Punishment

Oxford Handbooks Online

Punishment
Michael Tonry
The Oxford Handbook of Crime and Criminal Justice
Edited by Michael Tonry

Print Publication Date: Sep 2011 Subject: Criminology and Criminal Justice, Punishment Theories
Online Publication Date: Sep DOI: 10.1093/oxfordhb/9780195395082.013.0004
2012

Abstract and Keywords

This article discusses important ways of thinking about punishment. Section I discusses the history and recent past
of work by philosophers and legal theorists. Section II surveys major works by social theorists, and Section III
sketches work by political scientists on the day-to-day workings of criminal courts. The conclusion discusses a
number of difficult issues concerning punishment that implicate all of the different ways of thinking about it.

Keywords: crime, philosophers, legal theory, social theory, political scientists, criminal courts

We are not very good at talking about punishment of offenders. There are many different ways to talk about it and
many different reasons to do so. Practitioners talk about it in relation to their ideas about appropriate outcomes of
cases; policymakers in relation to public safety, public opinion, and political self-interest; philosophers in relation to
abstract consideration of retributive and distributive justice; social scientists in relation to what happens in courts
and prisons, and (so far as they can tell) why; and victims and offenders in relation to their personal experiences
and senses of justice and injustice.

The people doing the talking seldom do it outside their own settings, or to people outside their own circles.
American judges and lawyers want to make appropriate decisions about individual cases, but subject to practical
constraints; most, for example, believe that case pressures require that they offer defendants inducements to
plead guilty. Legislators usually want to enact sensible laws that are effective; many believe, however, that it is
sometimes appropriate to authorize or mandate unusually harsh punishments because public opinion is more than
usually concerned about crime or outraged about a recent case. Philosophers want decisions about punishment to
result from rational and detached reflection on what justice requires in this case under these circumstances. Many
victims want punishment to restore their losses and to vindicate their sense of grievance. Many offenders want it to
take account of the circumstances of their lives and the immediate situations in which crimes occurred. Minority
group advocates want punishment policies to take account of the implications of alternate approaches for the
interests of members of groups they care about.

That diversity of interests and perspectives means that most people’s views of the subject are partial and distorted,
like those of the blind men who were each (p. 96) asked to touch a different part of an elephant’s body, and then
to describe what they thought an elephant looked like. One common depiction of the punishment elephant is that it
is a rational if complex and flawed human process in which individuals attempt to make sensible decisions in light of
what they believe to be applicable rules, relevant normative considerations, and limited resources. In dealing with
individual cases, most people usually try to “do justice.” In the narrowest traditional ways of thinking, “doing
justice” involves careful assessment of a series of questions about offenders’ culpability, the harm they caused or
risked, and the likely consequences for society and offenders of possible alternative punishments that might be
ordered. This is how judges and lawyers usually talk about it.

The description in the preceding paragraph is not incorrect, but it is incomplete. It does not take account of the

Page 1 of 22
 Punishment

aims of policymakers when they criminalize behavior or enact punishment laws. They tend not to focus on
individual cases, or on how laws might affect individual cases but on other considerations. Sometimes they want to
reduce victimization generally by deterring, incapacitating, and rehabilitating offenders. Sometimes they want to
denounce and discourage particular troubling behaviors (new ones in recent years, for example, include Internet
frauds and identity theft) by making them illegal and authorizing punishments for them. Sometimes they want to
send moral messages (in recent years, for example, concerning the unacceptability of drug use and domestic
violence), almost without regard to whether the law or punishment will reduce the incidence of the proscribed
behaviors. Sometimes they want to enact expressive or symbolic legislation that responds to public anxieties,
emotions, and fears, whether or not there are reasonable grounds for believing the new laws will have useful
effects. Sometimes they want to make ideological statements about themselves and their political beliefs. Many
laws enacted to serve purposes other than crime prevention or imposition of just punishments create arbitrary
categories or rigid rules that result in punishments many practitioners consider inappropriate or unjust.

The traditional description of punishment does not take account of ideas that have emerged or reemerged in the
past forty years. Adherents of restorative (e.g., Braithwaite 2001) and community justice (e.g., Clear and Karp
1999) believe that punishment policies and processes should serve primarily to resolve conflicts, solve problems,
and strengthen communities, and that traditional criminal justice processes and ideas about punishment should be
reserved for a residual category of cases for which newer approaches are not feasible. Adherents of therapeutic
jurisprudence argue that practitioners in every case should take account of the therapeutic (and possible
destructive) effects of every action they take and process in which they participate—in regard to offenders,
victims, and practitioners themselves (Wexler and Winick 2003).

Nor does the narrow traditional description of punishment as “doing justice” take account of other institutions and
processes that shape it and cause it to perform social functions that most analysts and courtroom officials do not
recognize. These include such things as reinforcing social norms (e.g., Durkheim 1933 (p. 97) [1893]), protecting
the interests of economically or politically powerful groups (e.g., Rusche and Kircheimer 1968 [1939]), maintaining
existing patterns of racial, ethnic, and social hierarchy (e.g., Wacquant 2002a, 2002b), and shaping people to play
roles required by the social and economic systems of their times (e.g., Foucault 1979). Punishment thus sometimes
serves functional ends unrelated to its nominal purposes.

But it is still more complicated. Punishment laws, institutions, and practices are shaped by personal interests of
individuals (for example, the prosecutor who wants to be re-elected and the assistant prosecutor who wants to be
promoted; the legislator who wants to become governor; the legislative faction that agrees to vote in favor of a bill
on one subject in exchange for another faction’s agreement to vote in favor of a bill on some other subject).
Punishment decisions are also shaped by the institutional interests of organizations (for example, most cases have
to be dismissed or dealt with by plea bargains because there are not enough assistant prosecutors, courtrooms,
and jail cells). They are influenced by professional social pressures; courtrooms are complex places, and they
function better if people respect local expectations and conventions. They are also influenced by serendipity—the
judge who is afraid of young black men, the prosecutor who is highly emotional about sex crimes against children,
the probation officer with a Little Caesar complex.

Here are some of the different ways of thinking and talking about punishment:

• Some philosophers and legal theorists regard punishment policies and practices as systems for calibrating
amounts of punishment to offenders’ moral fault.
• Other philosophers and legal theorists regard punishment policies and practices as systems for maximizing
happiness or preventing crime.
• Still other philosophers and legal theorists believe that punishment policies and practices should take account
of both moral fault and crime prevention considerations.
• And others still believe that punishment policies and practices serve primarily to reinforce or undermine basic
social norms.
• Some analysts believe that punishment policies and processes (usually operating under other names such as
restorative justice, community justice, or therapeutic jurisprudence) should serve primarily to resolve conflicts,
solve problems, and strengthen communities.

Page 2 of 22
 Punishment

• Some social theorists explain punishment policies and decisions as devices for achieving functional goals
only peripherally related to crime or punishment.
• Some analysts explain punishment policies and decisions as devices for realizing personal goals of officials or
institutional goals of organizations that may or may not be closely related to crime and punishment.
(p. 98) • Some analysts explain punishment policies and decisions as primarily expressive, as forms of
communication through which the state and state actors express solidarity with social values, acknowledge
public fears and resentments, and respond to public concerns.

These different ways of talking about punishment come to a head in difficult cases and troubled times. Northern,
personally abolitionist judges called upon to enforce “Fugitive Slave Laws” before the Civil War offer a classic
example (Cover 1984). Was any option open to them other than to enforce the law or resign on principle? Should
they have disregarded their moral beliefs about slavery or disregarded the law? Should they have refused to
enforce the law because doing so would sustain slavery and further the interests of slave owners, or should they
have enforced the law because not doing so would undermine rule-of-law ideas upon which democracies depend?
Should they have hypocritically ducked the problem by looking for and finding technical reasons to dismiss the
cases other than on the merits? Should they have taken into account the implications for their own careers or
reputations of whatever choices they made? Should they have taken into account the complex political calculus
that led to the federal Missouri Compromise provision that fugitive slave laws should be enforced in states that did
not allow slavery under their own laws?

Non-Nazi professional judges in Nazi Germany are another classic example (Hart 1968). What should they have
done when facing defendants charged, accurately, with violating a law that made it a crime to provide assistance
to a Jew? Should they have refused to enforce the laws because they were immoral, or because their enforcement
might undermine social norms calling for compassionate assistance to troubled people, or because their
enforcement reinforced the legitimacy of the Nazi regime? Should they have taken into account dangers to
themselves or their families if they did not enforce the laws? Were they entitled to take account of the implications
of their choices for their own later careers?

United States federal district court judges offer a less obvious but more apt—because more familiar—example.
When called upon by federal law to impose a mandatory minimum sentence of twenty years on a young first
offender—the first college-attending member of a poor black family—who agreed for a few dollars to carry a
briefcase containing 100 grams of crack from St. Louis to Minneapolis, what should a judge do? Should he or she
take account of the defendant’s personal background and sympathy for the plight of disadvantaged black
Americans at this historical moment? Should the judge impose the sentence, and disregard his or her personal
belief that the sentence is unjust because grotesquely severe? Should the judge defy the law and impose some
other sentence (or figure out a disingenuous way to avoid doing so, as by refusing to accept a guilty plea and
acquitting the obviously guilty defendant)? Should the judge be influenced by knowledge that alcohol, a substance
commonly used by the middle and upper classes that causes much (p. 99) greater human suffering and
economic loss than any other recreational substance, is legally available, while less dangerous substances like
marijuana, cocaine, and heroin, more commonly used by poorer people, are illegal? Should the judge be influenced
by knowledge that cocaine-offense defendants, mostly black, who sell crack are subject to much harsher penalties
than those, mostly white, who sell powder? However polemically or neutrally the observation is made, it cannot be
pure coincidence that the law more heavily punishes the misbehavior of the less well-off and the nonwhite. Should
that matter when the judge sets a sentence or when observers try to understand what is going on?

