencies we promote), and how we propose to slice it (that is, what distributive mechanism and criteria we prefer).
Does it matter to the citizens, as the potential victims of crime, whether the state embraces a consequentialist or retributivist theory of punishment? Is a consequentialist theory of punishment even escapable? Responding to the call to re-examine consequentialist theories of punishment, let me take you along a series of steps, starting with my right to life & limb and ending with the state’s punishment of crime, and prosecutorial discretion. For the sake of discussion I will assume that we know that punishment is effective, that deterrence works; but that we don’t know to what degree exactly, or how. The assumption is that we do know that without punishment crime increases quickly and exponentially.
(1) The easy starting point: The government sends out a goon to kill me (without due process of law). I have a constitutional right not to have a government goon kill me. I don’t even have to wait until the goon is at my door; I can demand that goons not be sent out by the government.
(2) The government releases a pack of wild animals (I am thinking of something like bears that shoot laser beams out of their eyes). It is clear that at least some of these wild animals will attack citizens like me and harm or kill them. As a citizen, I have a right to demand that the government not release these dangerous wild animals (there is no social benefit in releasing these menaces).
(3) The government releases repeat murderers who have been shown in psychiatric tests to be incurably violent. (This example is not quite as far-fetched as the laser-eye bears, if you consider the sex-offender debate.) It is understood that these repeat murderers will kill again. Do I have a right to prevent the release of the murderers? The difference is that the murderers are citizens themselves, with rights and liberties; of course, in my example they are psychotic, which may make them more like wild animals than like citizens for purposes of public safety (what else would be the reason for calling certain criminals “predators”).
(4) The government decides to fix the budget crisis by no longer prosecuting or punishing murder and other violent crime (tax evasion continues to be punished). Murder, robbery, and rape all increase, as (potential) criminals are no longer deterred by the threat of punishment. It was my assumption, as you’ll recall, that deterrence is known to work to some degree. Unlike the murderous psychopaths of the last hypothetical, most of these no-longer-deterred criminals are sane citizens without prior criminal records: responsible, independent actors in society. Still, as a citizen I feel threatened. Do I have a right to demand that the government resume punishing crime? My argument is that I have a right to life, of which I may not be deprived without the due process of law, and that, since deterrence is known to work, the government’s failure to punish murder (increasing significantly the likelihood that I will be murdered) is like the government sending out a goon to kill me; or at least it is like sending out wild animals; or at the very least like releasing known, violent psychopaths.
The real-world implication of hypothetical (4) is prosecutorial discretion. In the U.S., prosecutors have discretion to investigate and prosecute only some crime. The reasoning is that prosecuting less than 100% of all crimes will still deter criminals, especially if you pick high-profile, visible crimes to prosecute; it will be sufficient to achieve the ends of the criminal law. But is there a minimum level of crime that prosecutors must prosecute? And would there be a private right of action to ensure this minimum level is met? Unfortunately, this question is not far-fetched. There have been several studies of racial bias in the use of prosecutorial discretion. Poor black neighborhoods may be left lawless, their murderers unprosecuted.
Notice the consequentialist argument: A certain level of prosecution of crime is sufficient for deterrence, for the prevention of crime. This argument is interesting because it limits rather than justifies punishment. Does it chime with a retributivist justification of punishment? Commonly, a distinction is drawn between the reasons for threatening punishment and the reasons for actually punishing the criminal. If we punish for no other reason than the crime itself (the retributivist approach), can we choose to not punish for reasons other than the crime itself? Put differently, does a retributivist theory of punishment commit society to the punishment of every crime? (You’ll notice that I’ve backed into Kant’s island parable.)
The implication of this line of thought is the inescapability of consequentialist theories of punishment. Perhaps we cannot do without them, and the task is not to use the weaknesses of consequentialist theories of punishment as proof that retributivism is right. Rather, the task is to fix the consequentialist theories. If you know how, please drop me an email.
The Wild West and duels are a staple of American popular culture. In Westerns, the duel marks the show-down between good and bad. But not only good and bad: The duel also marks the edge of the law. It is lawful to shoot the man who drew first; drawing first and killing is murder.
