Thinking About Crime

Thinking About Crime:
Sense and Sensibility in

American Penal Culture
Michael Tonry
OXFORD UNIVERSITY PRESS

Thinking about Crime
Selected and Recent Titles in
STUDIES IN CRIME AND PUBLIC POLICY
Michael Tonry and Norval Morris, General Editors
The American Street Gang:

Its Nature, Prevalence, and Control
Malcolm W. Klein
Restorative Justice and Responsive Regulation
John Braithwaite
Maconochie’s Gentlemen: The Story of Norfolk Island
and the Roots of Modern Prison Reform
Norval Morris
Can Gun Control Work?
James B. Jacobs
Penal Populism and Public Opinion:
Lessons from Five Countries
Julian V. Roberts, Loretta J. Stalans, David Indermaur,
and Mike Hough
Mafia Brotherhoods: Organized Crime, Italian Style
Letizia Paoli
When Prisoners Come Home: Parole and Prisoner Reentry
Joan Petersilia
The Contradictions of American Capital Punishment
Franklin E. Zimring
Thinking about Crime: Sense and Sensibility
in American Penal Culture
Michael Tonry
Thinking about CRIME
sense and sensibility
in American penal culture
Michael Tonry
1
2004
3 Oxford New York
Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
São Paulo Shanghai Taipei Tokyo Toronto
Copyright © 2004 by Oxford University Press, Inc.
Published by Oxford University Press, Inc.,
198 Madison Avenue, New York, New York 10016
www.oup.com
Oxford is a registered trademark of Oxford University Press
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior permission of Oxford University Press.
Library of Congress Cataloging-in-Publication Data
Tonry, Michael H.
Thinking about crime : sense and sensibility in American penal culture
/ by Michael Tonry.
p. cm. — (Studies in crime and public policy)
Includes bibliographical references and index.
ISBN 0-19-514101-6
1. Prisons—Government policy—United States. 2. Crime
prevention—Government policy—United States. 3. Discrimination in
criminal justice administration—United States. 4. Prisons in mass
media—United States. 5. Prisons—United States—Public opinion.
6. Public opinion—United States. 7. United States—Politics and
government—2001– I. Title. II. Series.
HV9471.T65 2004
365′.973—dc21 2003004245
1 2 3 4 5 6 7 8 9
Printed in the United States of America
on acid-free paper
Acknowledgments
Isaac Newton, not generally known as a modest man, once,

on being praised, accepted it as his due but also observed
that in reaching so high he had stood on the shoulders of
giants. There are not many Isaac Newtons, and I am certainly
not one, but it feels right to acknowledge people without
whose existence and works this slight book could not have
been written. I have tried to draw on the work of historians
and sociologists to understand why modern Americans think
as they do about crime and criminals, and how, understand-
ing that, we can do more that is good and less that is harmful.
This book could not have been written nor the ideas in it
developed had I not learnt from the writings of historians
John Boswell, Grant Gilmore, Paul Johnson, David Musto,
and Jaroslav Pelikan, and of sociologists Stanley Cohen, Émile
Durkheim, Norbert Elias, Kai Erikson, David Garland, Joseph
Gusfield, and Edwin Sutherland. With the exception of Gar-
land, none of these are friends. Most of them are dead. A
few may count as giants.
This book, whatever its demerits, would be even less had
friends not read and reacted to drafts. Among those who
tried to save me from error, not always, in their views, suc-
cessfully, are David Garland, Roxanne Lieb, Charles Loeffler,
Amanda Matravers, Norval Morris, Kevin Reitz, Julian Rob-
erts, Michael E. Smith, and Andrew von Hirsch. Sara Harrop,
Charles Loeffler, and Vickie Sheridan helped with references,
artwork, and avoidance of grosser infelicities of style. Dedi
X. Felman, my long-suffering and patient editor at Oxford
University Press, stayed the course, ran interference, and
did all the things good editors—in romantic stereotype but
seldom in life—do. I am lucky to have such good friends.
Earlier versions of bits of chapter 2 appeared in Crime and
Delinquency, of chapter 6 in the UCLA Law Review, and of
chapter 7 in Punishment and Society. They are reprinted by
permission.
vi ACKNOWLEDGMENTS
Preface
From capital punishment to three-strikes-and-you’re-out to

the highest imprisonment rates in the Western world by a
factor of five, the United States stands alone in what it does
to its citizens to prevent crimes and punish criminals.
There are good reasons to doubt that recent punishment
policies have had much to do with recent drops in crime.
The strongest is that crime trends for the past 40 years have
been broadly similar in every Western country, and in every
American state, while punishment policies and practices have
varied enormously.
The United States has a punishment system that no one
would knowingly have built from the ground up. It is often
unjust, it is unduly severe, it is wasteful, and it does enor-
mous damage to the lives of black Americans.
This book explains how contemporary American crime pol-
icies came to be as they are, and how they can be reconfigured
to be made more effective but less costly, and to do less
harm to offenders, their loved and loving ones, and their
communities. The story is partly about rising crime rates
and public fears, about cynical politics and pusillanimous
politicians, and about public opinion and sensationalizing
media. The story more importantly is about the ways long-
term trends in values and attitudes influence how Americans
think about crime and punishment, and how they think about
victims and offenders.
Social scientists use the word “sensibilities” to refer to pre-
vailing social values, attitudes, and beliefs. Sensibilities change
slowly over time and fundamentally shape what people think
and believe. Current crime control policies are to a large
part an outgrowth of American sensibilities of the past third
century. Nothing wrong with that, one might observe. Poli-
cies have expressed the views of the people, and that is what
democracy is all about.
There is something wrong with that, though, because we
know, looking backward, that the prevailing sensibilities of
an era often are deeply regretted later. Mainstream American
sensibilities in the past have supported policies—slavery,
the near-extermination of indigenous North American peo-
ples, the internment of Japanese-Americans, sterilization of
the mentally ill, ostracism of homosexuals—that we now
believe to be wrong in some timeless sense. Sensibilities of
particular times and places may have supported such atti-
tudes and practices, but few people today believe that justi-
fies them.
And so it is concerning crime and punishment. In some
respects this seems obvious. No one any longer calls for draw-
ing and quartering of offenders, for public execution by guil-
lotine, or for the use of torture to extract confessions. In
other respects, it may not appear so obvious, even though
practices that many Americans endorse—capital punish-
ment, three-strikes laws, prison sentences measured in de-
cades or lifetimes—are as unthinkable in other Western
countries as are lynchings and public torture in America.
Assessments of the acceptability of particular policies and
practices need to take account of basic human rights and
moral considerations. They also need to take account of what
historians can tell us about interactions between sensibilities,
behavior, and policies in the past.
We know from historical writings on recurring cycles of
support for romanticism and classicism in the arts, of toler-
ance and intolerance of homosexuality and religious plural-
ism, and of attitudes and policies on drug abuse, that human
sensibilities often move in regular cycles and that beliefs and
policies move with them.
viii PREFACE
In trying to understand and assess crime control and pun-
ishment, we need to take account of prevailing sensibilities,
of timeless human rights ideas, and of what we know about
past interactions between sensibilities, cycles of tolerance
and intolerance, and public policies. When we do that in
relation to drug policy, we learn that “wars on drugs” are
typically launched when drug use is beginning to decline,
that members of minority groups are typically targeted and
scapegoated, and that attitudes and policies typically soften
a decade or two later. A similar pattern characterizes “wars
on crime.”
Had American policy makers in the 1980s and 1990s heeded
the readily available historical literature, learned its lessons,
and more sensitively interpreted evidence of public concern,
the wars on drugs and crime would have taken very different
forms. Many fewer Americans would have been sent to
prison, and the criminal justice system would have intruded
into many fewer black Americans’ lives.
What was in effect a form of false consciousness led drug
and crime control policy makers to adopt policies that were
too harsh, too simple, and too wasteful. Once we understand
why American policies evolved as they did we can begin
the job of making them better.
That in a nutshell is the main argument. I develop it within
these eight chapters. Chapter 1 is a brief summary of the
whole. Chapter 2 asks, and then summarizes alternative an-
swers to, the question of why contemporary American crime
control patterns developed as they did. Many answers have
been offered, including rising crime rates, public opinion,
the politics of race, political cynicism, the weakened credibil-
ity of government, and post-modernist angst. None of these
things is irrelevant, but the best answer focuses on cycles
of intolerance and prevailing but fallible sensibilities. Chap-
ters 3 and 4 introduce basic ideas and literatures on cycles
of intolerance, sensibilities, and moral panics. Chapter 5 re-
turns to the main story line to show that American crime
trends closely resemble those of other countries, that crime
PREFACE ix
control policies are much different, and that prevailing sensi-
bilities explain both why we think our problems are worse
and our policies are better (and effective). To illustrate the
force of prevailing sensibilities, chapters 6 and 7 examine
contemporary policies that would have been unthinkable in
earlier times, and mainstream ideas from recent earlier times
that have been forgotten. The last chapter returns to the
premise that the United States now has a punishment system
that no one would knowingly have chosen, and shows how
it can be changed to do more good and less harm, especially
to members of American minority groups.
x PREFACE
Contents
1 Sense and Sensibility in American Penal Culture 3
2 Why So Many Americans Are in Prison 21
3 Cycles and Sensibilities 63
4 Moral Panics and “Windows of Opportunity” 85
5 Crime Trends and the Effects of Crime Control

Policies 97
6 Formerly Unthinkable Policies 141
7 Unthought Thoughts 169
8 Better Understanding, People, and Policies 195
References 229
Index 249
This page intentionally left blank
Thinking about Crime
1
Sense and Sensibility

in American Penal Culture
In searching for ways to address rising crime rates in the

final third of the twentieth century, American policy makers
got lost in a forest of good intentions, public anxieties, and
political cynicism. They created a punishment system that
no one would knowingly have chosen, but that we do not
know how to change. Current policies are too severe, waste
lives and money, and often produce unjust results. They
have produced an imprisonment rate five times higher than
that of any other Western country and seven to twelve times
higher than most.
Current policies have disproportionately damaged the lives
of black offenders and their families and have undermined
black communities. Nearly a third of young black men are
in prison or jail, or on probation or parole. More than 13
percent of black men between 25 and 29, more than one in
every eight, were in jail or prison in 2001. We lack compar-
able information about Hispanics, but they are the fastest
growing part of the prison population. Unless current policies
change, the criminal justice system will soon intrude as
deeply into Hispanic Americans’ lives and communities as
it now does into black Americans’ lives and communities.
Ordinary Americans made these things happen. Elected
politicians proposed policies and enacted laws, but they
would not have done it if they believed American voters would
disapprove. Journalists and academics assert that many politi-
cians who proposed tough policies did so cynically, out of
personal, political self-interest, and that many realized the
policies they proposed would not work and would do harm.
“Soft on crime” was a label few politicians dared risk during
3
the 1970s through the 1990s, and many supported policies
they believed unwise or unjust rather than risk losing an
election. An influential analysis by Katherine Beckett shows
that politicians typically led public opinion about crime
rather than followed it, in effect raising public fears and
anxieties and then proposing harsh, simplistic solutions to
ameliorate them. The bottom line, nonetheless, even if those
accounts are valid, is that the public could not have been
led someplace it was unwilling to go.
At the beginning of a new century, Americans are having
second thoughts about the wisdom of current antidrug and
anticrime policies, but elected politicians most places are
afraid to change them. To do so runs risks of being tarred
as soft on crime and, until significant numbers of politicians
take that risk and get reelected, most will not take the chance.
And until that happens, the unfortunate social facts described
in the opening paragraph will continue to be true.
There is lots of evidence that ordinary Americans want to
change current policies, and there is lots of evidence that
judges and prosecutors at local and state levels are respond-
ing. Policy makers in some states are responding, though
usually by tinkering at the margins rather than by overhaul-
ing current policies. Legislators and executive branch officials
in the federal government and in the large, heavily populated
states, however, are neither repealing nor fundamentally re-
casting failed policies. Most remain locked in the politics and
attitudes of earlier times and are hunkering down to protect
the status quo. The George W. Bush administration’s determi-
nation, reiterated in 2002, to preserve the federal laws that
punish crack cocaine offenses as severely as powder cocaine
offenses one hundred times larger is but one illustration. Most
crack is sold by young inner-city blacks and most cocaine by
older whites. As a result, the 100-to-1 law imposes much
longer sentences on blacks. Almost no one defends that out-
come on the merits, but political courage to change the law
cannot yet be found.
This book explains how contemporary policies came to be
as they are, and how they can be reconfigured to be more
4 THINKING ABOUT CRIME

effective but less costly, and to do less harm to offenders,
their loved and loving ones, and their communities. The story
is only partly about rising crime rates and public fears, about
cynical politics and pusillanimous politicians, and about pub-
lic opinion and sensationalizing media. The story is also about
the ways long-term trends in social values and public atti-
tudes, and short-term effects of moral panics, influence how
Americans think about crime and punishment. For two de-
cades, Americans thought they wanted single-minded tough-
ness and they got it. The question is why they thought they
wanted it.
“Moral panics” are part of the answer. They typically occur
when horrifying or notorious events galvanize public emo-
tion, and produce concern, sympathy, emotion, and overreac-
tion. Examples in recent years include the kidnapping of
Polly Klaas in California and the crack-overdose death of Len
Bias in Maryland. Results included, respectively, California’s
three-strikes law and the federal 100-to-1 crack cocaine sen-
tencing law. Sometimes moral panics are about general prob-
lems, for example, child abuse or school violence, and the
resulting emotions have been harnessed constructively. Of-
ten though, moral panics relating to crime lead to poorly con-
sidered and overly harsh reactions. In recent decades, moral
panics have magnified the effects of longer term changes in
values and attitudes.
The ways people think about contentious issues change
slowly but predictably. Social scientists use the word “sensi-
bilities” to refer to prevailing social values, attitudes, and
beliefs, and show how sensibilities change slowly over time
and shape and reshape what people think and believe. Cur-
rent American crime control policies are to a large part an
outgrowth of American sensibilities of the past third of the
twentieth century. Nothing wrong with that, one might re-
spond. Policies have expressed the views of the people, and
that is what democracy is all about.
There is something wrong with that, though. Looking
backwards, we know that American sensibilities of earlier
times supported policies—slavery, the near-extermination of
SENSE AND SENSIBILITY IN AMERICAN PENAL CULTURE 5
indigenous North American peoples, internment of Japanese
Americans during World War II, sterilization of the mentally
ill, ostracism of homosexuals, denial to women of a right to
vote—that we now believe to be wrong in some timeless
sense, and not merely in light of current sensibilities. Looking
outside the United States in our own and earlier times, similar
judgments are widespread about religious intolerance, op-
pression of women, and exploitation of child labor. Sensibili-
ties of particular times and places may support such attitudes
and practices, but few people in our time believe that justifies
them.
And so it is concerning crime and punishment. In some
respects this seems obvious. No one any longer calls for draw-
ing and quartering of offenders, for public execution by guil-
lotine, or for the use of torture to extract confessions. In
other respects, it may not appear so obvious, even though
practices that many Americans in our time endorse—capital
punishment, three-strikes laws, prison sentences measured
in decades or lifetimes—are as unthinkable in other Western
countries today as are lynchings and corporal punishment
in America.
Concluding that particular policies or practices are conso-
nant with current sensibilities is thus the beginning but can-
not be the end of assessments of their legitimacy. That evalua-
tion needs to take account of basic human rights and moral
considerations, whatever the public opinion poll results or
prevailing sentiments of a particular day or year. It also needs
to take account of what historians can tell us about interac-
tions between sensibilities, attitudes, and policies in the past.
We know from historical writings on recurring cycles of
support for romanticism and classicism in the arts, of toler-
ance and intolerance of homosexuality and religious plural-
ism, and of attitudes and policies on drug abuse, that human
sensibilities often move in regular cycles and that beliefs and
policies move with them. Persecution of “heretics” in the
history of the Christian church, for example, often coincided
with periods when the power and authority of the organized

church were in doubt. At other times, self-confident church
leaders comfortably tolerated dissenters.
In trying to understand and normatively evaluate crime
control policies we need to take account of prevailing sensi-
bilities, of “timeless” human rights ideas, and of what we
know about past interactions between sensibilities, cycles of
tolerance and intolerance, and public policies. When we do
that in relation to drug policy, we learn that “wars on drugs”
are typically launched when drug use has peaked and is
beginning to decline, that antidrug policies and law enforce-
ment practices become harshest during such times, and that
zealous drug law enforcement targets and scapegoats mem-
bers of minority groups. People violating drug laws are
treated much more harshly during these cyclical periods of
heightened intolerance than they would have been ten years
earlier or would be twenty years later.
Why these things happen is not hard to understand. For at
least two centuries, significant fractions of American society
have held moralistic views about drug dependence and be-
lieved that drug use is dangerous, unwise, and immoral. Dur-
ing times when drug use is increasing or has stabilized, other
voices argue that moral choices are not the law’s business
and some argue that drug use is a good or at least culturally
tolerated thing. Drug use peaks, and begins to decline, how-
ever, because the balance of views has changed. As the de-
cline continues, the balance of social attitudes changes. Fewer
people use drugs and feel a need to justify themselves. The
moral arguments against drug use become louder and com-
mand wider support. People of dissenting views become more
reluctant to speak out for fear they themselves will seem
foolish, irresponsible, or deviant. More people come to pro-
mote, support, or administer harsh policies and fewer are
prepared to oppose them. The pattern of harsher laws, more
vigorous enforcement, and scapegoating of minority groups
occurs.
Eventually, however, the fervor abates, the harsh policies
soften, enforcement becomes less single-minded, and the fo-

cus on minority groups blurs. Drug use stabilizes or begins
mildly to increase. People become more confident opposing
majority sentiment and a wider range of views is expressed.
The drug policy cycle just described has recurred in the
past thirty years. Drug use peaked in the late 1970s and early
1980s, the harshest laws were enacted and the most vigorous
enforcement occurred in the ensuing fifteen years, and signs
of gradual easing of both fervor and policy have been evident
since the mid-1990s. The signs include the movement to cre-
ate drug courts for drug-abusing offenders, the enactment
of laws in many states calling for diversion of drug-depen-
dent first- and second-time offenders into treatment pro-
grams, and the passage of referendums supporting medical
use of marijuana.
There is no reason to doubt that something similar is hap-
pening in relation to crime. By the most cautious measure,
crime rates peaked around 1990 (some would say around
1980) and fell continuously and precipitously during the
1990s. During the several decades when crime rates were
rising, there was vigorous debate about trade-offs in crime
control policy between civil liberties and public safety. That
debate had largely stilled by 1990 and remained still for much
of the following decade. Just as people become uncomfortable
during peak prohibitionist periods arguing for individuals’
rights to use drugs, people become uncomfortable during
peak anticrime periods arguing for the need to treat offenders
fairly and to recognize the complex causes of much criminal-
ity. No one who read American newspapers during the 1990s
could have failed to observe that both major parties were
consistently tough on crime and loathe to risk any other
public perception. Former President Bill Clinton famously
observed in 1994, “I can be nicked on a lot, but no one can
say I’m soft on crime.”
As a result, closely paralleling experience with drugs, crime
control policies became unprecedentedly harsh during the
early and mid-1990s. That was the time of three-strikes,
truth-in-sentencing laws, and zero-tolerance policing, and of

a striking increase in the severity of punishments and in the
numbers of people in prisons and jails.
And now, the fervor is abating, policies gradually are soft-
ening, a wider range of voices is being heard, and the prison
population is stabilizing and in some places declining a little.
Had American policy makers in the 1980s read the readily
available historical literature, heeded its lessons, and more
sensitively interpreted evidence of public concern, the war
on drugs would have taken very different forms. And had
American policy makers in the 1990s heeded history’s les-
sons, crime control policies would have been very different.
Many fewer Americans would have been sent to prison. More
public monies would have been available for health care,
education, and social welfare. The criminal justice system
would have intruded into many fewer black Americans’ lives.
If only American policy makers had not allowed themselves
to be carried away by short-term and predictable transient
emotions, America’s criminal justice system would look very
different.
The organization of American government makes it espe-
cially vulnerable to emotional overreaction. American policy
makers need to worry even more than policy makers else-
where that they will be carried away by shifts in sensibilities
about crime, drugs, and other unwanted behaviors, or by
moral panics, and adopt inhumane and unjust policies that
later they will regret.
American institutions provide less insulation from short-
term passions and emotions than do governmental institu-
tions in other Western countries. Most prosecuting attorneys
are elected and those who are not are selected according to
partisan political criteria. A large majority of American state
judges are elected. No informed person would deny that ap-
pointments of federal judges are fundamentally partisan and
political.
Only rarely in Western Europe or the other major English-
speaking countries are judges and prosecutors elected. Prose-
cutors and judges elsewhere are often career civil servants.