These are, of course, extreme cases. Most criminal cases are more mundane. What these cases share with
ordinary cases, however, is that much more was going on, and needs to be understood, than the simple
mechanical (or the subtle, principled) application to individual cases of legal rules about punishment. The narrow
traditional picture of punishment as “doing justice” shows only a small part of the story.

Different kinds of analysts ponder different facets of punishment. Philosophers try to clarify the relevant normative
principles. Legal theorists try to relate principles to policies and practices. Social theorists try to understand why
punishment takes the forms it does in particular places and times. Political scientists explain punishment as the
product of political judgments shaped by the interests of groups in society and of individual politicians, or as the
outcome of institutional considerations that constrain the choices of individuals operating in complex organizational

Page 3 of 22
 Punishment

settings. Most of these theorists do not talk to each other, or take account of one another’s perspectives, insights,
and writings. And few exponents of the diverse scholarly perspectives are very good at communicating with
practitioners or public officials. All of the separate ways of thinking and talking about punishment are incomplete.
Often they operate at cross-purposes.

The rest of this chapter more systematically discusses important ways of thinking about punishment. Section I
discusses the history and recent past of work by philosophers and legal theorists. Section II surveys major works
by social theorists, and section III sketches work by political scientists on the day-to-day workings of criminal
courts. The conclusion discusses a number of difficult issues concerning punishment that implicate all of the
different ways of thinking about it.

I. Philosophers and Legal Theorists

Consider capital punishment of murderers. Some people believe it is a Good Thing, because it deters would-be
killers and thereby saves lives. Others believe it is a Bad Thing, because it does not deter would-be killers, and
sacrifices the lives of murderers (p. 100) for no good purpose. Still others think disagreements about the practical
effects of capital punishment are irrelevant. They believe that some murderers should be killed, because they
deserve it, or that no murderers should be killed, because no one deserves that, irrespective of what the effects of
the killings might be.

A. Philosophers

It is common to describe the arguments about good and bad effects as consequentialist (or teleological), and the
arguments about moral prescription as deontological. Both are kinds of moral arguments, but they differ in what
they mean by morality. Consequentialists believe that means must be justified by their ends. Murderers may be
executed, if doing so on balance accomplishes something good. If killing murderers saves lives, or achieves other
valid public purposes that outweigh offenders’ losses of life, then that is fine. If it does not, then we can’t do it.
Jeremy Bentham (1970), the nineteenth-century English utilitarian theorist, was adamant that punishments could not
properly be imposed if they were “inefficacious,” by which he meant that their imposition would not make future
offending less likely. He used the term parsimony as shorthand for the proposition that no pain should be imposed
on wrongdoers in excess of that required to achieve valid social aims. If no such aims could be realized, then
parsimony forbade infliction of any punishment at all.

Unfortunately, no term of art exists that satisfactorily categorizes people who subscribe to deontological (or
nonconsequentialist) views. Few people write or talk about “deontologists,” probably because the term is outside
ordinary usage and sounds arcane or pretentious. “Nonconsequentialist” is unsatisfactory because it is a mouthful
and is a definition by negation. It indicates nothing about what people believe. Sometimes the term retributivist is
used in relation to punishment, but this does not work for capital punishment because some retributivists oppose it
on moral grounds while others who oppose it on moral grounds also oppose retributivism. Those who support and
oppose capital punishment on deontological grounds might be called moralists, but this too is unsatisfactory
because it implies that consequentialist arguments are not also moral arguments. They are. Nonetheless, I follow
common practice below and sometimes refer to deontological arguments as moral or moralistic ones.

People who subscribe to deontological theories believe that means must be justified per se, irrespective of ends. If
morality requires that offenders be executed, then they should be. The German idealist philosopher Immanuel Kant
(1965 [1787]) believed that the capacity for moral choice is what makes human beings human, and that failure to
attach consequences to moral choices implies that the person who made them is not morally responsible and
therefore not human. Respect for peoples’ autonomy thus requires that their wrongdoing be punished in the
appropriate way.

(p. 101) Kant described a hypothetical island society that was about to dissolve, with its members to disperse to
the ends of the earth. Kant famously argued that the murderers in prison should be executed before the ships
departed, even though no possible public benefit could be realized from their killings. His hypothetical was
designed to isolate the moral case for capital punishment. If the island society was going to dissolve, deterrent and
incapacitative considerations were irrelevant. No one would remain to be deterred or to benefit from reduced
threats of future crime resulting from the killer’s incapacitation. The sole question then became, Should the

Page 4 of 22
 Punishment

murderer be executed if there is no possible consequential benefit? Kant’s answer was yes. Conversely, if morality
forbids the taking of life by the state, then murderers should not be executed.

The “morality” referred to here is not like consequentialist morality, which in principle derives from knowledge of
the world: capital punishment does or does not deter would-be murderers. When we are confident we know
whether it does, we will know what consequentialist morality requires. For people who subscribe to deontological
punishment beliefs, the sources of morality are a priori. Moral truths come from God or his agents on earth or are
inferable from knowledge about the universe or are simply (alas) the unexamined, culturally transmitted,
conventional wisdom of particular groups.

These two sets of views create the frameworks within which philosophers analyze punishment problems.
Consequentialists and “retributivists” operate on different ground. The former do not acknowledge a priori moral
claims; they want evidence. Retributivists and other “moralists” do not care about evidence; they want moral
clarity. Concerning capital punishment, there are also irreconcilable differences among the “moralists.” Some
believe that a priori morality requires it, and others believe that a priori morality forbids it.

Thus there are at least three sets of irreconcilable views about capital punishment. This might raise skepticism
about the value of philosophical analyses of punishment. If philosophers cannot agree among themselves or find
ways to resolve irreconcilable differences among themselves, how can they help nonphilosophers gain increased
understanding? The answer is that people other than philosophers also divide into those who instinctively
subscribe to retributivist and consequentialist views, and if philosophers can help clarify and enrich understanding
of each, that is a good thing. It will help the rest of us clarify our own views and be able more clearly to understand
our differences with others. It may sometimes cause us to change our minds.

Some philosophers say that they work in the realm of “ideal theory” which is premised on the application of their
ideas in an ideal, or just, world. If the world is not just, as almost every philosopher who writes about punishment
acknowledges, ideal theories cannot tell us what to do (Murphy 1973; Duff 1986; Honderich 2006). They can,
however, provide frames of reference for characterizing and criticizing imperfect institutions in an imperfect world,
and thereby provide inspiration for (at least) incremental improvements.

(p. 102) Legal, social, and political theorists also have essential and fundamental insights to offer to thinking
about punishment. In writing about their views, I shift from a focus on capital punishment to criminal punishment
generally. Capital punishment nicely demonstrates differences between consequentialist and opposed moralistic
views, but it is seldom imposed. Each year, however, millions of other criminal punishments are imposed in the
United States (and equivalently large numbers in other countries). These range from minor fines to sentences of
imprisonment for life without the possibility of parole. The stakes are lower than when death is not a possibility, but
similar questions are raised about just punishments.

For punishment generally, “retributivist” is an adequate alternative to consequentialist because the opposed a
priori moral views about capital punishment do not arise for other punishments. Although retributivists believe that
offenders should be punished because they deserve it, just as some retributivists believe that some offenders
should be executed because they deserve it, no one argues that punishments short of death should never be
imposed (as moralistic death penalty opponents argue that death should never be imposed).1 So from this point I
no longer occasionally refer to moralistic but instead to retributive theories.2

It is commonly said that retributive and consequentialist positions are irreconcilable. Retributivists say that, in
principle at least, no matter what happens in the real world, consequences do not matter. Consequentialists say
that in principle any imposition of pain, including punishment of offenders, is a bad thing, and can be justified only
by reference to more-than-offsetting good consequences. For a very long time, a shouting match resulted.

Retributivists argued that the logic of utilitarianism required that utilitarians (then the only category of
consequentialists3 ) approve punishment of innocent offenders if sufficiently great beneficial consequences could
be expected. A commonly used illustration was the prosecution (and potentially the execution) in the early
twentieth-century American South of a black, alleged rapist who was known (only) by the judge to be innocent if
that would prevent the lynchings of other innocent black men (e.g., McCloskey 1968). Retributivists charged that
no morally acceptable theory of punishment could countenance such a result.

Page 5 of 22
 Punishment

Utilitarians denied that attack. Usually they argued that such actions could never be kept secret and that the legal
system’s legitimacy would be undermined. Huge public insecurities would result if citizens knew that they could be
“punished” even if they had committed no crimes. Those negative consequences of the punishment of innocents,
utilitarians argued, would outweigh any possible gain. Sometimes they argued that the punishment-of-the-innocent
problem is not a difficulty for utilitarian theories because criminal punishment presupposes that a crime has been
committed and consequently, by definition, an innocent cannot be “punished.”4

Moving from defense to offense, utilitarians argued that retributivism, whatever its claim to philosophical purity, was
at base no more than the expression of (p. 103) vindictive and vengeful human instincts, which should not be
given free rein. Retributivists, of course, denied this.

The philosophy of punishment was not a major specialty until the 1960s. Until then, philosophers worked within their
own traditions and disagreed among themselves. It is only a small exaggeration to say that no one else much
noticed or cared. Most criminal justice practitioners and legal theorists, while acknowledging that retributive
instincts existed and had sometimes to be acknowledged, subscribed primarily to consequentialism (e.g., Michael
and Adler 1933; Michael and Wechsler 1937; Henry Hart 1958; Allen 1959, 1964). Rehabilitative and incapacitative
ideas were widespread. The Model Penal Code (American Law Institute 1962), the most influential criminal law
document of the twentieth century, illustrates this. The Code, drafted in the 1950s, nowhere mentions “just
deserts,” “just punishment,” “deserved punishment,” or retribution as aims of sentencing or of the criminal law
generally. Its punishment provisions aim to empower judges, parole boards, and probation officers to tailor
punishments to individualized consideration of offenders’ needs for rehabilitation or incapacitation (Tonry 2004,
chaps. 6, 7).