The duel represents a self-defense situation. Killing in self-defense is permissible. The reason for this exception to the general prohibition against killing is that the state and law are unavailable to protect the assailed citizen. Citizens are not permitted to resort to violence to kill, not even the guilty. The husband returning to find his wife murdered and the murderer still present is required to call the police and let the state do the killing (in the form of the death penalty), or more broadly, the punishment of the criminal, for him. The state’s monopoly on the use of force against law-breakers does not apply when the state is unavailable: If the police are not there to protect the citizen, the citizen may use force to defend himself or herself against an unlawful attack. That is the meaning of self-defense as a justification.
It follows that self-defense marks the edge of the law. Inside the law, the protection of citizens is the task and prerogative of the state. Outside the law (call it the state of nature, if you will), everyone has to fend for themselves, and the only right is the might to either defend oneself or to take what one wants. Unlike the old European nations, whose memory doesn’t reach back to times when wild nature was brought under the sway of civilization, America remembers this time very clearly: the Wild West. The sense of an elementary struggle between civilization and savagery, between law and outlaw, is present in the American mind. While the times of the frontier are long gone (and the entire period lasted only fifty years or so), it is remarkable in that it was mythologized already as it happened. William “Buffalo Bill” Cody (1846-1917), went from Buffalo hunting in the real Wild West, beset by outlaws and horse-thieves, to the stage: to the symbolic representation of the Wild West. Wild Bill Hickock (James Butler, 1837-1876), one of the more notorious gun slingers and at one time marshal of Abilene, Kansas, appeared in Cody’s Wild West Show several times before returning to the real Wild West, where he was murdered during a poker game in Deadwood, Dakota Territory.
In the meantime the law of self-defense underwent a significant change. The original justification at Common Law required the attacked person to flee from the attack where possible and to resort to deadly force against the attacker only where retreat wasn’t possible. In 1876, the Ohio supreme court ruled that a “a true man is not obligated to fly” from an assailant. Similar rulings followed in many courts. Oliver Wendell Holmes wrote the opinion for the U.S. Supreme Court decision Brown v. U.S., 256 U.S. 335 (1921), in which the right to not retreat but to shoot and kill an attacker armed with a knife, was recognized. Holmes is said to have remarked that in Texas, above all, a man is not born to run away.
If you take these three elements together: The myth of the Wild West, the duel as the stylized clash of right and wrong, and self-defense as the demarcation of law and untamed nature; what emerges is the cultural perception of the edge of law. There is a sense in America that there is a limit to law and society, that there is a space outside law and society. Civilization, society, and (on a technical level) law are about inclusion and exclusion. The Wild West is the time in American history when the struggle for inclusion and exclusion was most palpable. The Frontier is the physical symbol for the space outside civilization and society.
More about inclusion and exclusion in American law & society will follow.
The historic references in this entry are drawn from Paul Johnson’s excellent A History of the American People (1999).
Having just finished reading Michael Pawlik’s marvelous new book Person, Subjekt, B??rger (Person, Subject, Citizen), one of the most sophisticated defenses of a retributivist theory of punishment so far in both the English and German literature, I feel compelled to re-examine whether the arguments for a consequentialist theory of punishment are really defective. Most retributivists, and Pawlik is no exception, back into their positions by finding fault with some critical part of the consequentialist account. Negative general and special prevention are said to violate the personality principle, by treating the offender as a means to an end, specifically, as a means to someone else’s end. Positive general prevention, once heralded as an alternative to both consequentialism and retributivism, has turned out to be a useful descriptive theory, but has not fulfilled its promise as a normative theory. Against this backdrop, I would like to examine the question, whether satisfying the personality principle (or any other version of the end in itself test) is really a necessary condition for fully justified punishment under conditions of modernity. I suspect that by combining sophisticated rational actor theories of modern behavioral economics (to build a model of the subject) and system theory (to build a model of society), the implausibilities of most consequentialist accounts can either be avoided entirely or at least significantly mitigated.