Where they are not, they are typically selected through non-
partisan merit procedures. Nowhere else do legislators try to
micro-manage sentencing and parole decisions. In the United
States, however, policy makers and public officials often
pride themselves on responding to changes in public opinion
and to heightened anxieties, and are accordingly less likely
than those elsewhere to stand firm against winds of emotion
when they sweep across the national consciousness.
American policy makers over the past thirty years did not
try to keep their heads and retain a sense of perspective. What
was in effect a form of false consciousness led to adoption of
drug and crime control policies that were too simple, too
harsh, and too wasteful. Centuries of experience with the
excesses of witch-hunts and the persecution of heretics have
taught church officials caution. Those who feel moved to
punish heretics know they must search their souls to consider
whether that is because controversial beliefs genuinely
threaten harm or because they are concerned that the author-
ity of the church is weakening. Likewise, makers of crime
control policy should examine their consciences to consider
whether they support one policy or another for valid reasons
or merely because of fleeting beliefs, opinions, and passions.
Because American policy makers lost their way, hundreds
of thousands of people are in prisons because they were in
the wrong place at the wrong historical moment. Their bad
luck was to have committed crimes, or been convicted and
sentenced, during a period when public and official attitudes
toward crime were harsh and single-minded.
Many thousands of aging offenders convicted of trifling
crimes in the mid- to late 1990s occupy California prisons.
Their misfortune is to have been sentenced before the Califor-
nia Supreme Court gave judges the authority to sidestep Cali-
fornia’s three-strikes law and before prosecutors throughout
the state began to restrict its application to genuinely serious
cases.
Many thousands of people are serving decades-long sen-
tences in federal prisons for non-violent drug crimes. Their
misfortune is to have been sentenced in federal courts before

avoidance of sentencing guidelines by federal judges and
prosecutors became common practice.
Hundreds of thousands of people, mostly but not only of
minority and disadvantaged backgrounds, have spent much
of their young adulthood in prison for drug crimes. Their
misfortune is that unwisely, but for young people not uncom-
monly, and typically as a result of peer influences and teenag-
ers’ sense of invincibility, they experimented with drug use,
got hooked, and got caught—in a time when antidrug poli-
cies were unprecedentedly harsh.
Reasonable people disagree over why policies are so harsh,
so many people are in prison, and so many of those are black
and Hispanic. In the most literal sense, the explanation is
that American politicians adopted unduly harsh policies and
the public let them do it. That does not explain, however,
why the public supported those policies. Conventional expla-
nations range from the inevitable effects of rising crime rates,
reflection of the public will, and belief in the crime-preven-
tive effectiveness of tough policies to political cynicism, par-
tisan manipulation of public anxieties, and indifference to
the well-being of black Americans.
Current penal policies have trapped American policy mak-
ers much as Brer Fox got stuck in Joel Chandler Harris’s tar-
baby story. Brer Rabbit, not at all interested in becoming
Brer Fox’s supper, persuaded him that the tar-baby would
be juicier and tastier. Brer Fox grabbed hold of the tar-baby’s
arm, only to find that he could not pull his front paw away.
When he tried to push away with the other front paw, that
too became stuck. In the end, as Brer Rabbit walked away,
smiling quietly, Brer Fox was hopelessly stuck. He had
grabbed hold of what he thought he wanted, and he could
not break free.
American policy makers, having grabbed hold of current
policies, are stuck and do not know how to get loose. It does
not matter that we now know that current crime policies
cannot be justified on the merits. Although many people
long believed that severe crime control policies generally, or
particular policies like zero-tolerance policing in New York
City and the three-strikes law in California, substantially re-
duced crime rates, most informed practitioners and scholars
no longer believe that to be true. At the margins such policies
probably influenced crime rates, but only as small parts of
much more complicated stories. Crime rates fell sharply in
all large cities during the 1990s, including the large majority
that did not adopt zero-tolerance policing, and in all populous
states, including the majority that did not enact three-strikes
laws. Particular policies and particular laws in individual
cities and states may have influenced crime rate declines in
those places, but most of the steep decline was the result of
normative and behavioral changes that affected the entire
country.
It is more accurate to refer to normative and behavioral
changes that affected most Western countries. Crime rates fell
in most countries in the 1990s, including Germany, England,
Canada, the Scandinavian countries, Scotland, the Nether-
lands, and the United States. Crime rates have even been
falling since the mid-1990s in many of the countries of Eastern
Europe, a region that has experienced enormous social and
economic disruption.
No one has a good explanation for why crime rates are
down in most of the Western world, but such things have
happened before. Crime rates fell in most Western countries
for a century or more beginning around 1830. Historians
differ in their inevitably speculative explanations as to why
that happened but not as to whether it happened. Something
similar is happening now, and it is affecting Maine, Mary-
land, Minnesota, and Manitoba on the western side of the
Atlantic and Manchester, Mainz, Malmo, and Messina on the
eastern.
To some readers it may seem counterintuitive that policies
like zero-tolerance policing and three-strikes sentencing had
only modest effects on crime. That is because prevailing sen-
sibilities in recent years made Americans want to believe
harsh policies would work. Many Americans appear to be-
lieve that crimes occur primarily because offenders are bad

people or because punishments are insufficiently severe. If
that is true, incapacitating the incorrigibly bad people and
threatening lengthy prison sentences for the others should
reduce crime rates.
In many other Western countries, by contrast, ordinary
crime is understood primarily to be the product of personal
disadvantage or disability and social disorganization, and it
seems counterintuitive that crime can be much affected unless
those underlying problems are addressed. Few people deny
that offenders have free will and choose to commit crimes.
Self-evidently, they have and they do. That recognition,
however, can coexist with recognition that childhood abuse,
poverty, mental disability, limited opportunities, lack of mar-
ketable skills, and socialization into deviant values often lead
individuals to make choices they would otherwise not make.
To many disadvantaged inner-city American teenagers, for
example, drug dealing, in comparison with other available
options, appears to be a sensible and relatively low-risk way
to earn money and improve their lives. So they choose to
sell drugs. Privileged teenagers much less often see it that
way because their circumstances are different, and other ave-
nues to success and happiness are open. So it is not that
disadvantaged kids do not make choices, but that the choices
available to them are often far less attractive than those open
to privileged kids. A just society would take those differences
into account in punishing offenders and in setting crime con-
trol policies.
Many European policy makers think this way now, and in
earlier times so did many American policy makers. Historian
David Rothman described the prevalent Progressive Era
views during the early twentieth century: “No one raised in
a slum could be held strictly accountable for his actions. The
wretchedness of the social setting was so great that responsi-
bility could not be assigned in a uniform and predictable
fashion. Elemental fairness dictated that the offender be
treated as an individual. It was not just a sensible and effec-
tive principle, but a just one.” Tony Blair’s Labour Govern-

ment in England famously expressed this in the back half of
its campaign slogan “Tough on Crime, Tough on the Causes
of Crime.”
In any case, for countries, states, or societies, changes in
crime rates or patterns do not necessarily lead to changes in
punishment policies or patterns. Put crisply, at a societal level
crime does not cause punishment. Imprisonment rates and the
severity of punishment move independently from changes in
crime rates, patterns, and trends. Governments decide how
much punishment they want, and these decisions are in no
simple way related to crime rates. This can be seen by compar-
ing crime and punishment trends in Finland, Germany, and
the United States between 1960 and 1990. The trends are
close to identical. Violent crime rates in all three countries
grew by a factor of 3-to-4 and homicide rates more than
doubled. Yet the U.S. imprisonment rate quadrupled in that
period, the Finnish rate fell by 60 percent, and the German
rate was broadly stable.
Likewise, punishment does not cause crime (or less cryptic-
ally, punishment policies do not have major effects on crime
rates or trends). This can be seen by comparing crime rates
over the past twenty years of comparable and closely related
countries that have very different punishment policies. Cana-
da’s incarceration rate fluctuated within a narrow band of
100 to 125 per 100,000 between 1980 and 2000 while the
U.S. rate more than tripled, growing from 200 per 100,000
in 1980 to 700 per 100,000 in 2002. Yet crime rate patterns
in Canada between 1980 and 2000 closely parallel those in
the United States with the crucial difference that they were
not as high.
And a Scandinavian example: crime rates in Finland and
Denmark have moved in parallel since the early 1960s, though
Finnish crime rates have been consistently lower. Finland’s
incarceration rate, twice as high in 1965 (130 per 100,000
compared to 65 per 100,000), fell by half, Denmark’s stayed
about the same, and both fluctuated in the low 60s per
100,000 in the 1990s. The Finnish incarceration rate in 2000
(56) was marginally lower than Denmark’s (61). In both coun-

tries the crime rate in 1990 was essentially the same in 2000.
Whether current American crime policies were an inexora-
ble product of recent crime patterns, and whether they can
be justified on the merits, or in either case not, they are
difficult to alter or repeal. Enacting a three-strikes law, a life-
without-possibility-of-parole law, or a law calling for manda-
tory minimum sentences is incomparably easier than repeal-
ing such laws. They were politically popular in the 1980s
and 1990s. No candidate or incumbent who supported them
imperiled his or her electoral prospects. In contemporary
America, politicians’ support for laws aimed at prevention
of crime, hostility to offenders, or well-being of victims does
not require political courage.
Political courage is required, however, to propose or sup-
port repeal of tough laws. Even in a time when crime has
fallen well down lists of the public’s primary concerns, a
vote to repeal, narrow, or weaken a three-strikes law can be
portrayed as soft on crime. This makes elected officials risk
averse. Conservative New York Republican Governor George
Pataki in 1995 first proposed reducing penalties for non-
violent offenders under the severe Rockefeller drug laws; in
mid-2003, the legislature had still not acted. Conservative
Republican U.S. Senator Orrin Hatch, according to a Senate
staffer, long believed reducing federal drug sentences was
“the right thing to do, but he couldn’t do it for political
reasons.” The United States Sentencing Commission in 1995
proposed repeal of the federal 100-to-1 policy, adopted in
1986, that penalizes crack cocaine offenses as harshly as pow-
der cocaine offenses involving quantities 100 times larger.
In 2003, the law remained unchanged.
Leaving things as they are poses no electoral risks. Inertia
may result in wasteful public spending or avoidable damage
to the lives and interests of offenders, their families, and their
communities, but responsibility for these things is diffuse. No
one loses elections for failing to lead or support a campaign
for repeal of laws that are tough on drugs or crime.
This is unfortunate and ironic, because there is plenty of
evidence that broad-based public sentiments support more

nuanced, compassionate, and constructive policies than those
now in place. Few people argue that violence or theft is morally
excusable, or that use of heroin, cocaine, or ecstasy is a good
thing. Many more people than a decade ago, however, accept
that human lives are messy and complicated, that everyone
makes mistakes, that sensible drug and crime policies should
recognize that, and that current policies do not.
What is the evidence for the preceding paragraph’s claims?
Drug policy referenda for starters. Despite unceasing opposi-
tion from the federal government, by 2002 voters in eight
jurisdictions (Alabama, California, Colorado, Maine, Nevada,
Oregon, Washington, and the District of Columbia) had en-
acted referenda legalizing medical use of marijuana, and
Hawaii did so by legislation. In July 2002, the California
Supreme Court, in The People v. Mower, 122 Cal. Rptr. 2d
326 (2002), held that a person who cultivates or possesses
marijuana for medical purposes may not be convicted of un-
lawful possession of drugs. Earlier, in 2000, California voters
enacted Proposition 36, which required large numbers of
first- and second-time drug offenders to be diverted from
prosecution and imprisonment to drug treatment. District of
Columbia voters in November 2002 adopted Ballot Initiative
62, which prescribed a similar preference for treatment over
prosecution.
More evidence can be found in results from public opinion
surveys showing that crime is no longer widely numbered
among the country’s most pressing problems. From 1980 to
1990, crime ranked most years in the top three. In March
2002, according to the Gallup Poll, crime ranked fifteenth
and drugs twelfth, behind poverty, dissatisfaction with gov-
ernment, and the high cost of living. In a 1994 poll by Peter
D. Hart Research Associates, 48 percent of Americans said
their preferred crime control strategy would be to address
the underlying causes of crime, and 42 percent preferred
stricter sentencing. In a 2001 Hart Poll, 65 percent of Ameri-
cans approved the root causes approach and only 32 percent
preferred stricter sentencing.

The general public holds complicated views about punish-
ment that are neither monolithically nor single-mindedly pu-
nitive. Ordinary citizens want criminals punished, but also
want them treated. They believe punishments in general are
too soft, but when they focus on particular cases typically
support sentences less severe than judges now impose. They
believe that crime is a product of bad character and bad judg-
ment, but also that drug and alcohol dependence, bad home
lives, and disadvantage are among crime’s major causes.
There is also evidence that many criminal justice officials
believe current policies are too severe and too expensive,
and long have done. A January 2002 New York Times article
by Fox Butterfield, for example, reported that some states
were changing their laws on sentencing and punishment in
various ways to reduce their rigidity and severity. A Ninth
Circuit Federal Court of Appeals in The People v. Andrade, 270
Fed. 3rd 743 (2001), struck down California’s three-strikes
law, as applied in that particular case. The court held that
a fifty-year sentence for Andrade’s theft of $150 in videotapes
from K-Mart violated the U.S. Constitution’s provision for-
bidding “cruel and unusual punishments.” (The U.S. su-
preme court, however, in mid-2003 held otherwise.) Several
years earlier, in The People v. Romero, 53 Cal. Rep. 2d 789
(1996), and in The People v. Alvarez, 60 Cal. Rep. 2d 93 (1997),
a California Supreme Court composed entirely of judges
appointed by conservative Republican governors Ronald
Reagan, George Deukmeijian, and Pete Wilson reinterpreted
the three-strikes law to empower judges to sidestep it when
they believed it appropriate to do so. More recently, there
is much evidence that the three-strikes law has fallen into
partial desuetude, invoked only for extremely serious cases.
Washington Post reporter Rene Sanchez wrote that, “prosecu-
tors insist they are careful to use discretion in deciding
whether to invoke a third-strike prison sentence.” Los
Angeles County prosecutor Steve Cooley ran for office and
was elected on a platform promising to use three-strikes pros-
ecutions only in very serious cases.

That is a lot of California three-strikes, but together with
the California drug referendum it starkly illustrates the prob-
lem. Voters, prosecutors, trial judges, and appellate courts
have acted in various ways to change California laws to make
them more flexible and discriminating. Legislators, however,
have not. The three-strikes law remains unchanged on the
books. Thousands of people are serving decades-long sen-
tences in California who, had they been sentenced a few years
later, would not (if drug offenders) have been sent to prison
at all or who (if third-strike property, drug, or minor violent
offenders) would not have been sent there for anywhere near
so long.
So we appear to be stuck. Voters and many criminal justice
officials would support repeal and refashioning of much of
the harsh penal legislation of the past quarter century, but
elected officials are too risk averse to do it. Crime and punish-
ment has come to resemble the signature issues of single-
issue interest groups such as the National Rifle Association
or the Right-to-Life movement. Only a few percent of voters
in a district may base their electoral choices solely on candi-
dates’ positions on gun control or abortion but in places
where victory margins are sometimes small, they may make
the difference. The safe course is to oppose gun control and
oppose abortion, and many electoral candidates do so most
of the time, whatever their private views may be.
Crime and punishment has become a comparable issue. No
matter how amenable most voters may be to sensible policy
changes, the existence of a few percent who will always
oppose candidates who appear “soft” on crime or likely to
“coddle” offenders can immobilize legislatures.
Crime and punishment may be even more immobilizing
than gun control and abortion. There is no “pro-criminal”
sentiment to offset “tough on crime” sentiment as there is a
gun control movement that opposes the NRA and a pro-
choice movement that opposes the right-to-life movement.
It was common in the 1990s for politicians to suggest that
one must be “for victims” or “for criminals.” Given the emo-

tive force of crime as an issue, few public officials are willing
to risk the “for criminals” label.
But, unless some elected officials are prepared to propose
fundamental policy changes, and work to achieve them, we
will not be able to let loose of the tar-baby. America will
remain stuck. Too many people who shouldn’t go to prison
will continue to be sent there and to stay there too long. Too
many lives will continue unnecessarily to be blighted.
Chapter 8 sets out three sets of proposals for correcting
and ameliorating the worst excesses of the American criminal
justice system. The first set is aspirational and calls on Ameri-
can policy makers to learn from both history and contempo-
rary experience that they should not allow themselves to get
carried away by the passions of the moment or to adopt
policies that will later be regretted.
The second set addresses features of American government
that create higher risks in the United States than in other
Western countries that policy makers will get carried away.
These include most prominently the American practice of
electing many judges and prosecutors, and taking partisan
political considerations into account when appointing most
of the rest. Judges and prosecutors in other Western countries
are either career civil servants who have undergone special-
ized university-level training or are appointed under nonpar-
tisan meritocratic procedures. The American system ties crim-
inal justice policies and decisions in individual cases to
election returns and to officials’ personal and political self-
interest. Rather than buffer policy and decisions about indi-
viduals’ fates from emotional overreactions, moral panics,
and changing sensibilities, American institutional arrange-
ments are almost designed to assure their influence.
The third set of proposals concerns changes to contempo-
rary criminal justice policies. Early release laws should be
adopted that allow panels of judges or corrections officials
to review the sentences of all offenders serving very long
terms to determine whether and when they can safely and
justly be released. Three strikes and mandatory minimum

laws should be repealed and replaced by sentencing guide-
lines that allow judges to draw sensible distinctions among
offenders. If that is not viable, all such laws should be
amended to authorize judges to disregard them in appropriate
cases. A variety of measures should be adopted to prevent
enactment of ill-considered legislation, including require-
ments that all proposed new laws be accompanied by appro-
priations of whatever funds will be required to implement
them and by disparity analyses that demonstrate their likely
differential effects on minority and women offenders. To be-
gin to lessen racial disparities in the courts and prisons, not
only proposed new laws but also all existing practices should
be subjected to disparity analyses. These would investigate
whether and how new and current practices treat minority
and women offenders worse than whites and men, and serve
as the basis for deciding whether disparity-causing practices
can be justified.
We know how to create a humane and effective criminal
justice system. Experience with sentencing guidelines shows
how to adopt comprehensive sentencing policies that treat
offenders fairly and consistently, requiring severe punish-
ments for those who deserve them and less severe but appro-
priate punishments for others. Accumulating evidence on the
effectiveness of drug courts and drug treatment shows that
greatly increased investment in drug treatment can prevent
crime, save money, and rebuild lives. Accumulating evidence
on the effectiveness of a wide range of other treatment pro-
grams has undermined the “nothing works” attitudes of the
1980s and 1990s and shows that diversion of public funds
from prisons to treatment programs would pay financial and
crime-prevention dividends. If American criminal justice sys-
tems would eliminate the worst policies of the last thirty
years, we could re-create systems that were no longer out-
of-step with our own history or with the practices of other
Western countries.

2
Why So Many Americans Are in Prison
American imprisonment rates, more than 700 per 100,000

residents behind bars at the end of December 2002, have
reached unprecedented levels compared with other times in
U.S. history or with current times in other Western countries.
In other Western countries between 50 and 150 residents per
100,000 are in prison or jail on an average day. In Sweden,
one in every 2,000 people is locked up; that is the lowest
rate. In Portugal, the highest, it is one in 650. By contrast,
in the United States, one of every 120 people is in prison or
jail; that is five to twelve times the rates in other Western
countries (A. Kuhn 1998; Bureau of Justice Statistics 2003).
When those under 16 and over 70 are disregarded, one in
eighty Americans each day awakens inside a prison or jail.
Nearly one of every seven black American men in their late
twenties is among them.
American punishment policies are unusually severe in
other respects. Only in the United States are constitutional
and other safeguards of criminal defendants systematically
being reduced; throughout Europe, under the influence of the
European Human Rights Convention and Court, defendants’
procedural protections have been expanding for the last
twenty years (Kurki 2001). In the United States, legislatures
and courts are cutting back on prisoners’ rights and privi-
leges; in Europe, they are steadily expanding under the in-
fluence of the European Convention for the Prevention of
Torture and Inhuman or Degrading Treatment or Punishment
(Morgan 2001). Among advanced Western countries, only
the United States retains and uses the death penalty, and
with increasing frequency. Only the United States has adopted
21
broad three-strikes and extensive mandatory minimum sen-
tencing laws.
Only in the United States is “prison gerontologist” an imag-
inable civil service career ladder. Only the United States uses
life-without-possibility-of-parole sentences; elsewhere even
murderers sentenced to life terms are eligible for parole or
executive-branch pardon, and are typically released after
eight to twelve years. In much of Europe, fourteen years is
the longest prison sentence that may lawfully be imposed.
Only in the United States are prison sentences longer than
one or two years common; in most countries, fewer than 5
percent of sentences are for a year or longer. In the United
States in 1994, the average sentence among people sent to
state prisons for felonies was seventy-one months, and that
was before the effects of three-strikes and truth-in-sentenc-
ing laws were being felt. Among those in prison in recent
years, more than half were serving terms exceeding ten years
(Bureau of Justice Statistics 1998, table 1.3). In 1997 the mean
average sentence an inmate in a state prison would serve
before release was ninety-one months (Bureau of Justice Sta-
tistics 2000, table 4.8).
All of this is a drastic change from earlier times. In the
1930s, for example, the United States had incarceration rates
comparable to or lower than those of many European coun-
tries, including England, France, Switzerland, and Finland
(Tonry and Hatlestad 1997, chap. 3). Through the early 1970s,
American rates were not much higher than those elsewhere.
More recently, in the 1960s, the United States was in the
international mainstream of criminal justice policies. In many
respects, it was leading the way toward creation of criminal
justice practices that were humane and effective, setting stan-
dards that other countries tried to emulate. The death penalty
was withering away, the incarceration rate was dropping and
comparable to those in other Western countries (Blumstein
and Cohen 1973; Zimring and Hawkins 1991), the courts were
establishing, expanding, and refining defendants’ procedural
protections, a prisoners’ rights movement was taking shape,

and crime control was not generally viewed as a partisan or
ideological issue.
So what happened? Why did American crime control and
punishment policies become so much harsher than in earlier
American times or in other places now?
Eight different plausible stories can be told. The first is
empirical. American crime rates are higher or have increased
more than other countries’, and punishment patterns and
policies simply reflect that reality. Crime rates increased and
carried punishments with them.
The second is psephological. No matter what the crime
trends and rates, the public demanded tougher penalties, and
elected officials bowed to that demand.
The third is cynically political. Conservative Republican
politicians realized they could use crime, and also welfare,
immigration, and affirmative action, as “wedge” issues de-
signed to separate white working-class voters from the Demo-
cratic Party. The text of this story is about crime; the subtext
is about race. Current policies are the result.
The fourth story is structurally political. Developments of
the past quarter century have fragmented the electorate into
a mélange of single-issue political groups. People have lost
confidence in government’s ability to achieve positive goals.
Politicians have had to seek broad-based support around
emotional but essentially negative issues, including opposi-
tion to crime, welfare, and immigration, that offend no politi-
cally powerful groups.
The fifth, the arrival of the “risk society,” and the sixth,
postmodernist angst, are related. I separate them somewhat
artificially to isolate their central claims. The risk society
story is that the insecurities and social isolation of our times
have made us preoccupied with uncertainty, danger, and
risk. Modern crime control and penal policies accordingly
are concerned above all to identify, quantify, and reduce
risk or the perception of risk. Insecurity is so profound and
so pervasive that traditional concerns about fairness, justice,
and equality have become unaffordable luxuries.
WHY SO MANY AMERICANS ARE IN PRISON 23