The primacy of consequentialist ideas began to wane in the 1960s,5 and by the 1970s a retributive resurgence
occurred. There was renewed interest in the philosophy of punishment. The first few major works attempted to
reconcile retributive and utilitarian theories. John Rawls (1955) and Edmond Pincoffs (1966) offered analyses that
assigned different kinds of theories to different realms. Utilitarian considerations were said to pertain to general
legislative consideration of criminal law doctrine and statutory frameworks for sentencing decisions. Retributive
considerations were said to be germane to judges’ decisions about punishment in individual cases. For a variety of
reasons, that analysis persuaded few people.

A breakthrough occurred in Punishment and Responsibility (1968), a book by the Oxford philosopher H. L. A. Hart.
Hart argued that the conflict between utilitarians and retributivists had been misconceived because it was based on
the assumption that criminal punishment raised only one fundamental question: “How is punishment to be
justified?” To the contrary, Hart argued, there are three separate important questions:

1. General Justifying Aim. How is the existence of a state institution of punishment to be justified?
2. Liability. Who may be punished?
3. Amount. How much punishment may be imposed?

Hart argued that a comprehensive theory of punishment might coherently provide different answers to those
questions. His own, which he called a “middle way,” was that prevention, a consequentialist idea, is the general
justifying aim, that—possibly with some exceptions6 —punishment should be imposed only on offenders, for
offenses, and that both retributive and consequentialist considerations were relevant to the amount of punishment.

(p. 104) For some kinds of theories, for example, what might be called “thorough-going retributivism,” Hart’s three
questions changed nothing. For a Kantian theorist who believed that moral principles require that offenders be
punished (e.g., the island example), essentially the same answers would be given to all three questions. What is
the general justifying aim of punishment? The imposition of morally deserved punishments. Who may be punished?
Offenders, for offenses. How much? As much as is deserved.

For other theories, however, things did change. A consequentialist could neatly sidestep the punishment-of-the-
innocent challenge. Crime prevention, or positive social utility, or happiness, or economic efficiency could be the
justifying aim, and decisions about the amount of punishment could be predicated on maximizing it. Liability to
punishment, however, could be based on retributive considerations.

I describe this as a breakthrough because it allowed a place for what have become known as mixed or hybrid

Page 6 of 22
 Punishment

theories (like Hart’s own). Academic lawyers leaped into the opening Hart created.

B. Legal Theorists

Legal theorists try to devise punishment theories that take account of the world’s imperfections and can inform real
decisions in a real world. If perfect justice cannot be achieved, efforts to achieve justice in practice can at least be
improved. Many people believe that punishment decisions should be related to offenders’ moral responsibility and
desert, and also that punishment decisions should take account of the likely crime-preventive effects of alternate
punishment possibilities. Theorists have therefore tried to figure out ways to bridge gaps between what I described
as irreconcilable views in philosophy. They might be thought of as applied philosophers. Typically they propose
hybrid theories.

In 1975 utilitarian ideas had been predominant in the English-speaking countries for a century, and served as the
theoretical rationale for what we now call indeterminate sentencing.7 That system, which some American states
today largely retain, gave officials broad discretion to individualize sentences to take account of individual
offenders’ circumstances. Legislatures defined crimes and set maximum authorized sentences. Judges could
usually impose any sentence, ranging from unsupervised probation to the statutory maximum prison sentence.
Prison officials could shorten prisoners’ sentences to take account of good behavior, participation in treatment
programs, and work, and could oppose or support prisoners’ applications for parole release. Parole boards could
release any prisoner who was eligible. In the most far-reaching indeterminate systems, in California and
Washington State, judges sentenced offenders to the state prison system for a term from one year up to the
statutory maximum; parole boards determined when they were released (Rothman 1971, 1980).

In the early 1970s, indeterminate sentencing fell out of favor. A number of things contributed to this. First, people
inside and outside the legal system began to be (p. 105) concerned about the possibilities of biased and
idiosyncratic decisions by officials accorded such broad discretion (Davis 1969). One influential critique
concerned unwarranted sentencing disparities (Frankel 1972). Second, the civil rights and prisoners’ rights
movements focused attention on the disproportionate number of blacks in prison and on the possibility that
indeterminate sentencing produced biased decisions that caused and worsened racial disparities (American
Friends Service Committee 1971). Third, the U.S. Supreme Court in the 1960s issued a number of decisions—most
notably Goldberg v. Kelly, 397 U.S. 254 (1970)—extending procedural rights to citizens in a wide range of their
interactions with the state. People affected by government decisions were at minimum entitled to be told what rule
they were accused of violating, to present evidence on their own behalf, to have the issues decided by an
independent fact finder, and to appeal decisions adverse to their interests. Critics soon observed that by those
standards prosecution and parole systems (Davis 1969) and sentencing (Frankel 1972) were essentially “lawless.”
There were no clear rules governing such decisions and possibilities of appeal were either nonexistent or erratic.
This critique had particular relevance to a sentencing process that provided no guidance to judges about their
decisions and afforded defendants no opportunities to appeal the sentences they received. Fourth, researchers
evaluating the effectiveness of rehabilitative programs, which in large part gave indeterminate sentencing its
rationale, concluded that few programs if any could be shown to reduce re-offending. The most famous synthesis
of treatment effectiveness studies, “What Works? Questions and Answers about Prison Reform” (Martinson 1974),
was generally interpreted to conclude that “Nothing Works” (it didn’t really; its conclusions were much more
qualified than that, but what it was perceived to conclude proved much more influential than what it did conclude).

This indictment that indeterminate sentencing caused unwarranted disparities, facilitated racially biased decision-
making, lacked minimum procedural safeguards, and was based on assumptions about rehabilitative programs that
were unwarranted was devastating. The results include the development of sentencing guidelines, “truth-in-
sentencing,” mandatory minimum sentence laws, three-strikes laws, and abolition of parole.8 The indictment also
left a void in normative thinking and writing. Few people had been writing about punishment based on retributive
premises.

The obvious answer to the problems of indeterminate sentencing was to establish some form of rules to guide
judicial and parole decision-making. Retributive ideas provided a ready rationale for rules. If a principal criterion of
punishment is that it be morally deserved, attention is drawn almost inexorably to the seriousness of the crime as a
measure of moral desert and to the idea of proportionality—harsher punishments for more serious crimes,
comparable punishments for comparably serious crimes. From there the leap was easy to sentencing guidelines

Page 7 of 22
 Punishment

that ranked offenses in terms of severity and criminal records in terms of extensiveness, and directed judges to
base their decisions on those factors.

(p. 106) Some philosophers, such as John Rawls (1955), Edmond Pincoffs (1966), and H. L. A. Hart (1968), had
anticipated the change in direction, and a few philosophers began to take account of it in their writing (e.g., H.
Morris 1966; Kleinig 1973), but legal theorists quickly staked out the territory. They had two comparative
advantages. Many—notably Alan Dershowitz, Norval Morris, Andrew von Hirsch, Franklin Zimring, and the writer of
this chapter—were involved in the sentencing reform movement as law reform activists and as scholars. This gave
them familiarity with the real-world issues involved. In addition, most were primarily interested in normative
analyses of sentencing (Hart’s “How much?” question), unlike philosophers who had traditionally written about, in
Hart’s terms, the General Justifying Aim.

Most of the analyses offered by legal theorists were hybrid theories with significant retributive elements. Sentencing
guidelines based on retributive ideas addressed all the major critiques of indeterminate sentencing: the absence of
rules and resulting unwarranted disparities—guidelines for decisions; broad discretions facilitating racially biased
decisions—narrowed discretion constrained by guidelines; lack of procedural fairness—published guidelines
coupled with a right to appeal decisions inconsistent with them; ineffective treatment programs—don’t allow judges
to base sentences on treatment considerations.

The legal theorists rushed in (e.g., Frankel 1972; N. Morris 1974; Dershowitz 1976; von Hirsch 1976; Morris and
Tonry 1978; Singer 1979). Two perspectives quickly became predominant. The first, called desert or
proportionality theories, proposed that punishments be linked to scales of offense severity and that little latitude
be accorded judges to take account of consequentialist concerns. The leading figure is Andrew von Hirsch of
Cambridge (1976, 1985, 1993, 2005 [with Ashworth]). The second, called limiting retributivist theories, are most
famously associated with the late Professor Norval Morris of the University of Chicago (1974, 1990 [with Tonry]).
Limiting retributivist theories posit that there is a range of not-undeserved punishments and that, within that range
and assuming other criteria are satisfied, judges may take consequentialist considerations into account. In the
1980s, desert theories were highly influential and government bodies as diverse as the Australian Law Reform
Commission (1980), the Canadian Sentencing Commission (1987), the Minnesota Sentencing Guidelines
Commission (1980), and the Home Office of England and Wales (1990) explicitly adopted desert theory as their
guiding normative rationale. More recently, limiting retributivist theories have become more influential as evidenced
by their adoption by the American Law Institute (2007; in the Model Penal Code, 2nd ed.) and a major English
sentencing reform body (Home Office 2001) as guiding rationales.

C. The Return of the Philosophers

Interest in retributive punishment theories revived among philosophers at the same time as legal theorists argued
among themselves. For the most part they focus primarily on Hart’s first question. Intuitionist theorists argue that
the widely shared (p. 107) intuition that people who commit serious crimes deserve to be punished provides an
adequate and persuasive justification for retributive punishment (e.g., Moore 1993). Benefits and burdens or
equilibrium theorists start from a social-contract premise by which citizens are deemed to have assented to rules
and conventions that provide a sense of security, which enables each to plan and live the life they choose: 9
offenders benefit from other peoples’ law-abiding ways and obtain an unfair advantage or benefit, which must be
repaid, or a law-abiding equilibrium restored, when they offend against others (e.g., Murphy 1973). Paternalist
theorists argue that the imposition of deserved punishments provides an occasion to teach offenders about right
values in order to enable them later to live more satisfying, law-abiding lives (e.g., H. Morris 1981). Censure
theorists argue that punishment is a blaming institution that appropriately censures offenders in proportion to the
seriousness of their wrongdoing (e.g., von Hirsch 1993). Communicative theorists argue that imposition of
deserved punishments provides a noncoercive occasion to help offenders and others understand why offenders
deserve to be punished in order that they may come to share that understanding (e.g., Hampton 1984; Murphy and
Hampton 1988). Communitarian theorists argue that punishment provides an occasion to help offenders
understand and comply with the values of communities of which they are members and thereby enable them more
fully to enjoy the benefits of membership in a socially organized community (e.g., Duff 2001).

Legal theorists and philosophers cannot tell us all that we need to know about punishment. They both start from the
assumption that punishments result from rational processes in which individual judges and other practitioners make

Page 8 of 22
 Punishment

better or worse decisions about individual cases. Social theorists and political scientists instruct that that is an
oversimplified view of the world.

II. Social Theories

Life, and punishment, are more complicated than philosophers and legal theorists usually want to acknowledge.
Social theorists try to describe the functions punishment performs and to explain why it takes the forms that it does,
rather than to justify it in normative terms. Social institutions and practices exist for reasons, and the reasons are
not always obvious.

Social theorists often write about punishment’s functions. By this they seldom mean, as lawyers and philosophers
might, crime-preventive utilitarian functions (or purposes) such as rehabilitation, deterrence, and incapacitation.
Instead they are interested in what broader social functions are served. Work by the sociologist Loïc Wacquant
provides a good illustration (2002a, 2002b). In a series of articles published since 2000, he argues that modern
American punishment policies and (p. 108) practices serve to maintain patterns of social, economic, and political
hierarchy in which blacks as a group are kept subordinate to whites. Punishment, in this analysis, is the latest in a
series of social and legal institutions that have in turn maintained a racial hierarchy in which whites dominate
blacks. The earlier ones were slavery, the “Jim Crow laws” and social conventions that allowed explicit racial
discrimination after the Civil War, and the urban ghettos of mid- and late twentieth century America. When one form
of hierarchy maintenance became no longer possible or effective, another one came into being. Lincoln’s
Emancipation Proclamation ended slavery, but Jim Crow took its place. The civil rights movement of the 1950s and
1960s ended legal forms of racial discrimination, but the urban ghetto kept blacks in their marginalized place. As
American cities revived late in the twentieth century, the law-and-order movement that began in the 1960s took
hold and assured that the life chances of blacks, especially poor black males, remained inferior to those of whites.
As a result, black men have a one-third chance of spending time in prison during their lives, compared with 11.3
percent of whites, and black men’s imprisonment rates are six to seven times those of whites (Bureau of Justice
Statistics 2003; Tonry 2011). Criminal justice policies and practices have the effect of assuring that blacks fare
worse than whites by every measure of social and economic well-being (others have developed similar arguments:
e.g., Loury 2007; Massey 2007).

Functionalist accounts are not conspiracy theories. Wacquant does not claim that a self-perpetuating cabal of
racist whites has met regularly for two centuries and devised and implemented a succession of strategies for
disempowering blacks and diminishing their chances of achieving satisfying lives. The argument instead is that
whites as a group realize psychic and other benefits from their dominant economic, social, and political positions
compared with blacks, and thus they develop institutions—often unconsciously and often rationalized in neutral,
nonracial terms—that operate to maintain white dominance. Of course, there are always racists and bigots of other
kinds, and they certainly played major roles in perpetuating slavery and Jim Crow. During the periods when the
urban ghetto and law-and-order policies have damaged black interests, bigotry, however, has been much less
evident and its presence has been vigorously denied. Policies on zoning, lending, and mortgage insurance that
long limited blacks’ housing opportunities, for example, were explained in terms of neighborhood ambience,
protection of property values, and sound investment practices. They were never openly rationalized in terms of
preferences for black exclusion from middle-class areas and concentration in disadvantaged areas, but that is
what they accomplished (e.g., Massey and Denton 1993).

The best single piece of evidence for Wacquant’s case in the criminal justice system is a federal law passed in
1986, the “100-to-1 rule” (in 2010 it was replaced by an “18-to-1 rule”). It punished crack cocaine offenses, for
which most of those arrested are black, as severely as powder cocaine offenses 100 times larger (for which most
of those arrested are white). That law quickly became the single largest (p. 109) cause of racial disparities in
federal prisons (McDonald and Carlson 1993), and its 18-to-1 successor continues to have that effect. Crack and
powder cocaine are pharmacologically indistinguishable, as state supreme courts have noted in striking down
crack/powder sentencing differentials, and as the U.S. Sentencing Commission has reiterated (1995, 2007). More
generally, the “War on Drugs” launched in the late 1980s disproportionately focuses on crack offenses and has
long generated arrest rates for blacks for drug crimes generally that are three to six times those for whites
(although blacks are less likely than whites to use illicit drugs). These laws are generally justified in terms of the
immorality and dangers of drug use, and are never justified in terms of their effects on poor, black Americans. Yet

Page 9 of 22
 Punishment

they have powerful adverse effects in reducing poor, black (especially male) Americans’ likelihood of achieving
satisfying lives (through diverse mechanisms that include stigma, stereotyping, removal from the labor market,
disqualification from many kinds of jobs, and disenfranchisement from voting). Policies adopted in the past twenty
years to invest less in treatment programs, to make prisoners ineligible for federal Pell grants that fund poor
peoples’ college educations, and to deny ex-prisoners eligibility for a wide range of social service programs
exacerbate the effects of criminal justice policies (e.g., Western 2006; Tonry and Melewski 2008).

Neither housing and mortgage lending practices nor criminal justice policies explicitly aim disproportionately to
damage black Americans or to benefit whites. In both cases, that is indeed what they do, and what they could have
been foreseen to do. Wacquant asks how and why those patterns came to be, and tries to show that the best and
most plausible explanation is the one he offers.

The Scottish sociologist David Garland (2001) tells a different story but reaches a similar conclusion. His account is
in some ways more damning than Wacquant’s because it makes the causes of black overrepresentation in courts
and prisons the products of deliberate eyes-open political decisions rather than of an invisible hand. Although
recent American and English crime-control policies are often publicly rationalized in terms of deterrence and
incapacitation, they are—Garland suggests—actually “expressive” efforts by governments to reassure anxious
publics and gain or regain citizens’ respect. Governments, he argues, recognize the modest limits of their
capacities to affect crime rates and patterns, but want to be seen to be doing something, and loudly proclaim that
what they do will work, whether they believe it or not. Expressive policies once adopted need to be implemented,
and to be seen to be implemented, but it must be done in such a way that it does not undermine government’s
credibility and support among the broad electorate and opinion leaders. Targeting expressive policies on the
crimes of the dispossessed and groups with little political power (Garland refers to “the criminology of the other”)
enables government to be seen to be tough without seriously undermining its broad political support. In the United
States, the dispossessed and the politically powerless are disproportionately disadvantaged members of minority
groups. In England they are the same groups and socially marginal members of the white, lower working class (p.
110) (English Labour politicians like to denounce “yobs,” “thugs,” and “louts.”) Though Garland travels a route
different from Wacquant’s, the destination is the same: an explanation for why in the United States black Americans
disproportionately comprise criminal court dockets and occupy prison beds.

Social theorists have offered many kinds of functionalist arguments. I used Wacquant’s illustratively because it is
controversial, easy to understand, and speaks to developments of our time. The three most famous functionalist
arguments relating to the criminal law are associated with the French theorist Emile Durkheim, the German theorist
Karl Marx and some of his followers, and the French theorist Michele Foucault.

Durkheim (1933 [1893]) argued that law is primarily a mechanism for declaring, reinforcing, and changing basic
social norms. The criminal law’s ostensible functions of incapacitation, deterrence, and rehabilitation are not, he
argued, very important (or especially effective). Criminal convictions and punishments serve instead to channel
public dissatisfaction and indignation with wrongdoers and through those processes reaffirm the behavioral norms
that were violated. Durkheimian ideas can also be expressed in terms of the law’s dramaturgical effects. Criminal
trials and punishments serve as morality plays that denounce wrong behaviors and validate good ones. Although
practitioners and lawmakers may act and talk as if the criminal law affects behavior by changing or controlling
peoples’ behavior, Durkheim would say, they are mistaken. What they are doing, even if it is not what they think
they are doing, is useful all the same.

Karl Marx might be expected to have written about the criminal law as an institution for furthering the interests of
the dominant economic classes, but he wrote little on the subject (e.g., Marx 1853). Others did, however, most
influentially the German scholars Georg Rusch and Otto Kirschheimer (1968 [1939]). They tried to show that the
criminal law, and especially prisons and capital punishment, were best understood as mechanisms by which
society adapted to changing labor market conditions. When economies are strong and labor is valuable and in
demand, prison use and executions fall: able-bodied workers are too valuable to waste. When economies are
weak and workers are idle and not economically valuable, prisons fill and executions rise (“life is cheaper”).
Countless people have observed that the criminal law focuses much more on poor than on privileged peoples’
crimes, prosecutes them much more often, and punishes them much more severely. As a functionalist matter,
these things are probably not a coincidence.

Page 10 of 22
 Punishment

Michele Foucault (1979) argued that punishment forms vary over time and serve different functions depending on
historical circumstances. Before the Enlightenment, and before the development of modern bureaucratized police,
court, and prison systems, punishments though rare were sometimes repulsively severe and served to emphasize
and celebrate the power of the king. Beginning in the nineteenth century, the primary function of punishment (and
of other mass institutions like schools, armies, industrial factories) became the socialization and reformation (p.
111) of people into conformity with the kinds of roles a mass and heavily bureaucratized society requires be
played. Thus as the function of punishment changed from expressing and celebrating the authority of the king to
socializing people into particular kinds of social and economic roles, the nature and forms of punishment changed.
Corporal and capital punishment and banishment declined in use, and the nineteenth-century’s rehabilitative and
reformative correctional institutions came into being.

To describe an institution or process in terms of social functions does not mean that people are not sometimes
consciously aware of those functions. Presumably some people recognize that mortgage-lending practices that
redline minority neighborhoods do damage to black people, are happy about that, and regard that as a good
reason to maintain the practice. Some people may feel that way about modern American drug policies.