The sixth, the postmodernist angst story, is that a wide
range of economic, social, and cultural changes have made
people anxious, risk averse, and insecure, and desperate for
comforting explanations and easy cures of what ails them.
No simple and reassuring explanations exist, so politicians
have provided scapegoats. Criminals are among the most vul-
nerable and viscerally plausible scapegoats and politicians
have tried to placate voters’ discomforts and win votes by
being tough on criminals.
The seventh is historical and is based on comparison of
social, political, and legal developments over the past three
centuries in the United States, Germany, and France. It at-
tempts to explain why American punishment policies and
attitudes are so different from those in the two most powerful
European countries. Its key themes are deep cultural differ-
ences in attitudes toward the degradation of prisoners and
toward the application of egalitarian values to offenders.
Eighteenth-century France and Germany were inegalitarian
societies with well-developed social hierarchies, while eigh-
teenth-century America already aspired to be an egalitarian
society without established hierarchy.
Those differences, it is said, explain punishment practices
and attitudes then and now. In France and Germany, high-
status criminals in the eighteenth century were often treated
with respect, housed decently and comfortably, and, if pun-
ished, punished in ways that seemed dignified. Common,
low-status criminals were treated roughly and without re-
spect. As part of the shift toward more egalitarian values, as
part of the rejection of the social hierarchies of the past,
France and Germany have “leveled up,” have attempted to
accord all offenders the dignity and respect previously
granted only the elite.
Eighteenth-century America, by contrast, was a place that
rejected notions of aristocratic or social hierarchy, and elite
offenders often were dealt with in the same ways as were
others. Fast-forwarded to the twentieth century, egalitarian
ideas about punishment persist, but in the absence of cultural
commitment to leveling up. Instead, something like leveling

down has occurred. Commitment to ideas that prisoners
should be treated with respect and dignity translates into
mild punishment practices. The absence of that commitment
opens up possibilities of harshness, debasement, and lack of
compassion.
The final story is historical: complex, regularly recurring,
but poorly understood interactions among crime trends, pub-
lic attitudes, and policy making shape our sensibilities and
through them our thought, policy debates, and policies. A
succession of upsetting incidents has produced a series of
moral panics that, among other things, has led to artificially
heightened anxieties and fierce overreactions. Current poli-
cies are a predictable and understandable, but regrettable,
result. This is much the most plausible and complete story
and I tell it in considerable detail in chapter 5.
Other stories have been offered besides these eight, but
they are so idiosyncratic or obviously inadequate that they
can be dismissed out of hand. One, associated with German
sociologist Zygmunt Bauman (1991) and Norwegian criminol-
ogist Nils Christie (2000) is that current policies are the inevi-
table outgrowth of “modernity.” The modern state goes about
its business in a depersonalized neutral way. The “punitive
shift,” according to criminologist Simon Hallsworth (2000,
p. 148), “is born then out of the progressive development
of ever more rational forms of organization and their applica-
tion in the criminal justice system.”
This ignores the commonplace realities that nothing like
American penal policies exist in any other Western coun-
try—suggesting that the modernity thesis applies nowhere
else, which would be odd. It also does not acknowledge that
arguments in favor of current American policies almost al-
ways include ideas about accountability, personal responsi-
bility, and moral culpability. Conservative American propo-
nents of recent policies such as James Q. Wilson, John DiIulio,
and William Bennett would be surprised to learn that their
proposals were not predicated on normative premises.
Another dismissable story is that contemporary policies
and practices, far from reflecting the onset of modernity,
reflect the erosion of modernity and “speak directly to puni-
tive passions and are profoundly inspired by them” (2000,
p. 156). Somewhat similarly, American sociologist Jonathan
Simon argues that a long-term civilizing process has somehow
been reversed, at least temporarily, and that “cruelty-seek-
ing” has become a basic theme in “contemporary penality”
(2001). Like the modernity story, this one suffers from the
difficulty that the phenomenon it explains primarily charac-
terizes the United States, though it purportedly applies more
generally. As chapters 3 and 5 show, people characterized
by majority culture as seriously deviant always suffer from
negative stereotypes and seldom benefit from others’ empa-
thy. That is not new, and it cannot explain why American
policies were adopted when they were, and not earlier, or
why similar policies were not adopted elsewhere. Rhetoric
about cruelty and punitive passions does not a plausible story
make.
The first of the eight main stories, about the effects of rising
crime rates on imprisonment rates, is flatly wrong. There is no
simple or direct relation between crime rates and punishment
policies or imprisonment rates, though rising crime rates are
important. Taken together, the first six stories, whatever their
surface plausibility, suffer the same fundamental defect as
the modernity and cruelty stories. They, or substantially com-
parable developments, characterize all wealthy Western coun-
tries but only the United States has adopted crime control
policies and practices of unprecedented severity.
Most of the more familiar explanations for why contempo-
rary U.S. crime control policies emerged as they did have
some merit, but none by itself is convincing. No single factor
could cause so massive a change in policy (Garland 1990,
2001a). A sophisticated and nuanced explanation would take
all into account.
Sociologist David Garland has recently attempted to ex-
plain why English and American crime policies have evolved
as they have, and he offers a rich, complex explanation, which
I discuss below. His explanation confronts a major difficulty,
however. American imprisonment rates grew by 400 percent
between 1970 and 2002. English rates grew by 80 percent
and remained lower in 2002 (around 125 per 100,000) than
American rates were in 1970 (140 per 100,000). The English
imprisonment rate in 2002 was not significantly higher than
the highest rates of other periods. The American rate is nearly
four times higher than at any time in American history before
the last thirty years’ run-up. Similar social, economic, and
cultural forces may have affected sensibilities, politics, and
policies in both countries, but the consequences were funda-
mentally different.
Historian James Whitman’s effort to explain differences
between France and Germany, and the United States, is full
of wisdom and insight, but it suffers from a temporal limita-
tion that parallels the geographical limits of Garland’s argu-
ment. Garland’s analysis does not hold up across space; Whit-
man’s fails across time. If comparisons of French and German
practices with American ones today can be explained by
reference to cultural attitudes toward egalitarianism and deg-
radation, why did things look so differently for the first
two-thirds of the twentieth century? Before 1970, American
punishment practices were among the mildest in Western
countries and the United States was at the forefront of efforts
to rehabilitate offenders and advance offenders’ and prison-
ers’ rights. France and Germany lagged far behind.
Many explanations have been given for why American
policies and practices developed as they did. Some are plainly
wrong. Most are at best fragments of the whole explanation.
Many people, however, subscribe to various of them and it
seems to me useful to canvass them and show why they are
inadequate before offering a better explanation.
Crime Rates and Trends
The first explanation for why so many Americans are in

prison, that U.S. crime rates are higher or faster rising than
other countries’, has virtually no validity. Crime rates in the
United States in the 1990s were, for the most part, not higher
than those in other Western countries. We know this from
the International Crime Victimization Survey, conducted by
national governments in most major Western countries since
1989 (e.g., Mayhew and van Dijk 1997; Kesteren, Mayhew,
and Nieuwbeerta 2000). For property crimes, the United
States is in the middle of the pack. Chances of being burglar-
ized, having your pocket picked or your car stolen are consid-
erably higher in England and several European countries.
For violent crimes, American rates are among the highest,
along with Australia, Canada, Spain, and France, but not the
highest. Chances of being robbed, assaulted, or victim of a
stranger rape are higher in several other Western countries.
Where the United States stands out is in gun violence. Our
rates of robberies and assaults involving guns, and of gun
homicides, are substantially higher than elsewhere (Zimring
and Hawkins 1997). That is important. However, less than
a fourth of those sentenced to prison are convicted of violent
crimes of any type, so that is not why U.S. prison patterns
and penal policies are so different.
If absolutely higher crime rates do not explain the severity
of American punishment policies, perhaps crime trends do.
Perhaps there is a necessary connection between crime rates
and imprisonment rates. When crime rates rise, imprisonment
rates follow, and that is why the number of people locked
up increased by nearly seven times in the past thirty years,
from around 300,000 in 1972 to 2,000,000 at the end of 2002.
There is a surface logic to this claim: more crime yields more
arrests yields more convictions yields more prison sentences
yields higher prison populations and imprisonment rates.
Whatever the logic, however, comparisons of crime and pun-
ishment trends in the United States with those in other coun-
tries show that there is no inexorable relation between crime
and punishment rates.
Figures 2.1–2.3 show trends in imprisonment, homicide,
and violent crime rates in the United States, Germany, and
Finland from the 1960s to the early 1990s. They stop then
because crime rates in the United States began a steady de-

cline in 1990–91 and my aim is to compare national experi-
ences while crime was rising. There are some differences in
what the data represent in the three countries. The imprison-
ment rates, based on the numbers of convicted offenders in
prison per 100,000 population, are comparable. The homicide
rates are calculated somewhat differently; the Finns, for ex-
ample, include both attempted and completed homicides
whereas the United States counts only completed homicides.
The violent crime rates include robbery, rape, and serious
assaults, and there are differences in the legal definitions of
these crimes in each country. For my purposes, the technical
differences in how crimes are defined are unimportant, how-
ever, since my interest is not in comparing the rates of crime
in the three countries but in comparing crime rate trends.
The figures show trends in what each country counts as
homicide and serious violent crime, and my interest is in the
trends and how they compare with imprisonment trends.
In the United States from 1960 to 1993, as figure 2.1 shows,
homicide, violent crime, and imprisonment rates rose to-
gether through 1980. Imprisonment rates, however, also rose
in 1981–86 when crime rates fell, and rose continuously there-
after including after crime rates peaked in 1991. Between
1960 and 1991, and ignoring intermediate peaks and valleys,
violent crime rates rose by a factor of four and homicide rates
nearly doubled. On the face of it, this pattern supports the
hypothesis that rising crime rates lead to rising prison popu-
lations. The German and Finnish experiences show the hy-
pothesis to be false.
That there is no necessary connection between crime and
imprisonment rates is shown in figures 2.2 and 2.3, which
show comparable data for Germany and Finland during the
same period. Violent crime rates overall and for homicide
increased in both countries about as much as in the United
States, but imprisonment rate patterns are completely dif-
ferent.
Figure 2.2 shows the German data. Violence rates increased
by three or four times and homicide rates more than doubled.

Figure 2.1 Imprisonment, violent crime, and murder rates in the
United States, 1960–1993 (per 100,000 population)
Note: Crime rates are somewhat differently calculated in the United
States than in Finland and Germany; U.S. incarceration rates do not
include jail inmates.
Sources: Bureau of Justice Statistics, Prisoners, various years
(Washington, D.C.: U.S. Department of Justice, Bureau of Justice
Statistics); Federal Bureau of Investigation, Crime in the United States,
various years (Washington, D.C.: U.S. Government Printing Office).
The imprisonment rate, however, fell through the mid-1970s

and remained essentially stable thereafter, fluctuating within
a narrow band.
Figure 2.3 shows the Finnish data. The crime trend pattern
is the same as in the United States and Germany: homicide
rates more than doubled and violent crime rates grew by a
factor of three or four. The Finnish imprisonment rate, how-
ever, 160 per 100,000 at the start, equal to or higher than
the American rate at the time, fell continuously to 60–70 per
100,000.
I could have used data from other countries to make the
same point (A. Kuhn 2001). Crime rate trends in France in
the same period were about the same as those in the United
States, Germany, and Finland. Imprisonment rates zigzagged,

Figure 2.2 Imprisonment, violent crime, and murder rates in
Germany, 1961–1992 (per 100,000 population)
Note: Violent crime and murder rates are somewhat differently
calculated in Germany than in the United States; the German
imprisonment rate excludes pretrial detainees.
Source: German Ministry of Justice (unpublished data provided to
author).
with periods of steep increase alternating with steep declines.

In France, large-scale pardons and commutations generally
accompany national celebrations like the inauguration of a
new president or the two-hundredth anniversary of the
storming of the Bastille. There were also a series of policy
decisions to reduce use of imprisonment by means of newly
authorized community penalties (Kensey and Tournier 2001).
It is well known why German imprisonment rates held
steady and Finnish rates fell. German governments in the late
1960s and early 1970s decided that prison sentences under six
months serve no valid purpose. By separating offenders from
jobs, families, and loved ones, and stigmatizing them with
the label “ex-convict,” short prison sentences damaged pris-
oners’ prospects of later living a law-abiding life. At the
same time, a prison sentence of a few months is too short
for treatment programs to have much effect. As a result, the

Figure 2.3 Imprisonment, violent crime, and murder rates in Finland,
1965–1994 (per 100,000 population)
Note: Violent crime and murder rates are somewhat differently
calculated in Finland than in the United States; the Finnish
imprisonment rate includes pretrial detainees.
Source: Finnish Ministry of Justice (unpublished data provided to
author).
German parliament enacted laws discouraging use of short

prison sentences and authorizing expanded use of commu-
nity penalties. In particular, a system of day fines was estab-
lished, enabling judges to impose burdensome punishments
that were simultaneously scaled to the seriousness of the
crime and calibrated to the offender’s wealth and income.
(One “day-fine unit” is usually one day’s net after-tax income,
with upward adjustments to take account of wealth.) The
government also established a program under which prosecu-
tors could dismiss charges against suspects who accepted
responsibility and agreed to pay the fine or perform the com-
munity service that would have been ordered had they been
convicted. No subsequent German government, including
Helmut Kohl’s conservative Christian Democrats, felt a need
to change those policies. The day-fine system and the prose-
cutorial diversion schemes have several times been broad-
ened to apply to more cases (Weigend 2001).
Government officials in Finland in the 1960s observed that
Finnish crime rates were approximately the same as those in
Denmark, Norway, and Sweden but that the Finnish incarcer-
ation rate was two-to-three times higher. That must mean,
they concluded, that too many people were in prison. Patrik
Törnudd (1993, p. 4), a participant in the discussions at the
time but writing a quarter century later explained: Finnish
officials, policy analysts, and academics “shared an almost
unanimous conviction that Finland’s internationally high
prisoner rate was a disgrace and that it would be possible
to significantly reduce the amount and length of prison sen-
tences without serious repercussions on the crime situation.”
They decided to reduce the use of imprisonment, and Finnish
governments ever since have held to that decision. Over a
quarter century, year-by-year, the incarceration rate fell un-
til, by the 1990s, Finnish crime and imprisonment rates were
comparable to those of other Scandinavian countries (Lappi-
Seppälä 2001).
There are two important points to be made about the con-
trast between the United States and the German, Finnish,
and French experiences. First, violent crime rates increased
rapidly in all four countries over several decades but only
in the United States did imprisonment rates increase contin-
uously or substantially. Second, the German, Finnish, and
French patterns resulted from policy decisions that increased
incarceration was neither an appropriate nor an effective re-
sponse to rising crime rates. American politicians decided
otherwise. American imprisonment rates did not rise because
crime rose. They rose because American politicians wanted
them to rise.
A benign account of the formulation of recent American
punishment policies is that policy makers believed that some-
thing had to be done to stop the increases in crime rates and
adopted deterrence and incapacitation as their strategies. The
only problem with this is that the most drastic policies were
adopted long after crime rates began to fall. As figure 2.1
showed, crime rates for most crimes peaked around 1980,
fell through the mid-1980s, rose for a while for reasons largely
associated with the crack cocaine epidemic, and have since
fallen sharply. The first three-strikes laws, however, were
enacted in Washington state and California only in 1993–94
and the federal “truth-in-sentencing” law authorizing $8 bil-
lion for state prison construction was passed in 1994. The
meanings of these data comparing U.S. crime and imprison-
ment patterns to those of other countries are complex but,
whatever else they may show, they do not show any simple
interaction between crime trends and imprisonment patterns.
Public Opinion
The second explanation for why American punishment poli-

cies are so harsh is that public opinion wished it so. Survey
results sometimes in the 1980s and 1990s showed that “crime”
and “drugs” were “America’s most pressing problem,” that
large majorities thought sentencing is too lenient, and that
the people wanted tougher punishments. On this account,
elected officials merely respected the public will and adopted
harsher policies. Imprisonment rose as a result. If that is wrong,
blame the public.
There are two serious and well-known deficiencies in this
story. The first relates to what “the public” wants. The an-
swer is complicated, but it is not very different from what
practitioners want—penalties no more severe than are now
imposed and substantial efforts to rehabilitate offenders.
The second relates to whether politicians follow or lead
public opinion. The answer is that they lead it, but claim
to follow it.
A mountain of public opinion scholarship and research
shows that the belief that the public is monolithically and
unqualifiedly harsh is fundamentally mistaken (Roberts et
al. 2002). Most people necessarily base their opinions about
punishment on what they know about crime from the mass
media. As a result, many people regard heinous crimes and
bizarre sentences as the norms, many believe sentences are
much softer than they are, and many believe crime rates are
rising when they are falling. As a result, majorities nearly
always report that judges’ sentences are too lenient. Yet, and
this is true not only in the United States but also in Australia,
Canada, and England, when people are asked to propose ap-
propriate sentences for individual cases, they generally pro-
pose sentences shorter than are actually imposed (Roberts et
al. 2002).
A different body of research powerfully demonstrates why
results of opinion polls sometimes should not be taken at
face value. Dan Yankelovich (1991), long-time head of a large
public opinion survey firm, urged that readers of polls never
lose track of the difference between public opinion and public
judgment.
Public opinion is what people say off the top of the head
as a first reaction. It may be uninformed or misinformed,
ignorant or insightful. For some purposes evidence of this
sort is useful. On simple subjects where only preferences
matter (Will you vote for Bush or Gore?), it may be all we
need to know, though even there we know that preferences
change quickly and often can not be taken at face value (for
example, throughout the 1990s, more people told interview-
ers they would vote for black candidates than did). On com-
plex matters (How should the United States balance environ-
mental concerns with energy needs? Should criminal penalties
be increased or decreased?), ordinary public opinion surveys
tell us what is on people’s minds, or what they know or think
they know, but not what their informed opinion would be.
Public judgment is what people say when they know
enough to have an informed opinion. A great deal of opinion
research on many subjects has been premised on this distinc-
tion. Here is how it typically works. At the outset a group
of people are asked to answer a series of questions about a
subject. The results are a measure of their opinion. They then
participate over a few hours or a few days in seminars and
discussions about the subject. The people running the project
try hard to be sure the information presented is balanced and
evenhanded, and participants discuss the subject at length.
Participants are then asked for their views. The results are
an indication of public judgment.
The results of the before and after surveys can be compared
and it can be seen what difference knowledge makes to what
people think. This has been done on numerous subjects, in-
cluding the environment, nuclear power, the death penalty,
and criminal punishment. Public judgment is nearly always
substantially different from public opinion. Having more and
better information often changes what we think. There is
nothing surprising about this. That is everyone’s experience
in day-to-day life—it is why we try to avoid making rash
decisions about things that are important to us—but it is easy
to forget when thinking about surveys of public opinion.
Research of this type on punishment often begins by giving
people vignettes that describe the characteristics of a crime
and a criminal. They are then asked what sentence they
would impose, provided information on sentencing practices
and options, and asked again—a few hours later or the next
day—what sentences they would impose. This has been done
in at least a dozen American states and, though the details
vary, in outline the results are always the same. Many hypo-
thetical offenders who at the outset would have been sen-
tenced to prison are instead sentenced to various community
punishments.
Another major and recurring finding is that when public
judgment is wanted, when ordinary citizens have the infor-
mation to reach informed conclusions, they have the same
complex and ambivalent attitudes toward criminals that
judges and lawyers have. They believe crimes are the prod-
ucts of bad moral choices, disadvantaged backgrounds, and
substance abuse. They want offenders to be punished and
to be rehabilitated. They are much more often willing to pay
more taxes for treatment programs than for prison building.
They insist on prison sentences only for the most violent
crimes (Tonry and Hatlestad 1997, chap. 5; Roberts et al. 2002).
Public opinion surveys show that crime is no longer widely
numbered among the country’s most pressing problems.