Similarly, people may recognize functions of institutions and want to regularize them. Scandinavian judges and
scholars, for example, subscribe to ideas akin to Durkheim’s about law as a mechanism for reinforcing basic social
norms (e.g., Andenaes 1974). They believe with Durkheim that incapacitation, deterrence, and rehabilitation have
at best marginal influence on crime; except under extreme conditions, people do or do not commit crimes because
they have or have not been socialized into values with which much criminality is incompatible. They also believe
that law’s role is collateral; families, schools, neighborhoods, churches, and other primary institutions do the heavy
lifting in instilling and reinforcing social norms. But they believe it important that the law back up those norms. As a
result, though criminal punishments in Finland are not severe by international standards, punishments are more
likely to be imposed than elsewhere (so that bad behavior has consequences) and are highly proportionate (so
that norms about the comparative seriousness of misconduct are not undermined). Finns refer to law’s “moral-
educative” or “general-preventive” effects as its most important ones (Lappi-Seppälä 2001).

In a somewhat different—but not completely unrelated—vein, English-language theorists have developed other
ideas linked to socialization, norm reinforcement, and moral-educative effects. One of these is the development in
philosophy over the past thirty years of what are called “communicative theories of punishment” (e.g., Hampton
1984; Duff 1986; Murphy and Hampton 1988). Such theories take diverse forms but at their core is the proposition
that punishment as a process ought to be centrally concerned with communication with offenders, observers, or
both, about good values.

A different, and more pernicious, development is the concept of “expressive” policies or punishments. In our time,
many punishment policies, especially in the United States and England, are said primarily to be expressive (e.g.,
Garland 2001). This can mean as many things as there are speakers. In the narrowest sense, it can mean what
Finns mean when they talk about moral education: the idea that criminal (p. 112) law and punishment endorse
prevailing social norms and may reinforce them. Somewhat more broadly, it can mean denunciation of wrongful
behavior and declaration that it is wrong. More broadly, and more commonly, it refers to the adoption of punitive
laws that are meant in some general way to express solidarity with public opinion and reassure the public that its
anxieties and fears have been taken into account and acted upon. Such things as three-strikes-and-you’re-out
laws, lengthy mandatory minimum sentence laws, and life-without-the-possibility-of-parole sentences are
examples.10 One characteristic of expressive policies is that most, for well-known reasons that have long been
understood, cannot accomplish the goals set for them. For example, laws requiring judges and prosecutors to
impose punishments they believe to be unconscionably severe are often circumvented, and are always
inconsistently enforced. Many expressive laws make no sense in instrumental, effectiveness terms. They also
make no sense in the terms in which philosophers and legal theorists, whether retributivists or consequentialists,
talk about punishment.

III. Political Science Accounts

Expressive laws and policies do not necessarily, or even probably, have the socializing, norm-validating effects
that Durkheim attributed to the criminal law. Those processes are natural, unself-conscious, and inexorable.

Page 11 of 22
 Punishment

Expressive laws and policies are more like judges’ efforts to deter crime through the legal threats punishments
express, a process about whose effectiveness Durkheim was skeptical. They are conscious efforts to harness the
criminal law to achieve something that politicians or public officials want.

Politicians and officials dealing with punishment issues may be moved by a wide range of considerations. At their
most benign, officials may be genuinely concerned to prevent crime, to calm an upset public, to reassure fearful
people, and to enhance citizens’ confidence in the state. At their most venal, officials may cynically manipulate
public fears and anxieties—exacerbating fears in order to be seen to be trying to assuage them—to gain public
support and win elections. No doubt in many cases motives are mixed.

What happens in criminal courts is powerfully shaped by political and managerial concerns. Political theory in its
grand sense examines classical (e.g., Locke, Hobbes) and modern (e.g., Rawls 1971) notions about the properties,
functions, and responsibilities of the state. Here I mean something much humbler: analysis of how personal and
institutional interests shape punishment systems, policies, and practices. Promotion and adoption of expressive
policies is an example.

Anyone who wants to understand punishment must try to understand how courts process cases.11 At least in the
United States, courts seldom operate as philosophers (p. 113) and legal theorists presuppose.12 The model of
punishment implicit in much normative writing is one in which defendants are charged with crimes that the
prosecutor believes they committed and believes he or she has the evidence to prove, in which judges or juries
consider the evidence and enter judgments of guilt when they believe the charges have been proven, and in
which punishments are imposed based on the offenses committed and proven. That is seldom what happens.

Three interacting sets of personal and institutional interests shape what happens. First, for the vast majority of
cases, resource constraints preclude the operation of the archetypal criminal process described in the preceding
paragraph. Neither prosecutors’ offices nor courts have sufficient resources to try every case. Only a tiny fraction
goes to trial, and juries are involved only in a comparative handful. However, every case must be dealt with
somehow. Methods must be devised for diverting many cases from the system altogether and persuading the vast
majority of the defendants who remain to plead guilty. A whole series of policies and processes are implicated:
prosecutorial charging standards; eligibility criteria for diversion to treatment, mediation, and other nonadjudicative
dispositions; prosecution resource allocation policies on priority offenses; prosecution plea negotiation standards;
eligibility criteria for judicial diversion programs; sentencing guidelines; and judicial authority to stay the entry of a
judgment or the imposition or execution of a punishment. Among other consequences of these processes are that
offenders are often not convicted of the crimes they committed. Many cases are resolved without convictions.
Among those offenders who are convicted, most plead guilty on the inducement that the charges against them will
be reduced in seriousness or number, or that an agreed sentence will be imposed.

Second, personal interests of judges and prosecutors influence their behavior. Elected chief prosecutors and
elected or politically selected judges may want to be re-elected or re-appointed, or to be elected or appointed to a
higher office. This at the very least creates risks that individuals’ political or career ambitions will influence how
cases are handled. In many jurisdictions, prosecutors run openly demagogic election campaigns meant to show
that they are tougher than their opponents; this almost inevitably affects how their offices handle at least some
categories of cases. Some assistant prosecutors also aspire to run for electoral office later or to be promoted within
their offices and must be concerned to keep senior prosecutors happy and not to do things that might lessen their
chances for electoral nomination or victory. Every elected prosecutor or judge must worry about cases in which a
seemingly “lenient” sentence is imposed in a case that has attracted or may attract media attention. And it would
be a foolhardy, dishonest, or self-deceiving judge or prosecutor who claimed that he or she always handles
emotional, high-visibility cases exactly as any other case is handled.

Third, courts are complicated organizations and work most effectively when people get along and behave in ways
that one another regard as reasonable. Research on “courtroom work groups” and “local legal cultures” shows
that judges, prosecutors, (p. 114) defense lawyers, and probation officers often establish informal understandings
about reasonable performance of their respective roles. These understandings, of course, vary over space and
time, but in general they acknowledge that resources and time are limited and that most cases must be disposed of
by mutual agreement. Courtroom practitioners learn that, to a considerable degree, in order to get along they must
go along with existing ways of doing business including styles of plea negotiation and going rates for sentences.

Page 12 of 22
 Punishment

They also learn that mundane things matter: judges and other lawyers do not want hearings or trials to take too
long, especially routine ones; they want to go to lunch on time, to go home early on Friday, and not to be required
to do busywork caused by someone else’s “unreasonable” behavior. At least in places where people are assigned
for extended periods to work in particular courtrooms, people newly assigned are quickly socialized into work-
group norms and expectations. Any actor is always free to defy existing norms, but only at the risk of being seen
by others as unreasonable and undependable. There are also retaliatory risks that others will deny benefit of
customary accommodations to the uncooperative lawyer and the benefit of plea negotiation conventions and
sentencing going rates to his or her clients.

Considerations like those just described partly explain why many criminal laws and punishments are not applied as
their proponents wished or intended or as the laws’ plainest, most literal meanings might imply. Laws are not self-
executing. Human beings must apply them. Whether and how they do that depends on whether they believe a
particular law is reasonable, sensible, and just; whether punishments authorized or required are appropriate and
consistent with prevailing local norms; and whether the law can be applied without unacceptable disruption to
existing ways of doing business. Many expressive and other laws mandating severe penalties are routinely
circumvented for these reasons. Even when such laws are applied to some cases, courtroom work groups
generally devise ways to do so without disrupting court operations or mobilizing additional resources. Evaluators
commonly report, for example, that offenders convicted of salient offenses receive harsher penalties, fewer people
than before are convicted of those offenses (the others are convicted of less serious offenses not covered by the
enhanced penalty), and overall, sometimes after a short period of perturbation, guilty plea and trial rates revert to
traditional levels (Blumstein et al. 1983, chap. 3).

Thus ends this travelogue of institutional and disciplinary frameworks within which people make serious efforts to
understand punishment. More could be visited. Psychologists, for example, offer instruction into ways human
beings process information, including how we attribute characteristics to others, including offenders whom we may
arrest, prosecute, or sentence (or in each case not), but those processes operate in many social settings.
Evolutionary psychologists investigate hypotheses that human beings have been naturally selected for punitive
responses to wrongdoing and that this shapes how we respond to criminals and think about punishment.
Anthropologists can instruct on how different societies in different places deal with offenders, and historians can
instruct on different times.

(p. 115) IV. Thinking about Punishment

Thinking about punishment needs to take account of the subjects discussed in this chapter separately and
together. We can think only of one thing at a time. It is always necessary artificially to isolate aspects of a subject
one by one if each is to be given its due. Philosophers’ ideal theories of punishment attend to norms and values
and cannot fairly be faulted for failing to integrate Foucauldian insights. Social theorists pondering punishment’s
functions should not be faulted for failing to examine normative analyses that may seem epiphenomenal to the
operation of deeper social and economic forces. Political scientists studying courtroom work groups use a
combination of qualitative and quantitative methods to study courts as complex systems and to study aggregate
flows of cases through them. Normative analyses of dispositions of individual cases are not central to their subject.

One might argue, and some people do, that enrichment of knowledge for its own sake is a worthwhile activity. That
may be so, but few philosophers, legal theorists, social theorists, or political scientists would describe their
analyses and conclusions about punishment in that way. The imposition of pains or burdens on individuals
because they are believed to have violated important behavioral norms is a rich, complex, human activity.
Scholars of every stripe want to understand it better, usually with some thought that their efforts may contribute to
doing less of it or doing it in ways that are more self-aware, just, or effective. Philosophers may not believe their
ideal theories are realizable but may hope that their illumination of problems nudges practices and policies in more
morally justifiable directions. Social theorists may not believe that their insights will lead to the dissolution or radical
reconstitution of social institutions, but may hope that their analyses make people who exercise power more
modest and self-aware. Political scientists may not expect their findings to lead to fundamental reorganization of
court processes, but may hope that policies and practices will be improved because of their work.

That we can understand why the literatures about punishment are separate and distinct, and largely unaware of

Page 13 of 22
 Punishment

one another, does not mean that we can adequately understand punishment without taking account of all of them.
The overriding rationale of this chapter is that we need to see as much of the elephant as we can if we are to hope
to understand or control it. Without better questions, we are unlikely to settle on better answers. Here are a few:

1. How can severe expressive punishment policies be justified?

There may be ways, but reconciliation with any of the mainstream philosophies or legal theories of punishment will
be hard. Consider California’s three-strikes law requiring minimum twenty-five-year sentences for any third felony
conviction including, famously, thefts of pizza slices on a playground and of four CDs from K-Mart. Every
retributive, desert, and limiting retributivist theory would lead to condemnation of such (p. 116) penalties on
proportionality grounds: the law mandates punishments for comparatively minor offenses that are much more
severe than those imposed for many robberies, rapes, and homicides. Consequentialist theories would condemn
such punishments for imposing pains that cannot be justified by larger offsetting gains and thereby violate
parsimony constraints. Durkheimian analysts would point out that such penalties are no more likely than deterrent
penalties to influence behavior and thereby reduce the incidence of crime, but to the contrary may undermine
basic social norms through the exemplary imposition of punishments that are grossly disproportionate to the
seriousness of the crimes that triggered them. Socialization is a gradual accretive process not a sudden
pulverizing one; moral confusion and ambiguity results when punishments do not respect conventional ideas about
what is more serious than what. Wacquant would point out that most who receive such extreme penalties are
members of minority groups and ask whether we should not be deeply troubled by that. Rusche and Kirchheimer
would point out that nearly all the people receiving such punishments are outside the regular labor force. Social
scientists who study courts would point out that such laws often are circumvented by practitioners and always will
be inconsistently applied.

The law-and-economics scholar Richard Posner (1977), assuming that deterrence is a—or the—primary purpose of
sentencing, once argued that only a small number of guilty offenders should be imprisoned and that they should
receive very long sentences. His starting point was the notion that the risk of punishment is the price an offender
thinking of committing a crime must be prepared to pay. The threatened punishment “costs” must be such as to
outweigh the expected gains. He then assumed that punishing one hundred offenders with one-year prison
sentences would provide the optimal punishment costs for a particular kind of crime. He observed, however, that
sentencing ten offenders to ten-year prison terms each would also provide the optimal punishment costs, but at
less expense and with less human suffering. Either approach would impose the same number of years of
imprisonment. Economists estimate the punishment costs of crime by calculating the statistical probability that an
offense will result in a conviction and punishment, and then calculating the average punishment imposed per
crime. It is important to note that the denominator in the punishment calculation is the number of crimes committed,
not the number of convictions or punishments. The cost thus is the average amount of punishment expected to be
incurred for each offense committed in the community.

What matters is not the number of offenders punished but the total amount of punishment imposed. By that
calculus, ten prison terms of ten years each involves the same aggregate amount of punishment as one hundred
terms of one year. Posner argues that the ten-year sentences would be preferable because they would cost less to
administer. And, because people discount the future—the sixth year of imprisonment is not as painful as the first—
the ten would experience less aggregate subjective suffering than would the hundred.

That’s one argument, based on the economic model of amoral people who make decisions entirely on the basis of
self-interested calculation, that could conceivably (p. 117) be stretched to justify California’s three-strikes law. It
is, however, based on assumptions of punishment’s deterrent effects that Durkheim rejected a priori and that most
social scientists (though not most economists) reject as unsound (e.g., Doob and Webster 2003; Pratt et al. 2006;
Tonry 2008; cf. Levitt 2002). It gains its shock value from the way it contravenes the widely shared intuition that
like offenders should be treated alike and for that reason, even were such a policy adopted, it would be unlikely
often to be applied.

2. How can policies and practices that disproportionately affect members of disadvantaged groups
be justified?
People who have not thought much about punishment may find this question glib in the extreme and observe that

Page 14 of 22
 Punishment

the criminal law applies to all people of all incomes and social classes: disproportionate numbers of poor people
commit thefts, steal cars, and sell drugs, and accordingly disproportionate numbers are punished. The law in its
majestic impartiality, Anatole France observed, equally forbids the rich and the poor to sleep beneath the bridges
of Paris. That answer begs the questions of why those crimes are more likely to be prosecuted (compared with, for
example, equally prevalent tax evasion and domestic violence) and to result in severe punishments. It also begs
the questions as to why poorer peoples’ deviance is more likely to be criminalized than richer peoples’ deviance,
or why alcohol is a licit substance and the less socially damaging marijuana, heroin, and cocaine are not.

Neither philosophers nor legal theorists are much help here either. As a practical matter, they ignore the problem.13
Both discuss just punishments in individual cases (when they discuss individual cases at all) as they appear for
sentencing in court following a conviction. Whether cases appear in court, however, as political scientists have
shown, and for what offenses, is the outcome of a series of decisions shaped by influences that lie largely outside
the reach of normative punishment theories.

Functionalist analyses could help explain why criminal law seems mostly directed at poorer peoples’ deviance. It is
relatively easy to imagine Marx’s, Foucault’s, and Wacquant’s arguments elaborated in ways that do this. The
usefulness of Durkheim’s arguments is less clear: whether the dramaturgical effects of punishment of members of
distinctly disadvantaged social classes is likely effectively to model good and bad behavior is at best an empirical
question with no obvious answer.

3. Should normative theories of punishment take into account the findings of empirical studies of the
operation of courts?
How might that be done? This would be hard, and almost no one has tried to do it.

Within a state, plea bargaining differences between counties (charge bargaining in some, sentence bargaining in
others; bargaining over the number of counts in some, and over the most serious charge in others) mean that the
people being sentenced for a particular crime, or set of crimes, in one county, may have engaged in very different
(p. 118) behaviors than people being sentenced for the same crimes in a different county. Similar issues arise
every day in most court systems when some defendants charged with robbery are allowed to plead guilty to theft
and others are convicted of robbery. Yet sentencing guidelines systems,14 and most philosophers and legal
theorists, treat the offense of conviction as the measure of an offender’s wrongdoing.

There is something awkward and, once brought into the light of day, unedifying about theories that elegantly argue
for treating like offenders in like ways and different offenders differently, without taking into account the empirical
reality that the offenses of which people are convicted and for which they are punished stand in at best uncertain
relation to the crimes they are believed to have committed. In earlier times, during the heyday of indeterminate and
individualized sentencing, punishing people in ways that were disproportionate to the seriousness of their crimes
was not widely seen as a serious problem. Consequentialist strategies of deterrence, incapacitation, and
rehabilitation could be invoked to justify widely divergent punishments suffered by like-situated offenders.

4. Should normative theories of punishment take into account the insights of social theorists such as
Durkheim, Marx, Foucault, and Wacquant, and how might that be done?
Their point is that things are not what they appear to be, but are really something else. The drawing, quartering,
and burning of a traitor is not a loathsome punishment but a validation of the king’s power. The penitentiary is not,
as some early reformers saw it, a place for penitence and revelation or, as some Progressive reformers saw it, a
place to rehabilitate offenders into law-abiding ways, but an institution for converting pre–Industrial Revolution
human beings into the shapes required for them to become factory workers, soldiers, and government
bureaucrats. Contemporary crime control policies and the “experiment” with mass incarceration are not good-faith
efforts to suppress crime and reassure citizens but devices for destabilizing and disempowering black
communities. Or maybe they are devices for helping politicians get re-elected or for maintaining the legitimacy of
government in citizens’ eyes.

To some extent functionalist analyses are paradoxical. They are hypotheses that things are not as they seem.
Many people, to the contrary, will say that things are exactly as they seem. Some will simply reject the hypothesis
that contemporary American crime control policies operate to maintain existing patterns of racial hierarchy. A

Page 15 of 22
 Punishment

social theorist might well cite that denial as a form of false consciousness, as evidence in support of the
hypothesis. Existing policies do, as an empirical matter, demonstrably disadvantage minority groups, and lessen
the life chances of many of their members. People’s inability to recognize that, or to be troubled by its implications,
can be argued to show that the system is performing its hierarchy maintenance function: blacks are disadvantaged
as the outcome of policy processes that whites control and honestly claim are not meant to be discriminatory. The
politically (p. 119) dominant white majority has the satisfaction of believing it has acted in a principled,
nondiscriminatory way while benefiting from an outcome that sustains its economic, social, and political
advantages.

It is not, however, impossible to gain policy insights from functionalist analyses. The relationship between
Durkheimian ideas about norm reinforcement and Scandinavian theories about the moral-educative effects of
punishment provides an illustration. Some existing practices—for example, shortening prison terms of aboriginal
prisoners in Canada and Australia because of their short life expectancies—can be described as efforts to address
the implications of functionalist analyses of the law as a device for maintaining existing patterns of racial
(Wacquant) or class (Marx) relations.

5. Should race, ethnicity, or nationality matter?

This, I think, is the hardest question of them all. What should a judge do who simultaneously subscribes to a
principled theory of punishment and yet believes, with Wacquant, that the American history of race relations
largely explains the 100-to-1 rule, the War on Drugs, and disproportionate black incarceration, or with Marxist
analysts, that the explanations can be found in class-based functional analyses of the criminal law. The
disproportionate black presence among American prison inmates mirrors comparably disproportionate presences
of Moroccans in Dutch prisons, Algerians in French prisons, Turks in German ones, and foreigners in Swiss ones
(Tonry 1997).