Crime ranked in the top three from 1980 to 1990, but by
March 2002 had fallen to fifteenth. In 1994 less than half
of Americans preferred policies that address the underlying
causes of crime over stricter sentencing. By 2001 nearly two-
thirds of Americans favored approaches that address the root
causes of crime (Peter D. Hart Research Associates, Inc. 2002).
A final point about public opinion on punishment is also
important. Prevalent sensibilities in recent decades have pre-
disposed Americans to heightened intolerance of crime and
criminals. Nonetheless, public opinion findings showing that
Americans regard crime or drugs as the nation’s most pressing
problem typically follow, not precede, media and political
concentration on crime. Politicians who attempt to win favor
by demonstrating their toughness nearly always say that they
are honoring citizens’ wishes. The evidence shows, however,
that emphasis by politicians and the media on crime issues
is what causes public anxiety to increase. This is best shown
in a book by Katherine Beckett (1997), who analyzed interac-
tions among media attention to crime and drug issues, politi-
cians’ relative emphasis on those issues, and the results of
opinion surveys. She did this in relation to crime policy in
the 1970s and drug policy in the 1980s. Content analyses
of newspaper and television coverage were compared with
public opinion survey results and showed a recurring pat-
tern. Politicians focused on crime policy, or the media increased
their crime coverage, or both, and then, after those things
happened, opinion surveys showed heightened public con-
cern about crime or drugs, and heightened support for tough
policies.
So public support for harsh policies has coincided with
their adoption. However, it is not public opinion per se that
leads to harsher policies, but politicians’ proposals and pos-
turing and sensational media crime coverage that lead to
changes in public opinion. Americans were prepared to sup-
port the crime control policies of the 1980s and 1990s. Evi-
dence on public judgment, however, shows that they would
also have supported policies that aimed to enhance offenders’
chances of achieving law-abiding lives. Responsibility for re-
cent policies lies not with the public but with the public offi-
cials who failed to provide balanced and humane leadership.
This leads to the third explanation for why American pun-
ishment policies are so harsh—that conservative politicians
for partisan advantage banked the fires of public fear of crime
and then offered harsh policies to dampen those fires. I refer
to conservative politicians even though by the 1990s moder-
ate and some liberal Democrats were equally outspoken in
their anti-crime toughness. No one seems seriously to ques-
tion that Bill Clinton’s decision never to let the Republicans
get to his right on crime was a tactical response designed to
check Republicans’ partisan use of the crime issue. It worked
for Clinton, but not for the tens of thousands of people whose
lives were changed for the worse by the consequences of his
strategy.
Partisan Politics
Crime and punishment have been high on American political

agendas since the late 1960s. Before Republican Barry Gold-
water raised “crime in the streets” as a partisan issue in his
unsuccessful 1964 presidential campaign, public safety was
generally seen as one among many important but unglamor-
ous core functions of government, like public health, public
transit, and public education. Public officials were expected
to do their work conscientiously and well, and systematic
knowledge was widely seen as relevant to the formulation
of policies and the improvement of institutions and practices.
Reasonable people differed over the best approaches for ad-
dressing particular problems. As American Law Institute de-
liberations in the 1950s of provisions of the proposed Model
Penal Code demonstrate, however, the debates were seldom
partisan or ideological (this is discussed in detail in chapter
7). Criminal justice policy was a subject for practitioners and
technocrats, and sentencing was the specialized case-by-case
business of judges and corrections officials.

In recent decades, however, crime control has been at the
center of partisan politics, and policies have been shaped
more by symbols and rhetoric than by substance and knowl-
edge. Political scientists and journalists tell the story of how
that happened. Until the 1960s, the Democratic Party had
dominated electoral politics in most of the South since the
end of Reconstruction. Policy differences and personal rival-
ries were fought over within a state’s Democratic Party,
rather than between parties. The civil rights movement, how-
ever, created a fissure between racial and social policy liberals
and racial and social policy conservatives, initially in the
South but eventually nationally.
Republican strategists seized the opportunity in the 1970s
and 1980s to appeal to “Nixon [later Reagan] Democrats”
by defining sharp differences between the parties on three
“wedge issues”: crime control, welfare, and affirmative ac-
tion. Those issues were meant to separate Democrats from
their traditional white working-class supporters by speaking
to legitimate anxieties in ways that, beneath the surface, also
appealed to racial enmity and stereotype. On the surface,
these three issues are about protecting citizens from crime,
public funds from cheating, and workers from unfair treat-
ment. Below the surface, they were largely about race, and in
political imagery were given black faces. Illustrations include
George Bush’s 1988 use of black Willie Horton to personify
the dangers of soft crime policies; Ronald Reagan’s 1980 use
of black Linda Taylor, Chicago’s “welfare queen,” to embody
welfare fraud; and Jessie Helms’s television ads in his 1990
Senate campaign against black Charlotte mayor Harvey Gantt
illustrating the dangers of affirmative action with images of
an unemployed white worker’s wringing hands.
On crime control, conservatives blamed rising crime rates
on lenient judges and soft punishments, and demanded “tough-
ness.” On welfare, conservatives blamed rising welfare rolls
on “welfare cheats” and laziness, and demanded budget cuts.
On affirmative action, conservatives blamed white un- and
underemployment on “quotas,” and urged elimination of af-

firmative action (Edsall and Edsall 1991; Applebome 1996;
Gest 2001).
Crime’s role as a wedge issue had important consequences.
Issues that are debated on television and examined in fifteen-
and thirty-second commercials necessarily are presented in
simplistic and emotional terms. The appeal palpably is to
public opinion, not public judgment. Matters about which
judges and prosecutors agonize in individual cases are ad-
dressed in slogans and symbols, which often lead later to
adoption of ham-fisted and poorly considered policies.
Few corrections officials, judges, or informed scholars, for
example, support broadly defined three-strikes laws, manda-
tory minimum sentence laws, or sexual psychopath laws in
the forms in which they are typically adopted. This is not
because they do not want criminals to be punished, or are
unsympathetic to victims’ suffering, or are indifferent to
public safety needs, but for practical reasons. Such laws sel-
dom achieve their nominal purposes—there is no credible
evidence that they are significant deterrents to crime—but
generally generate serious unintended consequences. They
are too rigid and often result in unjustly harsh penalties.
They result in circumvention by judges and lawyers who
believe their application inappropriate and unjust in many
cases. They are often redundant because people who commit
serious crimes almost always receive severe penalties anyway
(Tonry 1996, chap. 4).
Many more practitioners and scholars would support such
laws if they were narrowly drawn and carefully crafted to
encompass only genuinely serious crimes and genuinely
threatening offenders. However, in a “sound-bite politics”
era, few politicians were prepared to act as voices of modera-
tion and parsimony, and as a result new sentencing laws
often lacked those qualities.
As important, when crime control became one of the central
issues in American politics, it ceased being a specialized pol-
icy subject and became instead a symbol or metaphor for,
broadly, concepts like “personal responsibility” and vindica-
tion of victims’ interests, and, more narrowly, ideas about
criminals’ immorality and irresponsibility. Analysts of con-
temporary crime control policy often say that its principal
aims are “expressive” rather than functional (J. Kennedy
2000; Garland 2001a). A broadly defined sexual psychopath
law, three-strikes law, or mandatory minimum sentence law
may be ineffective or cruel or unduly costly, but none of
that may matter. If the law’s proponents, and voters, view it
as an expression of revulsion with crime and outrage toward
criminals, whether it will work or achieve just results in
individual cases is often politically irrelevant. When issues
are defined in polar terms of morality and immorality or
responsibility and irresponsibility, few elected officials want
to be found at the wrong pole.
Few informed people will disagree with the broad outlines
of this account. Many liberals might say that the conservative
emphasis on toughness was cynical and intellectually dishon-
est. Many conservatives might respond that they believe that
tougher penalties reduce crime rates and, through public
opinion polls and electoral support for “tough-on-crime” can-
didates, citizens have shown they support such policies; what
better basis for policy making can there be?
Of the explanations offered so far for the severity of current
punishment policies, this is the most plausible. Current Amer-
ican imprisonment policies, and the avoidable damage they
do to prisoners, their families, and their communities, and
the fiscal and opportunity costs they impose on governments
(Hagan and Dinovitzer 1999), may merely be by-products of
an effective political strategy for winning elections. How-
ever, that the Right won, whether cynically or honestly moti-
vated, does not explain why such political appeals were suc-
cessfully made and such policies adopted in the United States,
and not in other countries, and now, and not at other times.
Political Reconfiguration
So what is needed is an explanation for why crime and pun-

ishment served so nicely as a wedge issue, and why so many

elected officials were prepared in recent decades to behave
in ways that opponents and many observers perceived as
demagogic. Social scientists have offered analyses of political
and governmental trends of recent decades that attempt to
explain why crime has received so much more and more
sustained attention from governments and politicians than
have other equally important public policy issues.
Sociologist David Garland has argued that American and
English politicians and governments came to focus so much
on crime almost by default, and that vigorous anticrime ef-
forts should be seen as expressions of the weakness, not
the strength, of the state (Garland 2001a). Since the 1960s,
Garland argued, external forces have made it difficult for the
state demonstrably to solve its citizens’ problems. The state’s
credibility has suffered as a result. Governments cannot insu-
late their citizens from the disruptive effects of economic
recessions, globalization, and multiculturalism.
But crime is different: “The essential and abiding attrac-
tiveness of the ‘sovereign’ response to crime . . . is that it can
be represented as an immediate, authoritative intervention.
. . . Like the decision to wage war, the decision to inflict
harsh punishment or extend police powers exemplifies the
sovereign mode of state action” (Garland 2001a, pp. 134–45).
Garland explains that evidence—of what works, of foresee-
able effects—is irrelevant because the goal is not crime reduc-
tion per se, but the credibility of the state. Whether policy
initiatives have any effect on crime rates or not is beside the
point. What matters is that policy-makers be seen to be doing
something, and that policies express anger and outrage with
crime, and the people who commit it, and its consequences
for victims (Garland, 2001a, p. 110).
Sociologists Theodore Caplow and Jonathan Simon (1999)
offer two interconnected reasons (among others) why U.S.
crime policy developed as it did. One, paralleling Garland’s
ideas about the weakness of the state, is the anomaly that
enormous expansion over the last thirty years in the role of
government, particularly the federal government, coincided

with sharp and continuous decline in public confidence in
government. The second is the weakening of broad-based
political coalitions and the growth and influence of single-
issue political movements.
The scope of federal government activity has expanded
greatly. Before the 1960s, for example, many subjects now
seen as important federal responsibilities received little fed-
eral attention. These include health care, education, street
crime, consumer protection, occupational safety, employment
practices, child care, environmental protection, the arts, and
discrimination on grounds of age, race, sex, and disability.
Few people any longer argue that any of these subjects is
not the Congress’s business. The enormous expansion of the
federal government’s agenda transformed American politics.
The result, Caplow and Simon (1999) argue, was a spiral
of failure. Declining student performance was attributed to
lack of discipline and intellectual rigor in schools. Illegiti-
macy and chronic poverty were blamed on perverse incen-
tives provided by the main federal income support program,
Aid to Families with Dependent Children. Rising crime was
blamed on lenient judges and parole boards.
As evidence of government failure, Caplow and Simon cite
the massive escalation of costs of federal government pro-
grams between 1970 and 1995 without corresponding bene-
fits. Federal expenditures for health care outpaced inflation
by five to one, for education by three and a half to one, for
Aid to Families with Dependent Children (AFDC) by three
to one, and for criminal justice by six to one. None of these
systems was widely credited as being conspicuously success-
ful, but they are interconnected and their defects are mutually
reinforcing. The extraordinary costs of the health care system
prevented any serious effort to alleviate poverty. The defi-
ciencies of the antipoverty programs undermined the public
schools. The failures of the schools poured into the criminal
justice system.
All of this led, they say, to a remarkable collapse of confi-
dence in government. In response to the survey question,

“How much of the time do you trust the government in
Washington to do the right thing,” 75 percent of a representa-
tive national sample in 1964 answered “just about always”
or “most of the time.” When the same question was put to
a similar sample in 1995, only 25 percent gave those answers
(Caplow and Simon 1999).
For Caplow and Simon, as for Garland, penal policies and
politics are primarily about other things. Government can
express outrage and appear to respond to the public by taking
drastic and dramatic action against crime and criminals. En-
actment of harsh policies serves as an end, not a means.
The inherently expressive character of much crime control
policy distinguishes it from other policy subjects. Govern-
ment and government leaders get credit in most realms for
policies that achieve results, and lose credibility when poli-
cies fail. Announcement of new educational or health care
policies by itself is insufficient. What matters is whether they
satisfy their aims, or demonstrably move in that direction.
In recent years, the expressive content of crime control poli-
cies has been enough.
Caplow and Simon’s second observation is that American
politics after the 1960s moved away from broad-based parties
with traditional class and regional constituencies toward
“single issue” movements. The proportions of the electorate
with strong loyalties to the Republican or Democratic parties
steadily fell and the proportion of voters with no attachment
to either party grew. Elections increasingly are won, or lost,
not because voters support the values or platforms of a party,
but because of the ways they react to media imagery and
particular issues. In an era in which few people believe
strongly in government’s ability to do good, it is difficult to
base political campaigns on broad positive agendas. Some
popular subjects—national defense, antiterrorism programs,
and environmental protection are examples—attract such
wide support that they seldom distinguish parties or candi-
dates.
What is left are candidates’ images and personalities, issues
of overriding importance to single-issue groups, and things
to be against. About images and personalities I have nothing
special to say; candidates try to present themselves favorably
and modern electronic media make those efforts ever more
effective.
The single-issue groups, however, have transformed poli-
tics. Elections are won and lost over such issues as abortion,
affirmative action, gun control, school prayer, gay rights,
and capital punishment. These controversies, Caplow and
Simon (1999) note, do not lend themselves to the compromise
and horse-trading that long characterized American politics.
Single-issue organizations and campaigns invite people to
join the side of good against evil. To antiabortion activists,
abortion is cold-blooded murder. To their adversaries, the
issue is women’s ownership of their own bodies. To advocates
of strict gun control, the private possession of firearms is
foolish and dangerous; to their adversaries it is the keystone
of liberty. To proponents of gay rights, the issues are equality
and human rights; to opponents, traditional values and the
sanctity of the family. School prayer for proponents is about
belief in God and respect for private faith; for opponents,
about cultural hegemony, intolerance, and subtle coercion.
Candidates and parties are loath to take sides in any of these
conflicts. The well-organized pressure groups that represent
such interests have few means of achieving their goals outside
of federal courts, Congress, and the state legislatures. They
want to prevent other people from doing things, or to require
other people to do things, and only courts and lawmakers
have the power to do that. Supporters of these groups will
reliably vote against candidates on the wrong side of the
issue. Only a small minority of voters may care deeply about
the issue, but in a close election they can make the difference.
Faced with voters who split on so many issues and are
profoundly skeptical about the ability of government to im-
prove their lives, parties and politicians have been required
to emphasize policy initiatives that command broad support.
In our time, these initiatives are generally negative. Opposi-
tion to communism and the Soviet Union long played this
role. Early in the twenty-first century, opposition to terror-

ism and other countries’ fundamentalist regimes may be used
in this way, but it will seldom differentiate parties or candi-
dates. What is left are domestic initiatives—such as crime,
welfare, and immigration—that can be debated in moral
terms, respond to broad-based anxieties and empathies, and
affront no powerful constituency.
Like some of the other explanations for American punish-
ment policies, this is plausible and probably valid. Confidence
in government is low and single-issue groups are powerful.
These developments to some extent distinguish American
politics and government from those of other Western coun-
tries. Faith in government is higher in other countries as is
acceptance of the view that government has positive obliga-
tions to improve citizens’ lives. Disagreements about values
and morals exist in every country, but galvanizing single-
issue movements like the National Rifle Association or the
National Abortion Rights Action League are not common.
Anti-immigrant movements are the closest equivalent, but
that is more akin to U.S. anti-immigrant or crime movements
than to American single-issue movements. Nor in most other
countries are issues such as school prayer, capital punish-
ment, and affirmative action the subjects of great contro-
versy.
The question remains, though, why alone among Western
countries has the United States adopted penal policies of un-
precedented severity?
Risk Society
One explanation builds on literatures in sociology and an-

thropology on “risk” and “the risk society.” These literatures
have only recently been applied to crime and crime control
policy. The anthropological literature, most famously associ-
ated with Mary Douglas (1985, 1992), concerns the social
construction of risks and such questions as why in a particu-
lar place and time something is seen as unacceptably risky

and how to understand and evaluate actors’ differing percep-
tions of risk.
The sociological literature is most famously associated with
Anthony Giddens (1990, 1991, 1998) and Ulrich Beck (1992,
1996). Part of the backdrop for each is concern for the envi-
ronmental risks created by modern industry and technology,
and how those risks are assessed at individual and social
levels. Giddens, after noting the secularism and uncertainty
of our time, described contemporary society as a “risk cul-
ture” (1991, p. 3), in which all of life is subject to “contingen-
cies.” “Living in a ‘risk society’ means,” he wrote, “living
with a calculative attitude to the open possibilities of action,
positive and negative, with which, as individuals and glob-
ally, we are confronted in a continuous way in our contempo-
rary social existence” (1991, p. 28). For Giddens, this does
not mean that people living in contemporary societies are
more exposed to risk than people living in other times, or
are more anxious about risk, but are more self-aware.
Beck bridges Douglas’s interest in the social construction
of risk and Giddens’s interest in how the greater fluidity of
contemporary life influences our lives. As social structures
have weakened, and work, family, and gender roles become
less fixed, individuals have more control over their lives,
more choices to make, and less certainty about the future.
He refers to this evolving aspect of contemporary life as “indi-
vidualization” and it is fraught with risk (Beck and Beck-
Gersheim 1995). As with Giddens, there is neither necessarily
more nor worse risk now but greater awareness of it.
Deborah Lupton (1999), in a recent survey of writing on
risk, synthesizes the main themes:
[T]he contemporary obsession with risk has its roots in

the changes inherent in the transformation of societies
from pre-modern to modern and then to late modern (or
postmodern, as some theorists prefer to describe the con-
temporary era). Late or postmodernity generally refers to
broader socioeconomic and political changes that have-

taken place in Western societies since World War II, pro-
ducing the sense for many people that we are ‘living in
new times.’ . . . For the individual, it is argued, these
changes are associated with an intensifying sense of uncer-
tainty, complexity, ambivalence, and disorder, a growing
distrust of social institutions and traditional authorities,
and an increasing awareness of the threats inherent in
everyday life. (Lupton 1999, pp. 10–12)
So far, so good. Stripped of jargon, there is nothing here

that any regular newspaper reader will find surprising or
unfamiliar. What is the relevance of the risk literature for
understanding crime control or punishment policies? Not
much.
Lupton offers three observations about the implications of
risk analysis for crime. First, situational crime prevention
strategies focus on prevention of crime rather than on rehabil-
itation of offenders (1999, p. 94). This is true, but neither
momentous nor new. Situational crime prevention is a com-
plement to law enforcement techniques and not inconsistent
with rehabilitation efforts. The idea is that much crime is
opportunistic and spur-of-the-moment, and it can be reduced
by changing situations to make opportunities less attractive.
Examples include increased street lighting, use of CCTV, re-
design of public spaces, and substitution of paper or elec-
tronic tickets for cash payments. Such initiatives make crimes
more difficult, more likely to be observed, and less likely to
succeed. People have, however, always tried to lessen their
chances of victimization by locking doors, buying dogs, hir-
ing guards, carrying weapons, and avoiding dark alleys at
night. Modern prevention techniques are larger in scale and
more sophisticated in technology but not different in kind.
Situational crime prevention in any case is but a tiny part
of contemporary crime prevention efforts.
Second, emphasis on risk or dangerousness treats offenders
not as individuals but as members of “risk groups” (1999,
pp. 94–95). This also is true, and nothing new. Incapacitation
has always been the flip side of rehabilitation. The converse
of the belief that an offender has been rehabilitated and is
capable of living a crimefree life, and hence should be re-
leased from prison, is that he has not, is not, and should not.
Until recent decades, such decisions were usually based
on individualized assessments of rehabilitation and predic-
tions of risk, but they were intuitive and subjective, and less
reliable than are such decisions when based on validated risk
prediction instruments (Clear and Cadora 2001). Beginning
in the 1920s, sociologists and others began developing base
expectancy tables to predict offenders’, prisoners’, and parol-
ees’ likelihood of future offending (e.g., Burgess 1928; Ohlin
1951; Gottfredson, Wilkins, and Hoffman 1978). These are
actuarial calculations aimed at identifying groups, on the
bases of particular characteristics, whose members have great-
er or lesser probabilities of reoffending. Contemporary risks
and needs assessments, prediction instruments, and targeted
incapacitation policies are applications of that long-estab-
lished technology. Surveillance and punishment policies
based on actuarial calculations do raise questions concerning
whether, when, and within what limits such information may
justly be used in individual cases. There is nothing intrinsi-
cally new about any of this. The relevant conceptual, juris-
prudential, and empirical literatures all date back at least a
half century.
Third, “the figure of the criminal is frequently positioned
as risky and needing exclusion from others” (Lupton 1999,
p. 144). Among the implications of this are that people assess
some places—“the inner city, the shopping mall, housing
estates [public housing]”—where criminals are likely to be
as dangerous, and avoid those places, and that members of
some groups, such as “injecting drug users,” are stereotyped
as potential criminals. All true, and often regrettable, but
the discovery neither needed nor awaited the appearance of
the risk society literature.
Lupton, however, is a specialist in “cultural studies,” not
a crime specialist. The criminologists do no better a job of
demonstrating how risk society analysis aids understanding
of crime policy changes (Hope and Sparks 2000; Stenson and
Sullivan 2001). The best work on the subject, most of it
qualitative, examines the interacting ways objective risk,
anxiety, and fear of crime affect people’s daily lives in differ-
ent settings and circumstances (e.g., Jefferson and Hollway
2000). The worst of it is jargon-ridden and obscure (e.g.,
Stanko 2000), or ideological and polemical (e.g., Stenson
2000).
Social scientists not working in the risk society framework
also worry about increased use of prediction instruments
throughout the criminal justice system. Sir Anthony Bottoms
(1995) includes this as part of a new “managerialism,” and
Malcolm Feeley and Jonathan Simon (1992) portray it as “ac-
tuarial justice.” Both works emphasize increased use of risk
prediction, but neither claims that risk is a new consideration
in criminal justice decision making.
Most of the risk society literature discusses punishment
policies in England and the United States, as if the phenomena
to be explained are the same in both countries. Though most
who write on this subject are based in England, it does not
seem to have occurred to them that American penal policies
are incomparably harsher than the English and imprisonment
rates are five times higher. A theory that applies equally to
both cannot explain those differences and is invalidated by
them.
By this point, the explanation for contemporary American
crime control and punishment policies has gotten pretty com-
plicated. Crime rate levels and changes by themselves do not
have much explanatory power. Rising crime rates and mass
media developments provide a plausible basis for heightened
public concern, but ordinary people’s views are more com-
plex, ambivalent, and temperate than is widely recognized.
Frightening stereotypes and valid fears provide an intelligi-
ble reason why voters respond to “tough-on-crime” rhetoric
and appeals, but do not explain why politicians in our time
chose to campaign on those issues rather than others. The
account of structural changes in American politics is part of
the explanation—there is little support in the early twenty-
first century for ambitious broad-based policy initiatives by
government, and it is always easier to mobilize support
against something than for something. Social, economic, and
cultural changes in American life in the past thirty years
have made life more uncertain and, in the Giddens/Beck
sense, risky.
Like the dots in a pointillist painting, all of these claims
provide points of understanding, but they lack a pattern that
provides an intelligible picture. Three more complex stories
try to do that. One is David Garland’s story, discussed in
the next section, of the force of postmodernist angst. The
second is James Whitman’s explanation for why American
punishment policies and practices differ so greatly from those
in France and Germany. The last, best, and most complete
story is that the pattern can best be understood in relation
to the intertwining of cyclical changes in sensibilities with
recent moral panics and political opportunism.
Postmodernist Angst
This story overlaps the risk society story, but is more com-
plex, comprehensive, and sophisticated. The final third of
the twentieth century was a period of disruption, uncer-
tainty, and change. Few people were unaffected by reces-
sions, globalism, or economic restructuring; nor by social
changes associated with the women’s, civil rights, and gay
rights movements; nor by increased immigration, ethnic di-
versity, and cultural pluralism. Routines and expectations
had to change. Peoples’ lives became less constrained and
predetermined but greater autonomy and wider possibilities
brought greater uncertainty. Values and certainties were un-
dermined and questioned. And, for much of the last thirty
years, crime rates increased and fear of crime penetrated into
more people’s lives, and more deeply. Most crimes, especially
the street and violent crimes that are most feared, are commit-
ted by people at the social and economic margins, and dispro-
portionately by members of racial and ethnic minorities. Be-
cause most of the forces and developments that destabilize