Insofar as minority disparities can be attributed to social and economic characteristics associated with elevated
rates of criminality, minority disparities might be interpreted as another manifestation of a general social pattern
that links disadvantage to crime and punishment. Higher rates of crime and imprisonment can be expected to
decline as successive generations of immigrant groups become more fully assimilated. However, in some
countries, notably those with aboriginal populations (e.g., Australia, Canada, New Zealand) and the United States in
relation to black Americans, elevated levels of crime and imprisonment have endured for many generations. How, if
at all, can the criminal law and punishment doctrine take account of that?

Related issues have arisen in other places. In countries with large aboriginal populations, should normal prison
sentence lengths be discounted to take account of aboriginal citizens’ shorter life expectancies? Appellate courts
in Australia and Canada have explicitly discussed the question. If there are jurisprudential bases for such a
discount, the same argument might apply in many countries to members of ethnic minority groups afflicted by
significantly lower life expectancies than the majority population—for example, to blacks in the United States.

Should immigrants or visitors from countries with especially harsh legal systems be punished by different standards
than a country’s majority population? The issue (p. 120) usually arises in places like Germany or the
Scandinavian countries with relatively mild punishment traditions. Should, for example, Turks convicted of drug
trafficking be sentenced by Finnish standards (a few months or a year; seldom for many years) or according to the
much more severe standards of Turkish courts and prisons (Andenaes 1974)? Concern for equality before the law
leads many people to dismiss such a proposal out of hand. Its rejection, however, creates a perverse incentive for
would-be offenders from harsher places to come to Scandinavia, for example, to market their drugs.

None of these questions has easy answers. There are no easy ways to integrate the insights about punishment of
philosophers, legal theorists, social theorists, and political scientists. We will understand punishment better,
however, if we try to answer such questions and to integrate diverse insights.

References
Allen, Francis A. 1959. “Legal Values and the Rehabilitative Ideal.” Journal of Criminal Law, Criminology, and
Police Science 50:226–32.

Page 16 of 22
 Punishment

Allen, Francis A. 1964. The Borderland of Criminal Justice: Essays in Law and Criminology. Chicago: University of
Chicago Press.

American Friends Service Committee. 1971. Struggle for Justice: A Report on Crime and Punishment in America.
New York: Hill and Wang.

American Law Institute. 1962. Model Penal Code (Proposed Official Draft). Philadelphia: American Law Institute.

American Law Institute. 2007. Model Penal Code: Sentencing. Tentative Draft no. 1 (April 9, 2007). Philadelphia:
American Law Institute.

Andenaes, Johannes. 1974. Punishment and Deterrence. Ann Arbor: University of Michigan Press.

Australia Law Reform Commission. 1980. Sentencing of Federal Offenders. Canberra: Australian Government
Publishing Service.

Bentham, Jeremy. 1970. “The Utilitarian Theory of Punishment.” In An Introduction to Principles of Morals and
Legislation, ed. Jeremy Bentham, J. H. Burns and H. L. A. Hart. London: Methuen.

Bianchi, Herman, and René van Swaaningen. 1986. Abolitionism: Towards a Non-Repressive Approach to Crime.
Amsterdam: Free University Press.

Blumstein, Alfred, Jacqueline Cohen, Susan Martin, and Michael Tonry, eds. 1983. Sentencing Research and the
Search for Reform. Report of the National Academy of Sciences Panel on Sentencing Research. Washington, DC:
National Academies Press.

Braithwaite, John. 2001. Restorative Justice and Responsive Regulation. New York: Oxford University Press.

Braithwaite, John, and Philip Pettit. 1990. Not Just Deserts: A Republican Theory of Criminal Justice. New York:
Oxford University Press.

Bureau of Justice Statistics. 2003. “Prevalence of Imprisonment in the U.S. Population, 1974–2001.” Washington,
DC: Bureau of Justice Statistics.

Burgess, Anthony. 1962. A Clockwork Orange. London: Heinemann.

Burgess, Anthony. 1986. “Introduction: A Clockwork Orange Resucked.” A Clockwork Orange. New York: W. W.
Norton.

Canadian Sentencing Commission. 1987. Sentencing Reform: A Canadian Approach. Ottawa: Canadian
Government Publishing Centre.

Clear, Todd R., and David Karp. 1999. Community Justice Ideal. Denver: Westview Press.

Cover, Robert. 1984. Justice Accused: Antislavery and the Judicial Process. New Haven: Yale University Press.

(p. 123) Davis, Kenneth Culp. 1969. Discretionary Justice: A Preliminary Inquiry. Baton Rouge: Louisiana State
University Press.

Dershowitz, Alan. 1976. Fair and Certain Punishment. New York: McGraw-Hill.

Doob, Anthony, and Cheryl Webster. 2003. “Sentence Severity and Crime: Accepting the Null Hypothesis.” In
Crime and Justice: A Review of Research, vol. 30, edited by Michael Tonry. Chicago: University of Chicago Press.

Duff, Antony. 1986. Trials and Punishments. Cambridge: Cambridge University Press.

Duff, Antony. 2001. Punishment, Communication, and Community. New York: Oxford University Press.

Durkheim, Emile. 1933 [1893]. The Division of Labor in Society. Trans. George Simpson. New York: Macmillan.

Eisenstein, James, Roy Flemming, and Peter Nardulli. 1988. The Contours of Justice: Communities and Their Courts.
Boston: Little, Brown.

Page 17 of 22
 Punishment

Foucault, Michel. 1979. Discipline and Punish: The Birth of the Prison. London: Penguin.

Frankel, Marvin. 1972. Criminal Sentences—Law without Order. New York: Hill and Wang.

Garland, David. 2001. The Culture of Complaint. Chicago: University of Chicago Press.

Hampton, Jean. 1984. “The Moral Education Theory of Punishment.” Philosophy and Public Affairs 13(3):208–38.

Hart, Henry M. 1958. “The Aims of the Criminal Law.” Law and Contemporary Problems 23:401–42.

Hart, H. L. A. 1968. Punishment and Responsibility. Oxford: Clarendon.

Home Office. 1990. Crime, Justice, and Protecting the Public. Cm 965. London: Her Majesty’s Stationery Office.

Home Office. 2001. The Halliday Report: “Making Punishments Work: A Review of the Sentencing Framework for
England and Wales.” http://www.homeoffice.gov.uk/documents/halliday-report-sppu/.

Honderich, Ted. 2006. Punishment: The Supposed Justifications Revisited. London: Pluto.

Jacob, Herbert, and James Eisenstein. 1977. Felony Justice: An Organizational Analysis of Criminal Courts. Boston:
Little, Brown.

Kant, Immanuel. 1965 [1787]. “The Penal Law and the Law of Pardon.” The Metaphysical Elements of Justice,
trans. John Ladd. Indianapolis: Liberal Arts Press/Bobbs-Merrill.

Kleinig, John. 1973. Punishment and Desert. Dordrecht, Netherlands: Martinus Nijhoff.

Lacey, Nicola. 1988. State Punishment: Political Principles and Community Values. New York: Routledge.

Lappi-Seppälä. 2001. “Sentencing and Punishment in Finland: The Decline of the Repressive Ideal.” In Sentencing
and Sanctions in Western Countries, ed. Tonry and Richard S. Frase. New York: Oxford University Press.

Levitt, Steve. 2002. “Deterrence.” In Crime: Public Policies for Crime Control, ed. James Q. Wilson and Joan
Petersilia. Oakland, CA: Institute for Contemporary Studies Press.

Lewis, C. S. 1949. “The Humanitarian Theory of Punishment.” 20th Century: An Australian Quarterly Review.
3(3)3:5–12 (repr. in God in the Dock: Essays on Theology and Ethics, ed. Walter Hooper. Grand Rapids, MI:
Eerdmans).

Loury, Glenn C. 2007. “Racial Stigma, Mass Incarceration and American Values.” The Tanner Lectures in Human
Values delivered at Stanford University on April 4 and 5, 2007.
http://www.econ.brown.edu/fac/Glenn_Loury/louryhomepage/.

Martinson, Robert. 1974. “What Works? Questions and Answers about Prison Reform.” The Public Interest
35(2):22–54.

(p. 124) Marx, Karl. 1853. “Capital Punishment—Mr. Cobden’s Pamphlet’s—Regulations of the Bank of England.”
New York Tribune, February 17, p. 3.

Massey, Douglas S. 2007. Categorically Unequal. New York: Russell Sage Foundation.

Massey, Douglas S., and Nancy Denton. 1993. American Apartheid: Segregation and the Making of the
Underclass. Cambridge, MA: Harvard University Press.

McCloskey, H. J. 1968. “A Non-utilitarian Approach to Punishment.” In Contemporary Utilitarianism, ed. Michael D.

Bayles. New York: Doubleday.

McDonald, Douglas, C., and Kenneth E. Carlson. 1993. Sentencing in the Federal Courts: Does Race Matter?
Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics.

Michael, Jerome, and Mortimer Adler. 1933. Crime, Law, and Social Science. New York: Harcourt Brace.

Page 18 of 22
 Punishment

Michael, Jerome, and Herbert Wechsler. 1937. “A Rationale of the Law of Homicide.” Columbia Law Review
37:701–61(pt. 1), 1261–1335 (pt. 2).

Minnesota Sentencing Guidelines Commission. 1980. Report to the Legislature. St. Paul: Minnesota Sentencing
Guidelines Commission.

Moore, Michael S. 1993. “Justifying Retributivism.” Israeli Law Review 27:15–36.

Morris, Herbert. 1966. “Persons and Punishment.” Monist 52:475–501.

Morris, Herbert. 1981. “A Paternalist Theory of Punishment.” American Philosophical Quarterly 18:263–71.

Morris, Norval. 1974. The Future of Imprisonment. Chicago: University of Chicago Press.

Morris, Norval, and Michael Tonry. 1978. “Sentencing Reform in America.” In Reshaping the Criminal Law—A
Festschrift in Honor of Glanville Williams, ed. Peter Glazebrook. London: Stevens.

Morris, Norval, and Michael Tonry. 1990. Between Prison and Probation: Intermediate Punishments in Rational
Sentencing System. New York: Oxford University Press.

Murphy, Jeffrie. 1973. “Marxism and Retribution.” Philosophy and Public Affairs 2:217–43.

Murphy, Jeffrie G., and Jean Hampton. 1988. Forgiveness and Mercy. Cambridge: Cambridge University Press.

Nardulli, Peter, James Eisenstein, and Ron Flemming. 1988. The Tenor of Justice: Criminal Courts and the Guilty
Plea Process. Champaign, IL: University of Illinois Press.

Pincoffs, Edmund. 1966. The Rationale of Legal Punishment. New York: Humanities Press.

Posner, Richard. 1977. Economic Analyses of Law. 2nd ed. Boston: Little, Brown.

Pratt, Travis C., Francis T. Cullen, Kristie R. Blevins, Leah H. Daigle, and Tamara D. Madensen. 2006. “The Empirical
Status of Deterrence Theory: A Meta-Analysis.” In Taking Stock: The Status of Criminological Theory, ed. Francis
T. Cullen, John Paul Wright, and Kristie R. Blevins. New Brunswick, NJ: Transaction.

Rawls, John. 1955. “Two Concepts of Rules.” Philosophical Review 64:3–32.

Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press.

Rothman, David J. 1971. The Discovery of the Asylum: Social Order and Disorder in the New Republic. Boston:
Little, Brown.

Rothman, David J. 1980. Conscience and Convenience. Boston: Little, Brown.

Rusche, Georg, and Otto Kirchheimer 1968 [1939]. Punishment and Social Structure. New York: Russell and
Russell. New York: Columbia University Press.)

Singer, Richard. 1979. Just Deserts: Sentencing Based on Equality and Desert. Cambridge, MA: Ballinger.

Tonry, Michael. 1997. “Ethnicity, Crime and Immigration.” In Ethnicity, Crime, and Immigration: Comparative and
Cross-national Perspectives, edited by Michael Tonry. (p. 125) Vol. 21 of Crime and Justice: A Review of
Research, edited by Michael Tonry. Chicago: University of Chicago Press.

Tonry, Michael. 2004. Thinking about Crime: Sense and Sensibility in American Penal Culture. New York: Oxford
University Press.

Tonry, Michael. 2006. “Purposes and Functions of Sentencing.” In Crime and Justice: A Review of Research, vol.
34, edited by Michael Tonry. Chicago: University of Chicago Press.

Tonry, Michael. 2008. “Learning from the Limitations of Deterrence Research.” In Crime and Justice: A Review of
Research, vol. 37, edited by Michael Tonry. Chicago: University of Chicago Press.

Page 19 of 22
 Punishment

Tonry, Michael. 2011. Punishing Race: An American Dilemma Continues. New York: Oxford University Press.

Tonry, Michael, and Matthew Melewski. 2008. “The Malign Effects of Drug and Crime Control Policies on Black
Americans.” In Crime and Justice: A Review of Research, vol. 37, edited by Michael Tonry. Chicago: University of
Chicago Press.

U.S. Sentencing Commission. 1995. 1995 Special Report to the Congress: Cocaine and Federal Sentencing Policy.
Washington, DC: U.S. Sentencing Commission.

U.S. Sentencing Commission. 2007. Cocaine and Federal Sentencing Policy. Washington, DC: U.S. Sentencing
Commission.

von Hirsch, Andrew. 1976. Doing Justice: The Choice of Punishments. New York: Hill and Wang.

von Hirsch, Andrew. 1985. Past or Future Crimes: Dangerousness and Deservedness in the Sentencing of
Criminals. New Brunswick, NJ: Rutgers University Press.

von Hirsch, Andrew. 1993. Censure and Sanctions. Oxford: Clarendon.

von Hirsch, Andrew. and Andrew Ashworth. 2005. Proportionate Sentencing: Exploring the Principles. Oxford:
Oxford University Press.

Wacquant, Loïc. 2002a. “From Slavery to Mass Incarceration.” New Left Review 13 (January-February): 41–60.

Wacquant, Loïc. 2002b. “Deadly Symbiosis: Rethinking Race and Imprisonment in Twenty-First-Century America.”
Boston Review 27(2):22–31.

Western, Bruce. 2006. Punishment and Inequality in America. New York: Russell Sage Foundation.

Wexler, David B., and Bruce Winick, eds. 2003. Justice in a Therapeutic Key. Durham, NC: Carolina Academic
Press.

Notes:

(1) . In the 1970s and 1980s, “abolitionists” proposed the abolition of punishment, but what they meant in
substance was major diminution in resort to the criminal justice system and development of other mechanisms
operating under other names (for example, mental health or social welfare) for dealing with egregious offenses or
dangerous people (Bianchi and van Swaaningen 1986). More recently, many people have urged a shift away from
criminal justice toward restorative justice approaches, but almost invariably they allow for criminal justice handling
of egregious cases and incorrigible offenders (e.g., Braithwaite 2001).

(2) . Some people who argue that corporal punishments should never be imposed, and U.S. courts agree (e.g.,
Jackson v. Bishop, 404 F.2d 571 [8th Cir., 1968]) but that is not an argument about punishment per se. It is a human
rights argument akin to the argument that torture should never be used to obtain evidence from recalcitrant
suspects or witnesses. In much of the developed world, capital punishment is not an authorized punishment on the
basis of human rights arguments. The constitutional distinction between death and lesser forms of corporal
punishment as authorized punishments is distinctively American.

(3) . Other modern consequentialist theories would weigh the effects of alternative choices on “dominion”
(individuals’ capacities and opportunities to live satisfying lives within organized communities; Braithwaite and Pettit
1990) or on the community (Lacey 1988).

(4) . H. L. A. Hart (1968, chap. 1) disparaged this as an illegitimate “definitional stop”: an effort to evade the
problem by defining it away.

(5) . The novelist and Christian apologist C. S. Lewis (1949) and the novelist Anthony Burgess (1962, 1986) earlier
offered moral critiques of consequentialism, especially in its rehabilitative aspirations.

(6) . He had in mind strict liability crimes. Some like traffic offenses are administrative offenses that threaten only

Page 20 of 22
 Punishment

financial penalties (and sometimes loss of a driver’s license). Others are business regulatory offenses (such as
violation of pure food and drug laws) in which the stakes are so high and proof of fault is so difficult that
corporations and sometimes their executives are made strictly liable for their violation. The aim is to create
powerful incentives to assure compliance with the law. Prison sentences are rarely if ever imposed for such crimes.

(7) . Utilitarian ideas were influential in all developed countries. Most countries developed treatment programs and
institutions for offenders, and many developed formal systems of parole and probation. Retributive ideas however
also remained influential in most countries. Only in the United States were the institutions of the criminal justice
system radically transformed to reflect utilitarian ideas. Examples include the “inventions” of the reformatory, the
juvenile court, probation, parole, and the indeterminate sentence. Other countries later adopted some of the
American innovations but usually in less full-blown forms. None, for example, removed authority to set lengths of
prison sentences from judges and gave it to parole boards.

(8) . In addition, between the mid-1970s and mid-1980s a number of states starting with California in 1976 enacted
systems of “statutory determinate sentencing” that in fairly crude ways enacted statutory provisions which
prescribed punishments for particular offenses. Statutory systems were quickly recognized to be less effective
than sentencing guidelines at addressing the perceived weaknesses of indeterminate sentencing. After 1985 no
more statutory systems were enacted (Blumstein et al. 1983).

(9) . Stated more fully: it is assumed that citizens, if rational, would, if given the opportunity, assent to be governed
by a set of rules that forbids serious intrusions by others into their lives and by them into the lives of others.
Acceptance of such a body of rules would allow each citizen to plan and conduct his life or her life in the way he
or she wishes in the expectation that the rules will be observed by all. Accordingly, if those rules are violated, the
violator has benefited from others’ obedience to the rules has thereby obtained an unfair advantage.

(10) . A narrower example is provided by laws forbidding sex offenders to live within three miles of a school, a
distance that in many towns effectively forbids any sex offender to live there. Other expressive laws mandate
especially harsh punishments for drug sales within specified distances of a school; the distances often effectively
mandate the harsher punishment for any drug sale in the town.

(11) . I discuss the issues raised in this section and provide references to the major literature, in an earlier article
(Tonry 2006). The classic works by political scientists on the operation of criminal courts are Jacob and Eisenstein
(1977), Eisenstein, Flemming, Nardulli (1988), and Nardulli, Eisenstein, and Flemming (1988).

(12) . The issues raised in this section though they probably arise in weaker forms in every court system in their
strong forms are distinctively American. Punishment questions are universal even if their answers are inevitably
local. With small and partial exceptions (most notably in Switzerland), however, in no developed country other than
the United States are judges and prosecutors elected or openly selected on the basis of political criteria. And in no
other country are most criminal cases and many sentences resolved by means of plea negotiations.

(13) . That’s not quite true. It is more accurate to say that sometimes they recognize the problem (e.g., von Hirsch
1976, last chap.) and then step around it, and other times they acknowledge it as a reason why their ideal theories
are unsuitable for implementation in the real world (Duff 1986).

(14) . The exception is the U.S. Sentencing Commission’s “real offense” policy under which convicted offenders
were sentenced for the “actual offense behavior” found by the judge to have occurred using a civil law burden of
proof, irrespective of whether charges were not filed for part of that behavior, or been dismissed, or resulted in a
not guilty verdict.

Michael Tonry
Michael Tonry is Russell M. and Elizabeth M. Bennett Chair in Excellence and Director, Robina Institute of Criminal Law and
Criminal Justice, University of Minnesota Law School.

Page 21 of 22