our lives are distant and impersonal, but criminals are near
at hand and identifiable, broad-based anxieties are displaced
onto blamable criminals. Politicians know that government
action cannot fundamentally affect crime rates and patterns
but want to be seen to be doing something, and accordingly
promote and enact policies meant primarily to express disdain
for offenders and hurt them.
“Underlying the debate about crime and punishment,” Gar-
land wrote, “was a fundamental shift in interests” (2001a,
p. 76). Sir Anthony Bottoms of Cambridge University in 1995
famously characterized the sensibilities to which expressive
policies are a putative response as “populist punitiveness.”
He, like Garland, attributes populist punitiveness to in-
creased crime rates and the uncertainties associated with
what he, like Garland, calls “late modernity.” Populist puni-
tiveness exists, Bottoms says, and it explains why govern-
ments adopt harsh policies and why harsh policies are well
received. Garland’s story, however, goes further and explains
both why governments adopted particular policies and why
they won public favor.
Garland’s story, rich, subtle, and multifaceted, is told in
what he calls a “history of the present.” He sketches the
origins of modern criminal justice institutions, traces the
1970s loss of faith in individualized and relatively nonpoliti-
cized practices, describes social and cultural changes, surveys
changes in criminology, crime prevention, and crime control
policies, and then explains why policies changed as they did.
It is an ambitious work, and much the most distinguished
of its kind. What follows is not a review of the entire work
and cannot do it justice, but is an effort to distill central
themes.
Garland’s story has four main components. The first is
the salience of “high crime rates as a social fact” (2001a, p.
106). The second is the use of expressive punishments as
a demonstration of state sovereignty. The third is the risk
society story of instability, change, risk, insecurity, anxi-
ety, and displacement. The fourth is the “criminology of

the other.” I summarize these components briefly, quoting
Garland extensively, rather than paraphrase wordily and
imprecisely.
First, crime has become a “normal social fact,” more widely
distributed, and avoidance of crime a ubiquitous feature of
everyday life:
[B]y the 1970s society’s vulnerability to high rates of crime

came to be viewed for what it was—a normal social fact.
At the end of the 1990s, despite much publicized decreases,
American and British rates of crime and violence remain
at an historically high level. . . . Whatever successes po-
lice and politicians may claim, crime avoidance remains a
prominent organizing principle of every day life. (2001a,
p. 194)
Garland describes how political campaigns and the mass me-

dia amplified awareness of crime. Independently of that, he
argues that victimization is more widespread than in earlier
times. No longer primarily localized in low-income, disadvan-
taged, and minority neighborhoods, victimization has be-
come in a sense more fairly distributed and threatens affluent
and middle-class people.
Second, crime’s status as a new social fact exposed the
“myth that the sovereign state is capable of delivering
‘law and order’ and controlling crime” (2001a, p. 109).
This created a predicament for policy makers. They can
accept and adapt to the limits of sovereign power, or they
can pretend not to know it and retreat into an expressive
mode.
“[T]he essential and abiding attractiveness of the ‘sover-
eign’ response to crime (and above all of retaliatory laws that
create stronger penal sanctions or police powers),” Garland
writes, “is that it can be represented as an immediate, authori-
tative intervention. Such action gives the impression that
something is being done—here, now, swiftly and decisively”
(2001a, pp. 134–35; emphasis in original).

Third, the new social fact of crime, particularly as it im-
pinges on the middle class, importantly exacerbates the
“sense of ontological insecurity” of late modern society:
[T]his new element of precariousness and insecurity is

built into the fabric of everyday life. . . . Little surprise
too that people increasingly demand to know about the
risks to which they are exposed by the criminal justice
system and are increasingly impatient when the system
fails to control “dangerous” individuals. . . . Crime has be-
come one of the threats that the contemporary middle-
class household must take seriously—another problem to
manage, another possibility that must be anticipated and
controlled. (2001a, p. 155)
The vulnerability of the middle class is doubly important.

It provides receptivity to politicians’ populist appeals. Per-
haps as importantly, middle-class prosecutors, judges, proba-
tion officers, and prison officials shared the widespread sense
of precariousness and danger. Previously, these peoples’ so-
cial distance from “the poor people’s problem” of crime, and
their low levels of victimization, made compassion easy to
express and others’ vulgar punitiveness easy to disdain.
Their own sense of personal insecurity, however, made many
“more supportive of punitive responses to crime” (2001a, p.
150) and less committed to the more humane and constructive
policies that in earlier times they supported. A group that
might have been expected to resist expressive policies be-
came less inclined to do so.
Fourth, is Garland’s “criminology of the other, of the threat-
ening outcast, the fearsome stranger, the excluded and the
embittered [which] . . . functions to demonize the criminal,
to act out popular fears, and to promote support for state
punishment” (2001a, p. 137; emphasis in original). The out-
casts in the social and political climates of the 1980s and
1990s inevitably were the welfare poor, urban blacks, and
marginalized working-class boys.

So, putting all this together, why are harsh expressive poli-
cies adopted? “Because,” Garland writes, “the groups most
affected lack political power and are widely regarded as dan-
gerous and undeserving; because the groups least affected
could be reassured that something is being done and lawless-
ness is not tolerated; and because few politicians are willing
to oppose a policy when there is so little political advantage
to be gained by doing so” (2001a, p. 132).
This snapshot of Garland’s argument omits much that is
creative and insightful but does, I believe, capture its flavor.
The important question is whether it adds anything impor-
tantly new to the five stories already told. I think not. The
salience of the new fact of crime is part of every story, as are
the insecurities of “late modernity.” Garland’s sovereignty
argument bears strong resemblance to Jonathan Simon’s
governing-through-crime thesis. The centrality of expres-
sive punishments, and their frequent disengagement from
substantive crime prevention, is common ground. And
along the way, Garland develops, albeit less extensively,
the political cynicism and ideological subplots of the other
stories.
The postmodernist angst story by itself is not enough.
There have been other equally unsettled periods in American
history that produced widespread dislocation and insecurity
but did not produce the kind of sensibilities that led to con-
temporary crime control and punishment policies. It is easy
and not unnatural for people living of a time to think that
time uniquely challenged or troubled, but that is a reflection
merely of chronocentrism, the temporal equivalent of ethno-
centrism. Both where and when we stand, observe, and gen-
eralize shape what we see and think. Anyone who thinks
the three decades 1970–2000 were uniquely unsettling has
forgotten the three decades 1920–1950. That period experi-
enced the rise of Nazism, fascism, and Soviet communism,
as well as the Roaring ’20s, the Great Depression, World War
II, the Iron Curtain, and the onset of the Cold War. Nothing
very dramatic happened to crime control and penal policy

during those thirty years and the American system of indi-
vidualized and indeterminate punishment that had taken
shape by 1930 looked much the same in 1950.
More eloquently and comprehensively than anyone else,
Garland puts all the stories into one mosaic, but the whole
is not greater than the sum of its parts. What is still needed
is an explanation for why the United States adopted crime
control and punishment policies of unprecedented severity,
but neither England nor any other Western country that
experienced comparable crime, social, economic, and cultural
changes and trends did so.
Garland’s story has real power, but ultimately does not
offer a convincing explanation, and he himself seems not
entirely convinced. Having explained how crime trends, so-
cial and cultural changes, recognition of the limits of state
power, and the “criminology of the other” all shaped the
development of current policies, in the end he reduces those
developments to background conditions; they are not singly
or jointly causes. He insists that he does not mean “to imply
that political decisions and policies are determined, or made
inevitable, by events and circumstances occurring elsewhere”
(2001a, p. 139), and that contemporary crime control poli-
cies were “not inevitable” (2001a, pp. 76 and 201). He also
insists that “public attitudes about crime and control are
deeply ambivalent” (2001a, p. 203) and that “[t]he populist
current in contemporary crime policy is, to some extent,
a political posture or tactic, adopted for short-term politi-
cal advantage. As such it can quickly be reversed” (2001a,
p. 172).
Garland’s account of the relationship between postmodern-
ist angst, crime control policies, and punishment practices,
notwithstanding its many merits, does not explain why U.S.
policies are as they are.
Earlier I quoted a passage from Garland about the changing
sensibilities and interests that underlay the construction of
contemporary crime control and punishment policies. But
where do the sensibilities and interests come from? The sev-
enth and eighth stories answer that question.
Democracy and Degradation
The challenge historian James Whitman set for himself was

to understand why “over the last quarter century, America
has shown a systematic drive toward increased harshness by
most measures, while continental Europe has not” (Whitman
2003, p. 38). One explanation is that, “for the most part,
though, American-style politics has failed to exert an Ameri-
can-style influence in German or French criminal justice,”
and, as a result, “bureaucrats have succeeded in keeping
control of the punishment process, without becoming subject
to decisive pressure from a stirred-up public” (2003, p. 15).
That last observation, if warranted, and lots of evidence
suggests it is, powerfully refutes most of the preceding stories
about increasing punitiveness in America. France and Ger-
many have experienced rising crime rates, economic destabi-
lization, more punitive public opinion, ethnic tensions, and
postmodernist angst no less than the United States, and yet
“as of the year 2000, mildness is still, at heart, the watchword
of punishment practices in each country” (p. 70).
Whitman’s description of contemporary differences be-
tween continental Europe and America is not an explana-
tion. For that he goes back two centuries and identifies two
differences between Europe and America that reverberate
in our time. The first is the American distrust of government
and government officials that developed in the fifty-year
build-up to the American Revolution and has characterized
Americans ever since. This he contrasts to the strong Euro-
pean states of the eighteenth century and the continuation
of support for strong states today. Germany and France
have state apparatuses that, compared with American gov-
ernment, are relatively powerful and relatively autono-
mous.
The second difference relates to the alternate ways Euro-
pean and Anglo-American cultures ameliorated the conse-
quences of eighteenth-century status differentiation. Social
hierarchy and status differences were, of course, marked in
Europe, Britain, and the American colonies, and extended to
the ways people who committed crimes were dealt with. Sta-
tus mattered, and it can be illustrated by the forms of execu-
tion and imprisonment. Hangings are unattractive: the victim
slowly asphyxiates, the bowels let loose, and the body wracks
with spasms. Beheadings are nearly instantaneous and the
headless body, if supported, stays in place. In Britain and
Europe, low-status people were hanged; high-status people
were beheaded. And so with prisons: low-status people were
kept in crowded, squalid places, while high-status people
were allowed comfortable facilities to which visitors and per-
sonal servants had access.
Whitman describes starkly different ways that Europe and
Anglo-America responded to the status hierarchy in punish-
ment: Europe leveled up, aspiring to treat all offenders with
respect and civility, at least in principle like high-status pris-
oners; England and American leveled down, treating all pris-
oners in principle like low-status prisoners. In this frame-
work, Whitman argues that much American punishment is
degrading, as low-status punishment has always been de-
grading, while French and German punishment is premised
on treating prisoners with respect and sympathy. Similarly,
American egalitarianism leads to beliefs in uniform and man-
datory punishments, while the European response to historic
anti-egalitarianism leads to beliefs in individualization and
mercy. This is why, Whitman argues, the French regularly
and the Germans occasionally announce broad-based amnes-
ties and commutations, without enraged public outcry, and
it is why the trend in recent decades has been toward greater
mildness in punishments.
Whitman’s two theories, about strong and weak states, and
punishment systems that level up or level down, fit nicely
with contemporary differences in punishment systems. France
and Germany grant autonomous officials substantial author-
ity to individualize mild punishments. America, and to a
lesser extent, England, attempt through legislation to limit
the discretion of officials to deviate from comparatively uni-
form systems of harsh punishments.

Whitman’s analysis is creative and original and genuinely
enriches efforts to understand differences between legal sys-
tems. What it does not do, alas, is explain American policies
of the past quarter century. The difficulty is that three prop-
erties of French and German punishment practice that his
theory explains—moderation in punishment, individualiza-
tion, and mercy—until twenty-five years ago were more
characteristic of the United States than of Europe. Individual-
ization of punishment was the reigning premise of American
sentencing and punishment systems for a hundred years be-
ginning in the 1870s.
Parole is the starkest example. First established in Amer-
ica as a means to let prisoners out when they had become
rehabilitated, parole systems were established in most West-
ern countries, including by 1930 in every American state
(Rothman 1980). American parole boards had broad discre-
tion, at narrowest to release any prisoner who had served a
third of his or her sentence. The extreme cases were California
and Washington State, where every sentence was “one year
to the statutory maximum” and the parole board decided all
release dates. The English parole board was not established
until the late 1960s. Parole systems in most European coun-
tries had only narrow scope: prisoners typically became eligi-
ble for release only after serving half the sentence and had to
be released when they had served two-thirds (Bottomley 1990).
So Whitman’s argument, elegant though it is, fits 1975–
2002, but not 1875–1975. A different explanation is required
that can explain not only why American punishment prac-
tices and policies are as harsh as they are, but why they have
changed so dramatically in a quarter century. The final story
provides that explanation.
Cycles and Sensibilities
Each of the preceding seven stories helps explain the origins

of American crime control and penal policies. However, four

of them—rising crime rates, toughening public opinion, the
emergence of the risk society, and postmodernist angst—
characterize all Western countries and accordingly cannot
explain why U.S. policies are so much harsher than those
elsewhere. The democracy and degradation story offers an
imaginative theory to explain why American punishment
differs so greatly from French and German. Unfortunately,
however plausible it would be if only the past quarter cen-
tury needed explaining, its plausibility disappears in the face
of the preceding hundred years when the United States spear-
headed the movement toward humane and respectful han-
dling of offenders.
The two remaining stories—about cynical politics and
structural changes in governance—are a bit more distinc-
tively American, but not entirely. The particular racial sub-
text of American politics is ours alone, but nearly every West-
ern country has recently experienced nativist, xenophobic,
and anti-immigrant political movements that could have man-
ifested themselves as law-and-order movements. Jonathan
Simon’s account of the disproportionate influence of single-
issue interest groups describes a distinctively American phe-
nomenon, but the larger arguments that he and Garland offer
about governance apply to, at least, all the English-speaking
countries. But none of Australia, England, or Canada has so
far done more than flirt with American-style crime policies
and punishment practices, so those stories also do not explain
why here and not there.
The distinctive and overarching story of cycles, sensibilit-
ies, and moral panics has already been sketched. Sensibilities
associated with the developments described in the other sto-
ries coincided unhappily with a downswing in recurring pat-
terns of tolerance and intolerance of deviance to produce
widespread public susceptibility to calls for adoption of un-
precedentedly repressive policies. The emotional force of
ubiquitous mass media coverage of such events as the crimes
of Willie Horton, the murders of Megan Kanka and Polly
Klaas, and the crack overdose death of Len Bias produced
moral panics that provided occasion for such calls. And
American governmental institutions and political culture
provided many fewer buffers to the force of that emotional-
ism than do the governmental institutions and political cul-
tures of other countries. This large and complicated story is
told in chapter 5. Before that, chapters 3 and 4 provide fuller
accounts of cycles, sensibilities, and moral panics.

3
Cycles and Sensibilities
Capital punishment, indiscriminate private possession of

handguns, and mass imprisonment of black men in early
twenty-first century America will someday be widely de-
plored and deeply regretted. Nearly all Americans regret and
deplore the past subordination of black people in the Ameri-
can South, and yet it was approved and accepted by most
southern whites for at least a century after the Civil War.
The moral, religious, and social policy arguments against
slavery did not change but people’s beliefs and values did,
and now human slavery and deliberate racial or ethnic subju-
gation are, almost literally, unthinkable. In most Western
countries today, adoption of American policies on capital
punishment, handguns, and mass imprisonment is almost as
unthinkable as adoption of slavery would be. Prevailing sen-
sibilities elsewhere attach greater importance to ideas about
human rights and the value of human lives and life chances
than do prevailing American sensibilities.
So why are American crime control policies as they are?
Usually policies toward crime and punishment are portrayed
as the outcome of disagreements between liberals and conser-
vatives, or between advocates of due process and advocates
of public safety, or between people who care about criminals
and people who care about victims. The views caricatured
in the first member of each of those pairs, it is said, predomi-
nated in the 1960s and the views caricatured in the second
in the 1990s. And that is all we need to know to understand
why 140 per 100,000 Americans were in jail or prison in
1970 and over 700 per 100,000 in 2003. That is true in one
important albeit grotesquely oversimplified sense, but it is
63
deeply shortsighted. Policies do, of course, reflect beliefs and
values, but those come from somewhere, and somewhen. If
Americans were more self-aware of why we believe what we
believe, our policies about crime would be very different and
we would not be fated to suffer our descendants’ disapproval.
For some subjects, beliefs and values oscillate in predictable
cycles. Americans have lately been caught up in a period in
which excessive hostility and emotional overreaction toward
criminals and drug users could have been predicted (and
was) and current penal policies are the result.
Truths and Trivia
What we see, think, and believe depends as much on when

we stand as where. This is true of trivial things—people who
were once enamored of bouffant hairdos, polyester pantsuits,
or platform shoes have a hard time imagining that was ever
true and a harder time picturing themselves dressed that
way now. Fashions change, tastes evolve, and few of us are
immune to influence in such matters. What once we liked,
we now dislike, and vice versa.
We are almost as susceptible to influence, however, about
things that are important. Periods of romanticism and classi-
cism in the arts have alternated throughout history, as have
periods of tolerance and intolerance of homosexuality and
religious pluralism, and peoples’ opinions and beliefs have
oscillated with them. And the same thing is true in recent
centuries in relation to punishment, with moralistic ideas
about deserved punishments and moral responsibility alter-
nating with instrumental ideas about problem-solving and
human fallibility.
Beliefs and attitudes about punishment, religious plural-
ism, and sexual tolerance are different in kind from prefer-
ences for polyester pantsuits. Wearing unfashionable clothes
or hairstyles can be done naively or defiantly. It can be gauche
or anachronistic, elicit disapproval or ridicule, and result in

embarrassment or humiliation, but the sanctions for being
different are entirely social.
Being seriously on the wrong side of contemporary views
about religious beliefs can be lethal. Apostates have been
being killed for as long as organized religion has existed
(Erikson 1966; Johnson 1976). From shortly after the death
of Jesus, through the Protestant Reformation and at least
through the hangings of Quakers by Massachusetts Bay Cal-
vinists, Christians have been killing Christians because of
their heretical views. In numerous periods from the first cen-
tury a.d. to the eighteenth, interspersed among more tolerant
eras, Christian heretics were martyred because of their het-
erodox beliefs. As the confidence and strength of the early
Christian church waxed and waned, beliefs that were pun-
ished by death in one century were tolerated and sometimes
celebrated in the next. The radical simplicity and antimateri-
alism of St. Francis of Assisi in the thirteenth century might
well have been intolerable heresy in the twelfth or the four-
teenth. According to historian John Boswell, the “Franciscans
came perilously close to being declared heretical before their
final acceptance by the church” (Boswell 1980, p. 275).
Histories of homosexuality tell similar stories of periods of
tolerance that alternated with periods when the consequences
ranged from ostracism to death (Boswell 1980; Greenberg
1988). Boswell describes classical antiquity and the Early
Middle Ages as periods when homosexual relations were
widely accepted and celebrated, albeit separated by the sixth
through ninth centuries when homosexuality was strongly
disapproved.
Then the pendulum swung: “During the 200 years from
1150 to 1350, homosexual behavior appears to have changed,
in the eyes of the public, from the personal preference of a
prosperous minority, satirized in popular verse, to a danger-
ous, antisocial, and severely sinful aberration” (Boswell 1980,
p. 295). Homosexuality passed from being an uncontroversial
matter of preference or idiosyncrasy in most of Europe to
being a criminal act that most contemporary legal compila-
CYCLES AND SENSIBILITIES 65

tions made punishable by death. Sensibilities changed and,
more importantly, laws changed with them. The change was
lethal, as Boswell demonstrates:
Kings themselves were no longer safe. In the twelfth cen-
tury the king of France could elevate to the episcopate a
man thought to have been his bed partner, and the king
of England [Richard II] could fall head over heels in love
with another monarch [Philip II of France] without losing
support from either the people or the church. But by the
fourteenth century all this had changed, and its opening
decades witnessed first the downfall of the Templars [who
had been accused of sodomy, among other things] at the
hands of Philip IV and then the execution (at the hands
of his daughter Isabella) of the last openly homosexual
monarch of the Middle Ages, Edward II of England. (Bos-
well 1980, p. 298)
The animosity toward homosexuals is made “pellucidly clear,”
wrote Boswell, by the nature of the executions of Edward
and his lover Hugh le Despenser: “Hugh’s genitals were cut
off and burned publicly before he was decapitated, and Ed-
ward was murdered by the insertion into his anus of a red-
hot poker” (Boswell 1980, p. 300).
Comparative tolerance is typical now in Western countries,
though far from complete. Memories of Oscar Wilde’s final,
sad years and of intolerance in some parts of the contempo-
rary world remind us that things are often otherwise.
Being on the wrong side of contemporary views about pun-
ishment can also be fatal, or life-diminishing, but in a differ-
ent way and to different people. Being homosexual or sub-
scribing to particular religious beliefs is part of who people
are or want to be. In tolerant times, those are Good Things,
and in intolerant times, perilously they are not.
The central question concerning crime and punishment is
not whether crime is a good thing—not whether people
should be entitled to be criminals in the same way they are
free in tolerant times to be homosexual—but how the state

should punish criminals. The prevailing ethos can have dras-
tic consequences for people convicted of crimes, and include
differences between life and death or between prison sen-
tences measured in years or in decades. In earlier centuries,
Michel Foucault (1977) and many others relate, breathtak-
ingly cruel punishments were sometimes practiced. Foucault
famously began Discipline and Punish with a description of
the 1757 execution of Damien, condemned to have the flesh
torn from his breasts, thighs, arms, and calves by red hot
pincers, his hand burnt with sulfur, and the wounds dressed
with a blend of molten lead, boiling oil, burning resin, wax,
and sulfur. He was then to have his body drawn and quar-
tered by four horses.
That time has passed, but prevailing attitudes toward
punishment continue to change with the times. Thousands
of executions occurred in the United States in the early
twentieth century, but steadily declined in the 1940s, 1950s,
and 1960s. In more recent times, an American offender’s
chances of execution are vastly higher early in the twenty-
first century than they were in the 1960s. During the 1970s
and 1980s, the average lengths of prison sentences for seri-
ous crimes more than tripled, and then increased some more
in the 1990s.
Even a few years can make an enormous difference in pun-
ishments imposed and suffered for particular crimes. Young
men convicted of draft evasion late in the Vietnam War often
received the maximum lawful prison sentence of five years.
Within a few years, a short prison term or probation became
typical and shortly after that an amnesty was enacted. The
same behaviors, occurring at about the same time, incurred
vastly different consequences depending on when the sen-
tence was imposed or the case dealt with.
There are many other examples of punishments varying
greatly depending solely on when they were imposed. Before
the U.S. Supreme Court declared capital punishment for rape
unconstitutional under the Eighth Amendment in Coker v.
Georgia, 433 U.S. 584 (1977), many actual or alleged rapists,

especially black men accused of raping white women, were
executed.
As tolerance of drug use and Drug War fervor fluctuate,
authorized and imposed sentences for drug crimes vary radi-
cally. The U.S. Congress enacted mandatory minimum sen-
tences for drug crimes in the 1950s and 1960s, then in 1970
at the urging of, among others, Republican Congressman
George H. Bush, repealed nearly all of them, because they
were seen to be too rigid and too harsh. In the 1980s and
early 1990s, at the urging of, among others, Republican Vice
President and later President George H. Bush, the U.S. Con-
gress enacted new mandatory minimum sentences of unprec-
edented severity for drug crimes. People involved in 500-
gram heroine transactions could receive very different pun-
ishments depending on whether the transaction occurred
in the 1970s, the 1980s, or the 1990s. Closer to many more
people’s homes, because of the commonality of use or exper-
imentation with marijuana, penalties for small-scale traf-
ficking in marijuana have fluctuated nearly as widely as
for cocaine trafficking. Marijuana offenses punishable by
multi-year prison sentences in the 1960s in many places
were largely ignored in the 1970s. In 2001 handling of mari-
juana cases is wildly inconsistent, ranging from de facto
decriminalization and widespread open use in some parts
of the country to inconsistent but sometimes harsh handling
elsewhere.
Punishment practices and policies change for reasons, but
the quality of the reasons may be good or bad. It is important
to be sure that the reasons why punishment policies are to be
made harsher are sound, and that they result from something
other than raw emotion or short-term upset. Recent toughen-
ing of penalties and policies concerning domestic violence
or child abuse, for example, can probably pass that test. Both
behaviors fundamentally damage victims’ lives and neither
appears before to have received sustained attention from the
criminal justice system. So might penalty increases be justi-
fied for particular crimes, corporate and white-collar crimes,
for example, for which reliable evidence gives reason to be-
lieve that ethically defensible increases would measurably
prevent future victimization. When, however, the reason is
short-term overreaction or political expediency or animus to
minority groups, the test is failed.
Stark shifts in social practices, including punishment, occur
because many or most people in a time and place share percep-
tions and beliefs that justify them, unmindful or indifferent
that their perceptions and beliefs may be time-bound, and
that they themselves in a few years or decades may see them
to have been wrong. In the abstract, we understand that things
change. That’s life. In the particular, individuals’ lives are
diminished, shortened, and taken not only because of who
they are or what they believed or did but also because of
when they are, or believe, or do.
Aristophanes’ “The Frogs” tells of two boys playing near
a stream who find a frog and, for fun, kill it. They killed
the frog in jest, Aristophanes reports, but the frog died in
earnest. People whose lives are diminished, ruined, or taken
because of the penal ethos of the era in which they live are
like Aristophanes’ frog. Impersonal forces of history or the
penal policies of a moment may explain what happened to
them, but their suffering and loss are real.
It would be easy but wrong to think that cruel punishments
are primarily of antiquarian interest, that we are more civi-
lized now, that we do not allow our collective judgment to
be overwhelmed by passing emotion, or that we do not do
things to offenders that we or our descendants will later
regret. There is no reason to suppose that people now in
positions of power are any less susceptible to the ethos of
our times than our ancestors were to theirs. The United States
in the final decade of the twentieth century was in the throes
of a cyclical period of extreme intolerance of drug users and
people who commit crimes. Many Americans early in the
twenty-first century have a hard time recognizing the cruelty
of current drug and crime policies, because they have become
used to them, but most people in earlier periods of intolerance
were equally unaware of injustices that later generations con-
demned.

Sensibilities
Sensibilities are time- and place-bound ways of thinking that

include ideas and express values that are widely shared and
little questioned. Sometimes sensibilities change slowly and
sometimes they change rapidly, but when a particular sensi-
bility is widespread it influences what people think, say, and
believe.
Sensibility is a central idea in this book, so I devote several
pages to explaining what I mean by it. Two influential con-
ceptions of sensibility have recently received attention in
writings on crime and punishment. For English penologist
Nigel Walker, sensibilities are characteristics of individuals
that make them more or less susceptible to different kinds of
pain (and pleasure). For sociologists Norbert Elias and David
Garland, sensibilities are characteristics of societies and are
the products of long-term economic, social, and psychological
developments that shape the values and beliefs of members
of a society.
By sensibility, I mean the ethos or zeitgeist of a moment
that influences but does not determine what most people
think and believe about a particular subject. It differs from
Walker’s conception because it refers to something social and
collective rather than idiosyncratic and individual, and it
differs from Elias’s and Garland’s because it is more contin-
gent, and more susceptible to conscious reflection and recon-
sideration.
Nigel Walker (1991) uses the term to describe the unique
characteristics of individuals that cause them to experience
punishments differently. He points out that a serious effort
to impose comparable punishments on like-situated offenders
would take account of offenders’ diverse sensibilities and
adjust punishments accordingly. Different offenders, Walker
argues, for example, will experience imprisonment in differ-
ent ways. For an active Los Angeles gang member, a three-
year sentence may be a rite of passage. For a single man in
his twenties, not otherwise involved in crime, it may be
unpleasant and disruptive. For a young single-parent mother,

it may be soul-destroying. For an employed middle-aged head
of household, it may be the destruction of a family and ruin-
ation of a life. For a sickly 75-year-old, it may be a death
sentence. For an effeminate teenage boy, or a claustrophobe,
it may be sheer terror. Three years in prison for one person
often will not result in the same kind or amount of suffering
as it does for another. To pretend otherwise is to ignore basic
differences between human beings, to fail to take differences
between individuals seriously. Walker makes these points
not prescriptively, but argumentatively, to demonstrate that
punishment theories or policies that call for equal punish-
ments for broad categories of crimes or criminals are psycho-
logically and ethically misconceived. In Walker’s sense, sen-
sibilities differ between individuals and, for individuals, over
time. For Elias and Garland, sensibilities differ between socie-
ties and, within societies, over time.
Modern attention to the nature, course, and power of sensi-
bilities at societal levels derives from the work of the late
German sociologist Norbert Elias (1978 [orig. 1939], 1982 [orig.
1939]). For Elias and for David Garland, a Scottish sociologist
now at New York University, who is partly responsible for
revived interest in Elias’s work, sensibilities are characteris-
tics of societies. They are the prevailing ethoses of time-
bound societies and shape the beliefs and values of individu-
als. Elias used the idea of sensibilities to explain changes in
societies’ norms and practices, and Garland uses it to explain
contemporary changes in penal policy. Societies differ in
their cultural, historical, material, religious, and social cir-
cumstances, and these differences shape the prevailing sensi-
bilities.
Elias, in The Civilizing Process (1939), connects changes in
sensibility and individual psychology with wider changes
in social, economic, and geopolitical organization. These
changes, as Garland summarizes:
eventually have an effect on the psychic organization of

the individuals involved and, in particular, on the struc-
ture of their drives and emotions. . . . To the extent that

this process of socialization is successful, the emotions and
behavior of the individual become more evenly ordered,
less spontaneous, and less given to wild oscillation be-
tween extremes. Individuals are thus trained and psycho-
logically equipped to sustain social conventions and to
display a particular pattern of sensibility. (1991, pp. 144–45)
For Elias, some changes in sensibilities have been more or

less linear and together make up what he calls the “civilizing
process.” Operating through changing sensibilities, the civi-
lizing process has been underway since the late Middle Ages.
It made the sight of brutality, suffering, pain, and violence
unpleasant and unappealing. These things became increas-
ingly unacceptable or, at least, were moved out of sight. In
our time, they have been stylized into the seemingly-real-
but-understood-to-be-unreal cartoons of modern mass enter-
tainment and thereby detached from day-to-day life (Bettel-
heim 1977).
Elias described “typical civilization curves.” One example
is the history of meat preparation and eating: “The curve run-
ning from the carving of a large part of the animal or even the
whole animal at table, through the advance in the threshold of
repugnance at the sight of dead animals, to the removal of
carving to the specialized enclaves behind the scenes, is a
typical civilization curve” (Elias 1978, p. 121). Sixty years after
Elias wrote, most people buy meat in tidy plastic-wrapped
packages even further removed from the visceral realities of
killing and dismembering. Increasing numbers of people in
Western countries, more in Europe than in North America,
have become vegetarians, and their numbers will grow, fur-
ther extending Elias’s meat-eating civilization curve.
Capital and corporal punishments have followed a similar
pattern. Albert Camus, in Reflections on the Guillotine (1960),
chronicled the gradual evolution of capital punishment in
France from drawing-and-quartering in a public square pre-
ceded by painful embellishments like disembowelment, suc-
cessively to strangling, hanging, guillotining in public, and
guillotining behind prison doors. Camus explained the se-
quence in terms of the increasing revulsion people felt about
killing human beings. Only by moving the corporal reality
and mess behind doors and out of sight could the killing
continue. Later, after Camus wrote, French sensibilities no
longer tolerated even secretive executions and capital pun-
ishment stopped.
The implicit march toward increased decencies in Elias’s
civilization process may or may not be overly optimistic.
That sensibilities in Elias’s sense exist, change over time,
and influence individuals is, however, incontrovertible. Our
beliefs, preferences, tastes, and values are shaped by the
culture and society in which we live. These differ and change.
Those differences are much of what makes us who we are.
Just as it is easy to forget that we speak prose, it is easy to
overlook basic matters about which most people in a time
and place unthinkingly agree.
The relevance of Elias’s work for thinking about punish-
ment and penal policy is demonstrated in David Garland’s
1990 book Punishment and Modern Society. Garland showed
that efforts to understand penal policy that incorporate only
normative analyses and arguments about practical effects are
incomplete. They must also take account of the social func-
tions punishment serves.
In his later book, The Culture of Control: Crime and Social
Order in Contemporary Society (2001a), Garland attributes re-
cent trends in punishment policies and practices to changing
sensibilities associated with the emergence of economic and
social changes that he and others encapsulate in the term
“late modernity.” Garland tries to show how secular changes
have shaped sensibilities. Changes in crime rates and patterns
are not irrelevant to penal policy changes, but larger social
forces do the heavy lifting.
Garland’s relevance here is that he offers a sophisticated
account of the role of sensibilities in the making and imple-
mentation of penal policy. Sensibilities as Elias described
them are shaped by the kinds of social changes Garland
describes, and in turn shape how we think and what we
believe.
Note the verb: shape. Close synonyms might include influ-
ence, conduce, and predispose, but not determine, cause,
or foreordain. Sensibilities exist, and they change, but only
rarely do they determine. They are more malleable than is
implied by Elias’s and Garland’s functionalist notions of
deeply bedded and evolving social norms that shape human
psychologies.
Elias and Garland do not allow for regular alterations in
sensibilities, but some at least do alternate, particularly in
relation to behaviors widely regarded in a place and time as
deviant. Nor do they treat sensibilities as the proper object
of self-conscious assessment. If men and women are thinking
animals, we can search for and appraise the sensibilities that
shape our thought, and reject, or work to constrain, those
we disapprove or recognize to be destructive.
Evolutions in sensibilities can be seen in the United States
in relation to capital punishment, not exactly in parallel to
Camus’ story about France, but with similarities (and with
the crucial difference that the United States resumed use of
capital punishment with enthusiasm). As in France, execu-
tions were moved from public squares to prisons, and meth-
ods of dispatch shifted from the cruel and messy to the hy-
gienic and painless—from hanging, shooting, and electrocution
to death by lethal injection. Many punishments—maiming,
branding, whipping, torturing, beheading, and burning at
the stake—are, for most people, as unthinkable in the United
States as they are in other Western countries.
American constitutional law doctrine about the death pen-
alty has, until recently, followed a course that resembles a
weak and crabbed version of Elias’s civilizing process hy-
pothesis and demonstrates changing penal sensibilities. The
U.S. Supreme Court resorts to consideration of “evolving
standards of decency” to decide whether particular punish-
ments are unconstitutional under the U.S. Constitution’s
Eighth Amendment prohibition of “cruel and unusual pun-
ishment.” Several Supreme Court Justices, most prominently
Thurgood Marshall and William Brennan, argued that evolv-
ing standards of decency by the early 1970s were no longer
compatible with the death penalty and urged that it be de-
clared unconstitutional. Supreme Court Justices Harry Black-
mun and Lewis Powell expressed the same view after their
retirements.
For a time in the 1960s and early 1970s, it appeared possible
that a majority of the Supreme Court might agree. That did
not happen, but the Court did decide in Coker that evolving
standards forbade capital punishment for rape and by impli-
cation for all lesser crimes of sex, violence, and property. In
the June 2002 case of Daryl Atkins, a Virginia man with an
IQ of fifty-nine, the Supreme Court decided that a “national
consensus” now exists that mentally retarded offenders are
categorically less culpable than the average offender and for
that reason cannot constitutionally be executed. The Court
usually looks for mundane empirical evidence on which to
base a conclusion about evolving standards, usually by
counting up the number of U.S. states that do or do not
permit a practice. What saved Atkins, unlike the many re-
tarded persons who have been executed since the court de-
cided Penry v. Lynaugh, 492 U.S. 302 (1989), upholding the
practice of executing mentally retarded offenders, is that
states have been changing their laws. When Penry was de-
cided, two states forbade such executions and fourteen states’
laws did not authorize capital punishment at all. By the time
Atkins v. Virginia, 122 S. Ct. 2242 (2002), was decided, eigh-
teen states forbade such executions and twelve states plus
the District of Columbia did not authorize any capital punish-
ment. Thus, thirty states, a majority, do not execute retarded
offenders. Standards of decency, in the Court’s view, have
changed enough that offenders such as Atkins should be
allowed to live. A counting exercise such as that is an odd
way to decide what is ultimately a moral or human rights
issue, but the sequence of cases does illustrate an evolution-
ary process.
Developments regarding capital punishment illustrate a ba-
sic difference between the United States and other Western
countries. Something like Elias’s civilizing process appears
to have operated in every country, and capital punishment is
now almost everywhere forbidden. For the past thirty years,
however, the United States has moved in a different direction.
Why? Because American constitutional and governmental
arrangements provide less insulation from the effects of cy-
cles of intolerance. Xenophobia, animus towards immigrants
and ethnic minorities, and harsher attitudes toward crime
and disorder have characterized many countries in recent
years. In some, including the United States, opinion surveys
show majority support for the death penalty and harsher
crime policies. In other countries, however, crime policy is
not a central feature of partisan politics and electioneering
and judges and prosecutors are either civil servant or ap-
pointed in nonpartisan ways. Officials setting policy can gen-
erally devote most of their attention to substantive matters
and comparatively little to the results of public opinion sur-
veys.
By contrast, crime has been a partisan political issue in the
United States for nearly forty years, and most prosecutors and
judges are elected or appointed in openly political processes.
Crime control features, often polemically, in executive and
legislative elections, and policies are often heavily influenced
by measures of public opinion, however well or poorly in-
formed that opinion may be. American governmental arrange-
ments, when public opinion turns harsh and mean-spirited,
rather than ameliorate the effects of public emotion, often
exacerbate them.
The important point here, though, is that something like
sensibilities exists, whatever we call them, and they influence
what we say and do and think. But there are three important
qualifications.
First, almost never is their influence monolithic. No matter
what ways the sensibility winds are blowing, there are almost
always people who adopt contrarian stances: who opposed
slavery in eighteenth-century Virginia, promoted female en-
franchisement in the nineteenth century, opposed Japanese
internment during World War II, and favored decriminaliza-
tion of drugs in the 1980s.

Second, some normative disagreements are so deeply rooted
that they are not especially susceptible to changes, at least
short-term changes, in sensibilities. Examples in our time
concern abortion, capital punishment, and gun control. The
proportions of the population subscribing to one or the other
side of these controversies ebb and flow, but large numbers
of people hold to their views no matter which is ascendant.
Third, most importantly, some ideas are timeless. Most ba-
sic Western ideas about law, ethics, and morality are exam-
ined in Aristotle’s Nicomachean Ethics and Plato’s Republic
and Laws. Concepts such as virtue, justice, fairness, freedom,
and equality go in and out of fashion, and conceptions
change, but they are seldom completely forgotten. Sometimes
they are forgotten or overlooked by many people and are
omitted from the orthodoxy of the day. Sometimes they are
or imply inconvenient truths that are dangerous to express,
and many people are afraid to express them. Sometimes cogni-
tive dissonance, the human tendency to rationalize and accept
what we cannot change, operates and many people come to
believe what is convenient or safe.
It became fashionable in the 1980s and 1990s in the litera-
tures of postmodernism and multiculturalism to denigrate
such notions as “timeless ideas.” There is something, of course,
to be said for the multicultural argument. Non-Western socie-
ties have different histories and cultural traditions than those
shared by Western countries and ideas like “freedom” and
“individual rights” may have different genealogies and cul-
tural meanings. Within Western countries, though, the no-
tion of timeless ideas is coherent and meaningful. Ideas such
as liberty, tolerance, and justice are widely shared. They do
not exist in nature as does a rock or a tree, but they are
durable social constructions in Western culture and are at
the hearts of documents such as the French Declaration of
the Rights of Man, the U.S. Bill of Rights, the United Nations
Declaration of Human Rights, and the European Convention
of Human Rights. People who want to argue that all cultural
meanings are contingent and hence time- and place-bound

may be right in a trivial, reductionist sense, but as a practical
matter they are wrong. So I use the term “timeless idea”
understanding that it is inaccurate in a geological sense, but
believing it a useful shorthand in our time, place, and culture
for core cultural ideas that are widely shared and have been
for centuries.
Some people always hold onto the timeless ethical ideas.
Erasmus understood that Protestant persecution of apostates,
sometimes unto death, was wrong after the Reformation, and
no less wrong than were Roman Catholic persecutions of Jews
and heretics during the Inquisition, even if the prevailing
sensibilities of the time made that a dangerous view to express
(Johnson 1976). George Washington and Thomas Jefferson
knew that slavery was wrong, as did many white American
southerners before and after the Civil War, even if Washing-
ton freed his slaves only at his death and Jefferson never
did (Ellis 2002). There were people who opposed Jim Crow
laws, the Red Scares of the 1920s, and McCarthyism, and
sometimes spoke out about it, even at danger to their own
lives and interests.
There are also mundane tastes and preferences that vary
with time and place and are the opposite of timeless. Some
people did not like Rudy Vallee, the Beatles, or The Fellowship
of the Ring. It is difficult to devise timeless analyses of the
merits or demerits of Nehru jackets, stiletto heels, gangsta
rap, 100 percent cotton, or the color mauve. Their appeal
for some people depends on personal idiosyncrasy, and for
others on what most or a sizable number of people in the
larger society or a smaller subgroup think at a particular
time. Because we are social animals, we are influenced by
social influences and prevailing opinions.
The trick, as a people, is to achieve clarity on which sub-
jects, tastes, and preferences are timeless and which are
ephemeral. The former matter, and we need always to keep
them in sight, even though that is not easy. As to the latter,
we can comfortably let a thousand flowers bloom, and revel
in their gaudy impermanence and exotic scents, knowing
that others equally alluring will soon blossom in their places.
Cycles
Historians have shown how tastes and perceptions change

over time, and how the beliefs and values of people change
with them. Jaroslav Pelikan, in Jesus through the Centuries
(1985), showed how each era reinvents or reinterprets the
historical Jesus: in some eras he is the Messiah, sometimes
a social and political reformer, sometimes the founder of the
church, sometimes a charismatic leader, sometimes a spiritual
ascetic. The historic Jesus was who he was, but people in
different times saw what they wanted or needed to see. Over
shorter periods, historians have documented similar meta-
morphoses in the characterization of such prominent figures
in American history as Washington, Jefferson, and Lincoln.
Pelikan, unlike Elias, does not describe a perceptible if
sometimes interrupted march of progress, but a series of con-
ceptions of Jesus that reflected the needs, beliefs, and sensi-
bilities of successive times. Nor does Pelikan describe cyclical
patterns of sensibility. Sensibilities do, however, fluctuate
and cycles of taste and belief recur. In one representative
realm, literature and the arts, Grant Gilmore described typical
fluctuations in aesthetic preference, taste, and belief:
We have become used to the idea that, in literature and

the arts, there are alternating rhythms of classicism and
romanticism. During classical periods, which are, typi-
cally, of short duration, everything is neat, tidy and logi-
cal; theorists and critics reign supreme; formal rules of
structure and composition are stated to the general ac-
claim. During classical periods, which are, among other
things, extremely dull, it seems that nothing interesting
is ever going to happen again. But the classical aesthetic,
once it has been formulated, regularly breaks down in a
protracted romantic agony. The romantics spurn the ex-
quisitely stated rules of the preceding period; they experi-
ment, they churn around in an ecstasy of self-expression.
At the height of a romantic period, everything is confused,
sprawling, formless and chaotic—as well as, frequently,

extremely interesting. Then, the romantic energy having
spent itself, there is a new classical reformulation—and
so the rhythms continue. (Gilmore 1974, p. 102)
Historian David Musto has described similar cycles in

American attitudes toward alcohol and drug use (1973, 1987a,
1987b). At least three times since the beginning of the nine-
teenth century, the United States moved from periods of
widespread, tolerated recreational use of alcohol and drugs
to puritanical periods of uncompromising prohibition. The
first period of extreme intolerance began in the 1820s and
culminated in prohibition of alcohol in eight states by the
1850s (Gusfield 1963). The temperance movement of the late
nineteenth century led to national Prohibition; more general-
ized intolerance of drug use and users produced the first
major federal narcotics laws, the Harrison Act of 1914 and
the Marijuana Tax Act of 1937. The first signs of the contem-
porary period of extreme intolerance appeared around 1970,
when the Nixon administration declared its war on drugs.
Those attitudes toward drug use took firm hold in the early
1980s and the United States remains in their grip.
According to Musto, peoples’ normative beliefs about drug
use vary with successive phases of cycles of tolerance and
intolerance of drug use. Live-and-let-live attitudes prevail
in periods of relative tolerance, like the 1870s–1880s and
the 1950s–1960s. In the late nineteenth and early twentieth
centuries, for example, cocaine and opium (and derivatives)
were widely used in patent medicines, most addicts were
conventional, law-abiding people, predominantly women,
and cocaine was widely seen as a harmless recreational drug.
Conan Doyle’s Sherlock Holmes was a regular user of cocaine
but that made neither the author nor the character notorious.
In the 1960s, marijuana was widely and openly used; it
and many hallucinogens were seen by many as recreational
drugs less harmful than alcohol. Numerous local jurisdictions,
including prominently Ann Arbor, Michigan, and Berkeley,
California, effectively decriminalized use, possession, and
small-scale trafficking of marijuana. Alfred Blumstein (1993)
showed how the realization that white middle-class kids were
getting criminal records and sometimes going to prisons for
minor drug offenses led to a drastic reduction in drug arrests
of young whites in the early 1970s. President Carter in 1977
called for federal decriminalization of private marijuana pos-
session. A few years earlier Peter Bourne (not then, but later
Carter’s primary drug policy advisor; now he would be called
the “drug czar”) wrote that cocaine “is probably the most
benign of illicit drugs currently in widespread use. At least
as strong a case could be made for legalizing it as for legalizing
marijuana” (Musto 1987a, p. 265).
Conversely, during periods of extreme intolerance, recre-
ational drug issues tend to be seen in black and white, and
adoption of policies of unrelenting vigor is not uncommon.
Liberal Republican Governor Nelson Rockefeller of New York,
for example, in 1972 proposed and won adoption of the
Rockefeller Drug Laws that established long mandatory mini-
mum sentences for drug offenses, forbade plea-negotiating
prosecutors from waiving prison sentences, and paid for ad-
ditional courts and judges to implement the new laws. Every
American state during the 1970s and 1980s adopted manda-
tory minimum sentence laws for drug crimes. The U.S. Con-
gress in 1986 enacted the harshest drug sentencing laws in
U.S. history, including the 100-to-1 law that punishes crack
cocaine trafficking, then as now primarily committed by
blacks, as harshly as powder cocaine offenses, then as now
primarily committed by whites, 100 times larger. In 1988,
legislation was enacted providing for appointment of a fed-
eral drug czar to lead the war on drugs. By 1989 the per
capita drug arrest rate had risen by a factor of six in ten
years, and by 2000, the number of people held in federal
and state prisons and jails for drug offenses had climbed to
324,489, up 1,370 percent from 1980 (when there were
23,900). In 1994 Congress authorized the death penalty for
various drug-related crimes. At various times during the
1980s and 1990s, Americans responding to opinion polls
named drug abuse as or among America’s most pressing prob-
lems.
Musto has several times written of the seeming anomaly
that prohibitionistic sentiments become strongest and drug
policies harshest after drug use has begun to decline (e.g.,
1987a, b). At various periods, drug use comes into vogue, use
increases, and dangers of abusive use become more evident.
Eventually, as Daniel Kagan has summarized Musto’s argu-
ment:
The trend reverses; drug use starts to decline faster and
faster. Public opinion turns against drugs and their accept-
ability begins to evaporate. Gradually, drug use becomes
associated, truthfully or not, with the lower ranks of soci-
ety, and often with racial and ethnic groups that are feared
or despised by the middle class. Drugs become seen as
deviant and dangerous and become a potent symbol of
evil. Trailing behind this decline come large-scale legisla-
tive and law enforcement efforts . . . aimed at curtailing
drug sales and use through energetic prohibition and en-
forcement and ever-harsher punishments against sellers
and users. During this period, public opprobrium intensi-
fies into outright fear, hatred of drug dealers and users,
and a burning anger and intolerance toward anyone and
anything associated with drug use. (Kagan 1989, p. 7)
During periods of relative tolerance, people feel comfort-
able invoking traditional American notions of individualism
and personal autonomy to justify making their own choices
about drug use. Drug use is widely seen as only mildly devi-
ant or not deviant at all. People feel comfortable arguing on
the merits for the benefits and pleasures of drug use, for
individuals’ rights to make their own choices, and against
state intrusion into those choices.
In periods of extreme intolerance, drug use is widely seen
as deviant. Few people feel comfortable risking moral disap-
proval or stigmatization by arguing in favor of drug use or
tolerance of drug users. Musto notes that, “in the decline
phase of drug use . . . we tend to have an overkill, that is
to say people become so righteous and so zealous that we
can have excesses in the name of fighting drugs. There is
very little opposition to draconian policies because no one
wants to stand up for using drugs“ (1987b, p. 43).
The important thing about the Gilmore and Musto ac-
counts, for my purposes, is that they tie sensibilities and
beliefs to recurring segments of cyclical patterns. Gilmore
observes that, during classical periods, rules of composition
and structure are stated “to the general acclaim.” Formalism,
in other words, is not foisted upon a society or an era, but
embodies aesthetic qualities that people value, or believe they
value, which is almost the same thing. Romantic aesthetic
values likewise become predominant in a period not because
they are compelled but because they are celebrated. Musto
writes that people “become so righteous and so zealous” dur-
ing the decline phase of drug use that they adopt views and
support policies that most people in other times would see
as draconic. People behave like this not because they are
coerced or cajoled into it but because that is how they feel.
It is a truism that our upbringings, past experiences, mate-
rial circumstances, and immediate environments shape the
way we see and understand the world. People tend not to
be conservative until they have things to conserve, and so
the frequent pattern that youthful radicals become middle-
aged moderates and elderly conservatives. Or, the somewhat
different Chinese developmental stereotype that people are
Confucianist while young, Realist in maturity, and Taoist
when old. And history is full of stories of whole societies
overwhelmed by fear, insecurity, or anger to do or tolerate
horrible things that later appear unthinkable.
It is one thing, however, to acknowledge that our character-
istics, interests, and experiences influence our wants and
beliefs, and quite another to urge, with some of the sillier
postmodernists, that what we see depends solely on where
we stand. Since one standpoint is as good as any other, it is
said, we can not see anything (or we can see many things,
but no perception is any more “privileged” or any less “con-
tingent” than any other). That may be a plausible position
to adopt in relation to aesthetic assessments of bellbottoms,
bouffant hairdos, or gangsta rap but not in relation to funda-
mental normative issues over which human beings have puz-
zled for millennia.
Even during periods of intolerance, some people recognize
cruelty and excess. Thoughtful people have always recog-
nized the polar possibilities in the value-laden domains of
religion, sexuality, and criminal punishment, the alternatives
that lie between them, the practical and principled arguments
that can be made for one position or another, and the inexor-
ability of the processes by which one set of beliefs displaces
another. Unfortunately, all that tends to get forgotten by
most people in the heat of a historical moment.
Human beings have been round the bend and back enough
times that we should be able, by learning from history, to
escape being condemned to repeat at least its worst excesses.
Witches and heretics have burned sufficiently often, to later
regret, that there is no need to burn them again in order to
be able again to regret it. That lesson should have been learnt
for good by now and, at least in most Western countries,
probably has been. Similar lessons can and will be learned
about tolerance of human sexual diversity, experimentation
with and use of intoxicants, and punishment of wrongdoers.

4
Moral Panics and “Windows of Opportunity”
Historians teach that crime rates and patterns change

slowly, in accord with regular long-term trends, that sensi-
bilities generally change with them, and that some sensibili-
ties fluctuate in accord with cycles of tolerance and intoler-
ance. Sociologists have shown that public attitudes and
opinions about crime sometimes change rapidly and deci-
sively. The interaction between long-term cycles and rapid
changes in attitudes goes a long way toward explaining
American crime control and punishment policies of the past
thirty years.
During the past half century, historical sociologists and
sociologists of deviance have conceptualized a phenomenon
called a “moral panic” that helps us understand times when
public passions take over and produce decisions, policies,
and behaviors that might not otherwise have happened.
Moral panic theory offers insights into the development of
contemporary crime and drug policies, and the politics that
surround them.
The term was coined and the concept popularized by Brit-
ish sociologist Stanley Cohen (1972). He was trying to explain
how an unruly weekend in an English seaside resort in 1964
developed into a national crime crisis that, in turn, led to
overwrought newspaper headlines, widespread anxieties,
police abuse of authority, poorly considered legislation, and
exemplary punishments. Stuart Hall, another British sociolo-
gist, used the concept to explain what he saw as overreaction
to an outbreak of London street muggings in the 1970s. Here
is Hall’s description of a moral panic:
85
[W]hen the official reaction to a person, group of persons,
or series of events is out of all proportion to the actual
threat offered, when “experts” perceive the threat in all
but identical terms, and appear to talk “with one voice”
of rates, diagnoses, prognoses, and solutions, when the
media representations universally stress “sudden and dra-
matic” increases (in numbers involved or events) and
“novelty,” above and beyond that which a sober, realistic
appraisal could sustain. (Hall et al. 1978, p. 16)
For Cohen and Hall, a moral panic was something negative,

irrational, and regrettable; shocking or frightening inci-
dents occurred, raw emotions took over, fears magnified,
panic set in, inhibitions weakened, and public officials over-
reacted. Whether, however, the result of a moral panic is
negative and to be regretted depends on its nature and
result, and who makes the assessment. The term is usually
used by activists and the media and the use is generally
negative.
Some venerable earlier sociological works can also be reread
as accounts of moral panics expressed in other language. Out
of chronological sequence, I discuss three seminal writings:
Cohen’s “Clacton riots,” Kai Erikson’s classic 1966 explana-
tion of the Salem witch trials, and Edwin Sutherland’s land-
mark 1950 account of the origins of twentieth-century sexual
psychopath laws.
Those works illustrate three features of moral panics that
are germane to my cycles-and-sensibilities theme, and how
it can help us understand changes in punishment and drug
policies. Cohen shows how notorious events can, for a short
time, cause people and governments to lose their heads and
overreact. Erikson shows that moral panics often occur at
times of broader uncertainty and insecurity; the Salem witch
trials had much more to do with Salem than with witches.
Sutherland demonstrates that moral panics about some sub-
jects, in his case sexual predators, tend to recur at more or
less regular intervals.

Mods and Rockers
Easter weekend 1964 in Clacton, a small, slightly scruffy,

seaside resort on England’s east coast was cold and wet, the
coldest in eighty years. People were disgruntled, shopkeepers
because business was slow on the usually busy weekend,
young people because they were uncomfortable and there
was little to do. A few small groups scuffled and threw rocks
at each other and two groups of working-class youth, called
Mods and Rockers, began separating out. There were fights,
loud music, and some vandalism, and motorcycles roared
up and down. Eventually there were ninety-seven arrests,
twenty-four criminal charges (two or three for violent crimes),
and £513 in property damage (then around $1,500; in modern
equivalent $10,000–15,000). Smaller scale “riots,” involving
less violence, less damage, and fewer arrests, broke out in
other seaside resorts on three-day weekends later in 1964
and early in 1965.
The following Monday, all the English national newspa-
pers ran histrionic stories, usually on page one: “Day of
Terror by Scooter Groups” (Daily Telegraph), “Youngsters
Beat Up Town—97 Leather Jacket Arrests” (Daily Express),
“Wild Ones Invade Seaside—97 Arrests” (Daily Mirror).
Within weeks the Malicious Damage Act 1964, a toughened
law on vandalism, was introduced, enacted, and imple-
mented.
Cohen reconstructed in detail what happened in the riots
and, afterwards, in police, pretrial, and court proceedings.
Among the things documented were public hysteria, media
hyperbole, police harassment and intimidation of young peo-
ple, abuse by judges of pretrial detention, and vindictive
punishments. Anyone interested to know whether I have
exaggerated Cohen’s account and conclusions will have to
read his book but most will conclude that I have not.
Whatever the exact nature of the Clacton episode, reactions
clearly were exaggerated. This is shown by the following
statement made in parliamentary debate by Henry Brooke,
MORAL PANICS AND “WINDOWS OF OPPORTUNITY” 87

the Home Secretary (the English equivalent of the U.S. Attor-
ney General):
Some of the reports of what happened at Clacton over the
weekend were greatly exaggerated. . . . At Clacton more
than 1,000 young people came by one means or another,
apparently with little money on them, intending to sleep
wherever they could find some form of shelter. The
weather was bad over the Easter weekend and there was
little or nothing to do. They became bored, tempers flared
and a certain amount of fighting broke out. There was
nothing like a riot or gang warfare. Clacton was not sacked.
(S. Cohen 1972, p. 136)
Brooke proceeded, reports Cohen, to note that acts of as-
sault, theft, or malicious damage were isolated and committed
by small groups of individuals. The incidents nonetheless,
as they developed in the media and in popular and political
imagery, reified the Mods and Rockers into threatening youth
gangs and created a sense of crisis in English seaside resorts
for years to follow.
Things like the Clacton riots have happened throughout
history. Something or someone emerges who dramatically
threatens established values and interests, the media and state
agencies overreact and exaggerate the nature and scale of
the threat, public opinion becomes polarized and demands
decisive government responses, public officials adopt ex-
treme policies, and no one has very much patience for sugges-
tions that the problem is less serious or more complicated
than it looks.
Quakers and Salem Witches
Kai Erikson’s Wayward Puritans (1966) describes three chal-

lenges that affected Massachusetts Bay Colony in the seven-
teenth century. Two look like moral panics. The first, the
persecutions in 1656–65 of Quakers, whose radically decen-
tralized religious beliefs were seen as an attack on Puritan

Christianity, resulted in cut-off ears, executions, and private
violence. The persecutions ended only after Charles II or-
dered that they stop.
The second was the event we now know as the Salem witch
trials. It began early in 1692 in the Salem home of Reverend
Samuel Parris, where a number of girls aged 9 to 20 met
regularly with a slave woman from Barbados named Tituba.
Some of the girls began to act secretively and strangely. The
two youngest “began to exhibit a most unusual malady. They
would scream unaccountably, fall into grotesque convul-
sions, and sometimes scamper along on their hands and knees
making noises like the barking of a dog” (Erikson 1966,
p. 142).
Salem’s only doctor could find no medical explanation. He
concluded that the Devil had come to Salem and the girls
were bewitched. Ministers advised that the Devil must be
met head on and that the girls would have to identify the
witches who were harassing them. The girls identified three
women, including Tituba. At the trial, the girls rolled around,
as Erikson describes:
[in] apparent agony whenever some personal fancy (or the

invisible agents of the devil) provoked them to it. It was
a remarkable show. Strange creatures flew about the room
pecking at the girls or taunting them from the rafters, and
it was immediately obvious to everyone that the women
on trial were responsible for all the disorder and suffering.
When Sarah Good and Sarah Osborne were called to the
stand and asked why they sent these spectres to torment
the girls, they were too appalled to say much in their own
defense. But when Tituba took the stand she had a ready
answer. . . . Tituba gave her audiences one of the most
exuberant confessions ever recorded in a New England
courtroom. She spoke of the creatures who inhabit the
invisible world, the dark rituals that bind them together
in the service of Satan; and before she had ended her
astonishing recital she had convinced everyone in Salem
Village that the problem was far worse than they dared

imagine. For Tituba not only implicated Sarah Good and
Sarah Osborne in her own confession but announced that
many people in the colony were engaged in the devil’s
conspiracy against the Bay. (Erikson 1966, pp. 143–44)
The three defendants were convicted. Two were executed.

The third died in prison.
A year later, Salem’s witchcraft hysteria ended. In De-
cember 1692 Massachusetts Bay Colony Governor Phips re-
prieved the final eight persons sentenced to death, discharged
every prisoner, and issued a general pardon to all persons
under suspicion. By then, however, nineteen people had been
executed, one man had been crushed to death under rocks
for refusing to testify at his trial, two died in prison, 150
people were in custody, and 280 more had been accused.
Kai Erikson did not describe the Quaker persecutions or
the Salem trials as “moral panics.” The term was not yet in
use. Instead, he viewed these events in a Durkheimian frame-
work, as “crime waves,” which, together with responses to
them, helped identify and protect the boundaries of norma-
tively acceptable behavior. Why those behaviors, and why
such extreme responses? The answers for Erikson are that
the Colony was a Puritan theocracy that had outlived its
historical moment, and the boundaries were genuinely ambigu-
ous and in doubt. The bloody repression was an act of cultural
self-defense: “Historically, this kind of behavior is often associ-
ated with people who are no longer sure of their own place
in the world, people who need to protect their old customs
and ways all the more narrowly because they seem to have a
difficult time remembering quite who they are” (1966, p. 114).
In chapter 2, I discussed claims that contemporary Ameri-
can drug and punishment policies are a product of existential
angst (Garland 2001a) and emergence of a “risk society” (Gid-
dens 1990). I rejected those arguments because the social and
economic changes of the final third of the twentieth century,
and the heightened senses of insecurity and anxiety they
fostered, occurred in all Western countries, but no other
country adopted anything like American crime control poli-

cies. They may have been necessary, but they were not suffi-
cient conditions. As Erikson’s account of Salem makes clear,
however, deep insecurities do provide fertile soil for moral
panics.
Sexual Psychopaths and Predators
There have been at least three periods of heightened concerns

about sexual predation of children in this century, beginning
in approximately 1910, 1940, and 1990. The best known book,
Moral Panic: Changing Concepts of the Child Molester in Mod-
ern America (1998), by Philip Jenkins, tells of three periods
that seem each to have been moral panics precipitated by
isolated but horrifying incidents of child sexual abuse. The
incidents galvanized media attention, public fears, political
reactions, repressive legislation, and enhanced enforcement.
The latest, in our time, is manifested in the rounds of passage
in the 1990s of sex-offender registration and notification laws
and sexual psychopath laws providing for civil commitment
of sexually “dangerous” people (Lieb, Quinsey, and Berliner
1998).
Nearly a half century earlier, long before moral panic the-
ory took shape, Edwin Sutherland, then the most prominent
American criminologist, explained why the sexual psycho-
path laws of the 1940s were passed. The background con-
ditions included press concentration on sex crimes, which
“produces a widespread uneasiness which, given a few local
incidents, readily bursts into hysteria” (1950, p. 144). The
process:
The diffusion of sexual psychopath laws, consequently,

has occurred under the following conditions: a state of
fear developed, to some extent, by a general, nationwide
popular literature and made explicit by a few spectacular
sex crimes; a series of scattered and conflicting reactions
by many individuals and groups within the community;
the appointment of a committee which organizes existing

information regarding sex crimes . . . and presents a sex-
ual psychopath law to the legislature. (1950, pp. 146–67)
The same things happen today, only more, and faster. Al-
though Sutherland wrote of national news coverage, he was
trying to explain why individual U.S. states adopted sexual
psychopath laws. Horrifying incidents happened in individ-
ual states, which, in climates already made sensitive to such
things by national news coverage, overreacted, created com-
mittees, and passed laws that Sutherland deeply disapproved.
They were based, he said, on a series of propositions about
identification and treatment of sexual offenders that are “all
false or questionable, [but] they have nevertheless been very
effective in the diffusion of the laws” (1950, p. 147).
In our time, however, any incident anywhere, given a
drumbeat of national publicity, can influence policies in ev-
ery state. Polly Klaas was killed in California, but the concerns
her story raised influenced passage of California’s three-
strikes law and similar new laws, though generally more nar-
rowly drafted, in many other states. Megan Kanka was killed
in New Jersey, but within a few years, prodded by congres-
sional threats of loss of federal funding, all fifty states passed
Megan’s laws requiring registration of sex offenders and al-
lowing notification of their presence in the community (Lieb,
Quinsey, and Berliner 1998). This is partly because national
media, especially television, permeate nearly every pore of
American life in vivid, repetitive, often hysterical colors. It
is also partly because conservative American politicians have
for nearly two decades been playing the crime card and ex-
acerbating public fears and then proposing or enacting re-
pressive legislation in order to allay them (Edsall and Edsall
1991). Law reform need no longer await the outcome of delib-
erations of Sutherland’s expert commissions.
The term “moral panic” can be used in ways that range from
loose polemic, in which it operates as hyperbole to describe
whatever contentious issue is on the writer’s mind, to careful
depiction of a complex of social, psychological, and policy

responses to troubling events. Usually the term connotes
something rash and impulsive, and is usually used as a device
for explaining things the writer does not like, but the con-
cept, properly understood, is more neutral than that.
Reasonable people, for example, differ about the desirabil-
ity of the sex-offender registration procedures required by
“Megan’s Laws.” Opponents or skeptics see them, or at least
some of their features, as the undesirable fruits of a moral
panic—overbroad, stigmatizing, and inviting vigilantism.
Proponents, however, see that tragic death as having, at last,
focused attention on the need for more effective protection
of children from sex offenders and provided some momentum
to the effort to achieve it.
In crude and oversimplified terms, the enactment of Meg-
an’s Laws might be seen as a victory for law-and-order con-
servatives and a loss for due process liberals. Of course, it
is more complicated than that and views on a subject as
difficult as this one seldom closely parallel standard political
lines.
Moral panics focus attention on a troubling event or prob-
lem and generate emotions that can be harnessed. Sometimes
something happens as a result that otherwise would not have
happened. Sometimes the result is unequivocally bad as, for
example, when a crime by an immigrant provokes mob at-
tacks against immigrants. Sometimes horrible events provide
windows of opportunity for desirable policy changes. Desir-
ability is in the eye of the beholder.
Here is an example. The gunshot deaths of Martin Luther
King and Robert Kennedy precipitated passage of the federal
Gun Control Act of 1968, the first meaningful federal gun
control law of modern times. To gun control advocates, those
tragic deaths provided the occasion and momentum to win
enactment of important and necessary legislation that might
otherwise have been unattainable. From that perspective,
the emotion and upset generated by the two assassinations
produced something good.
To gun control opponents, however, the 1968 Act was the
unwanted effect of a moral panic about gun deaths and the

consequent reaction produced bad policy. To them, the 1968
legislation was ill-conceived and in derogation of fundamen-
tal American values (and they worked for decades, eventually
successfully, to weaken it). Thus, if Megan’s Laws might be
seen by some liberals as the unwanted fruits of a moral panic,
some conservatives saw the 1968 gun control legislation in
exactly the same way.
So we need to distinguish the dynamics of moral panics
from their outcomes. Sometimes a perceived crisis can trigger
changes that warrant applause and sometimes that warrant
condemnation. Deciding which is which inevitably is conten-
tious and implicates questions of values and ideology.
Thinking about this may be helped by switching from the
sociologists’ conception of moral panics to political scientists’
parallel but less polemic concept of “windows of opportu-
nity.” A sizable literature on how social science research in-
fluences policy making developed in the 1970s and 1980s.
The aim was to understand why credible research findings
seemed seldom to influence policy making in direct or straight-
forward ways (e.g., Lindblom and Cohen 1979; Weiss 1983,
1986; Tonry and Green 2003). The precipitant was frustra-
tions felt by many social scientists that research findings were
ignored even though they seemed clearly to demonstrate
ways to improve public policies. In our era, similar frustration
is often expressed by researchers concerning policy makers’
failure to change punishment policies aimed at deterring
crime in the face of evidence casting fundamental doubts on
the deterrent effects of harsh or increased penalties (e.g.,
Hood 2002).
The literature showed that research findings influenced
policy making much more often than despairing researchers
recognized, but often partially, indirectly, and after the pas-
sage of considerable periods of time. Various theoretical and
conceptual frameworks were developed that could be used
to examine the influence of ideology, politicians’ self-interest,
and bureaucratic inertia. The simple and seemingly obvious
but not unimportant points were made that whether policy
makers take account of research findings depends in part on
whether they are paying attention and whether the time is
right. When the time is right, a “window of opportunity”
opens through (or during) which people are especially open
to persuasion. The trick is to identify which windows are
open and when and to devise effective communication strate-
gies to bring the right information to the right people.
Liberal reformers and others who believe that policy mak-
ing should be rational and evidence-based may see an upset-
ting event as opening a window of opportunity for the mak-
ing of better policies. The King and Kennedy assassinations
may, in this light, have opened a window through which
evidence could pass about the use of handguns in violent
crime, handgun accidents, and handgun suicides, with the
salutary result that legislation was enacted that created
tighter controls on gun distribution. Likewise, more recently,
identification through DNA and other evidence of innocent
people who have been wrongly convicted of murder or rape
and sentenced to life imprisonment or death may provide a
window of opportunity for meaningful reform or repeal of
capital punishment laws.
That is fine as far as it goes, but so phrased it overlooks
that people other than researchers and due process liberals
have strong policy preferences and that windows of opportu-
nity can open for them as well. Some people, for example,
hold strong views about moral responsibility and the right-
ness of severely punishing or executing people who commit
serious crimes and may be frustrated that their views are not
more fully reflected in policy. The death of Polly Klaas, and
the reaction it provoked, could be seen as opening a window
for more persuasive communication of moral views about
deserved and severe punishments.
Whatever the terminologies, notions like moral panic and
window of opportunity represent two-edged swords that can
cut to left or right, toward sympathetic understanding or
condign punishment, toward due process or crime control.
Which way the sword cuts depends on the environment in
which it swings. The political environment of the United
States during the last three decades of the twentieth century
was dominated by moral and political conservatives. They
won many of the contentious conflicts over crime control
and drug policy, sometimes by capitalizing on emotions gen-
erated by events and upsets that liberals saw as moral panics.
That the right so often successfully took advantage of dis-
tressing events to promote their policy agenda may be why
“moral panic” as used by academics generally has negative
connotations. There are other reasons. Evidence and rational
analysis should be central features of policy making. Deci-
sions made in the heat of an overwrought moment often
ignore both, and everyone, whatever their politics, should
regret that.
We should not lose sight of an important difference be-
tween moral panics that precipitate passage of gun control
legislation or repeal of death penalty laws and ones that pre-
cipitate passage of three-strikes laws or repeal laws that pro-
tect defendant’s procedural protections. The latter take away
people’s liberty and lives, sometimes mistakenly. The former
merely restrict peoples’ access to guns or frustrate their
wishes to see others killed.
The anger, emotion, and urgency that moral panics gener-
ate can be harnessed to various ends. Sometimes, as with
gun legislation, the momentum of a moral panic can overcome
the resistance of special interests and achieve a public good.
More often, they produce the kinds of results that Cohen,
Hall, Erikson, and Sutherland decried. The great danger is
that they lead to actions and policies based on stereotype,
anger, and emotion, rather than on careful assessment of
problems, cool reflection, and rationality.
During recent decades of decline in rates of crime and drug
abuse, American sensibilities, as David Musto predicted,
have been harsh and punitive, have been predisposed to
target members of minority groups as enemies, and have
been intolerant of the views of dissenters from the current
conventional wisdom. Penal sensibilities associated with a
period of declining crime rates amplified the effects of a series
of moral panics and made things worse.

5
Crime Trends and the Effects

of Crime Control Policies
Prevailing sensibilities in the contemporary United States

have produced, or at least allowed, drug and crime control
policies of severity unprecedented in recent American history
and unmatched in other Western countries today. There was
a de facto moratorium on capital punishment in the United
States in 1970 and belief was widespread that the Supreme
Court would soon declare the death penalty unconstitutional.
For eight years, from 1968 to 1977, no one was executed. In
2001, capital punishment was authorized in thirty-eight
states and the federal system, sixty-six people were put to
death, and at year’s end 3,625 were on death row (Death
Penalty Information Center 2002). In June 2001 the execution
of Timothy McVeigh, shown “live” on closed circuit televi-
sion to more than 230 people, was the first execution by
the federal government since 1963. By contrast, in 2001, no
Western European country, or Australia, Canada, England,
or New Zealand, allowed capital punishment. Courts in South
Africa and Hungary had declared their countries’ capital
punishment laws unconstitutional. The dozens of Eastern and
Southern European countries joining the Council of Europe
agreed to forego capital punishment as a condition of mem-
bership.
Concerning imprisonment, the U.S. prison and jail incarcer-
ation rate of more than 700 per 100,000 population in 2002
was five times the rate of 144 per 100,000 in 1970 and is nearly
five times higher than that of any other Western country. In
1970 Congress repealed most federal laws requiring manda-
tory minimums. In the states, mandatory minimum sentences
were modest in scope and severity, usually at most calling
97
for one- or two-year minimum sentences. There were no mean-
ingful three-strikes or sexual psychopath laws. Few prisoners
were sentenced to terms of life-without-possibility-of parole
or to non-parolable terms measured in decades. All of those
sentences, including lengthy mandatory minimums, are com-
monplace in America in 2001 and rare or nonexistent else-
where.
Why are contemporary American sensibilities as they are
and why do they allow harsher punishments than character-
ize other times and places? The best available evidence shows
that gross crime trends are determined by fundamental social
and structural forces that affect most Western countries, and
that they follow much the same broad patterns irrespective of
national differences in crime control policies and punishment
practices. If that is true, and the relevant historical and empir-
ical literatures demonstrate that it is, it is unrealistic to sup-
pose that short-term changes in penal policy or practice can
affect rates, patterns, and trends in any particular country
other than at the margins. Other countries experienced crime
rises in the 1970s and 1980s, followed by declines in the
1990s, the same patterns the United States experienced, but
did not adopt penal policies of unprecedented severity. Why
did the United States?
Historians and sociologists of deviance offer guidance. His-
torians show that crime rate patterns in Western countries
are broadly similar and that attitudes, public debate, and
policy vary in predictable ways over extended periods in
relation to rises and falls in cycles of deviant behavior. Sociol-
ogists show how, usually unpredictably, short-term moral
panics occur during which problems are exaggerated, public
attitudes become polarized, traditional values of moderation
and balance are cast aside, moral entrepreneurs become in-
fluential, and extreme policies are adopted. The misfortune
for our time is that long-term deviance cycles during which
intolerance and excessive severity are to be expected have
coincided with a series of moral panics, and the long-term
cycles and the short-term panics have each exacerbated the
other’s effects.

The historical and sociological literatures explain and docu-
ment social processes that shape how people perceive threats
in the world around them. Both fit within what might loosely
be called a Durkheimian framework, after the French sociolo-
gist Émile Durkheim. Durkheim argued that one of the crimi-
nal law’s functions is to identify and reinforce basic social
ideas about right and wrong, that “crime brings together
upright consciences and concentrates them.” The criminal
law is seen as performing a dramaturgical function, with
punishment directed primarily at the community and not at
the offender, and working not directly through deterrence
and incapacitation but indirectly through affirmation and
reiteration of basic norms.
Crime and punishment are seen as parts of every human
society, mechanisms that help set and then illuminate the
boundaries of acceptable behavior. The primary aims of the
criminal law and punishment are to state, restate, and rein-
force prevailing norms: “[Punishment] does not serve, or else
serves only quite secondarily, in correcting the culpable, or
in intimidating possible followers. From this point of view,
its efficacy is justly doubtful and, in any case, mediocre. Its
true function is to maintain social cohesion intact” (Durkheim
1933 [orig. 1893], p. 108).
Durkheim’s dramaturgical functions of law are among the
ways that socialization occurs and through which sensibili-
ties are shaped. Criminal law and punishment, however, are
not the primary means of socialization into right values. That
function belongs to primary social organizations such as the
family, the church, the workplace, the community, and to
kinship and friendship networks.
Cycles of tolerance and intolerance and moral panics oper-
ate at an even deeper contextual level, however, than do
legal institutions, and have greater influence on men’s minds
and actions. If legal institutions and their operations serve
as a backdrop to the socializing and norm-declaring roles of
primary institutions and networks, cycles and moral panics
operate as a backdrop to them. However, because of the ways
they influence sensibilities, and through sensibilities how
THE EFFECTS OF CRIME CONTROL POLICIES 99

people perceive and interpret the world, cycles and panics
play the more powerful roles.
Four principal claims are made in this chapter. First, human
life is characterized in many spheres by cycles, by regular
oscillations between recurring states or conditions. Second,
in the United States and in other Western countries for which
data are available, crime rates change slowly, over extended
periods, and for reasons that have no apparent relationship
to changes in crime control policies or criminal justice institu-
tions. Third, in relation to drug policy there is a regular
interaction between drug use cycles and punitive antidrug
policies, with the harshest policies being adopted when drug
use is declining rather than, as at first impression probably
to most people seems more likely, when drug use is rising.
Fourth, though to my knowledge there is no historical litera-
ture on interactions among crime rate cycles, sensibilities,
and crime control policies, there is little reason to doubt that
the patterns characterizing drug use cycles apply to crime
and crime control policies.
Oscillations
Human behaviors, values, and beliefs oscillate over time,

moving back and forth between what are widely seen as
fundamentally different positions. The business cycle’s peri-
ods of alternating expansion and contraction are a cliché,
but the world economic downturn of 2001–2003 confirms
its underlying reality. For the past two centuries in Western
countries, regular shifts have taken place in the breadth of
popular support for conservative and liberal political parties,
and governments have shifted with them. Histories of Chris-
tianity document regular fluctuations over 2,000 years be-
tween periods of doctrinal orthodoxy and fierce, often deadly,
resistance to apostasy, and periods of heterodoxy and toler-
ance. Histories of homosexuality document a similar aeons-
long fluctuation between periods of comfortable tolerance
and lethal intolerance.
Grant Gilmore describes the regular alternation between
classical periods’ celebration of formal rules of structure and
composition in the arts and the confused, sprawling, form-
less, and chaotic agonies of the romantic periods that follow.
Gilmore is important not only because he illustrates recurring
alternation between opposed perspectives, but because he
describes the psychological power of orthodoxy. Prevailing
sensibilities affect how people think. During classical peaks,
tradition, form, regularity, and structure are admired because
most people believe formalism is an aesthetically important
ideal, and at romantic peaks most people believe that experi-
mentation, spontaneity, and exuberance are important val-
ues. What this shows is that people’s beliefs are at least influ-
enced and often determined by the values of the era in which
they live and, accordingly, that many individuals would have
believed different things had they lived in different times.
People’s beliefs about norms, values, and ideals are impor-
tantly different from their beliefs about other things, for
example, technology. Beliefs about optimal characteristics of
internal combustion engines are not fundamentally cyclical
but evolutionary. At any developmental stage, some engines
are preferred because they are cleaner, more efficient, more
reliable, or more durable but, by those criteria, earlier engines
are almost inevitably inferior to those later developed. For
such things, something like perfectionism is an intelligible
frame of reference. Ideas about romanticism, homosexuality,
religious pluralism, and punishment are not like that.
Long-Term Crime Trends
Crime rates rise and fall over extended periods for reasons
that appear to have little to do with the actions or policies
of governments. Historians who have studied such matters
agree on a number of propositions about the history of crime
rates.
First, violent crime rates in the British Isles and continental
Europe fell steadily from the twelfth century, the earliest

time for which quantitative data are available, until the 1950s
or 1960s (Eisner 2001, 2003). Figures 5.1, 5.2, 5.3, and 5.4,
from Swiss historian Manuel Eisner’s work, show long-term
trends in homicide rates in England, Scandinavia, Holland
and Belgium, and Germany and Switzerland. The figures use
a log linear scale to show the data. This allows each band
shown to represent an order of magnitude difference from
those above or below, but within each band the intervals
are comparable. The dots on the figures represent quantita-
tive data from individual studies of official records for partic-
ular places for particular times. Any individual study can
be second-guessed in terms of the reliability of the data
source, the representativeness of the incidents that were re-
corded, and the adequacy of population estimates used to
calculate homicide rates. So long, however, as they fall within
the “right” band of the figure, being off even by a large
margin does not upset the basic patterns.
Looked at as a whole, the data tell remarkably consistent
stories in all four regions. In each, homicide rates were be-
tween 30 and 100 per 100,000 population in the earliest pe-
Figure 5.1 England—local estimates and national series

Source: Eisner 2003. History of Homicide Database.

Figure 5.2 Scandinavia—local estimates and national series for
Sweden
Figure 5.3 Netherlands and Belgium—local estimates and national

series

Figure 5.4 Germany and Switzerland—local estimates and national
series
riod and then, in a series of sharp falls alternating with

lengthy plateaus fell below 10 per 100,000 in the eighteenth
or nineteenth century to 1 or 2 per 100,000 in the twentieth
century. Eisner offers a variety of explanations that range
from broad theory, including Elias’s civilizing process and
increasing state roles in order maintenance, and individual-
ism, to technology (steadily improved medical care has re-
duced the likelihood of death from many types of assault).
The fundamental and uncontroversial point, however, is that
homicide rates in Western Europe decreased continuously
for at least five centuries.
Second, if only more recent periods for which more, and
more reliable, data are available are considered, violence rates
declined in Western countries from the early to mid-nine-
teenth century until the 1950s or 1960s and rose thereafter
but to levels well below either the medieval or nineteenth
century starting points (Gurr 1981, 1989; Lane 1992, 1999).
American historians refer to this pattern as a “U-shaped” or
“reverse J-shaped” curve.
Roger Lane, America’s leading historian of crime and polic-
ing, has written: “In the United States, as everywhere in the
developed world, rates of homicide and other violent crimes,
if graphed, would form a kind of long, ragged, reverse J-
curve. That is, while short-term events might create zigs
and zags, the basic curve dropped from the mid-nineteenth
century to the mid twentieth” (Lane 1999, p. 205).
Crime rates then rose for several decades in the United
States and in all other Western countries. More recently,
however, crime rates have again begun to decline in many
Western countries (including Australia, Canada, Denmark,
England and Wales, Finland, Germany, the Netherlands, and
the United States) as measured by official police data on re-
corded crimes, data from surveys in which people are asked
about crimes by which they have been victimized, or, in most
countries, both. It is too early to know whether significant
increases in violent crime rates in most Western countries
since the 1960s signaled a change in the eight- or two-century-
long trends or whether they were merely short-term pertur-
bations like previous up-ticks associated with wars and peri-
ods of sustained unrest. Short-term perturbation is the best
guess. The falling crime rate in the 1990s in most Western
countries, and within federal countries, in most states, irre-
spective of wide variations in crime policies, imprisonment
rates, and punitive sensibilities, suggests the latter.
Figure 5.5 depicts the U/reverse J-shaped American vio-
lent-crime-rate pattern for most of the past two centuries. It
also shows major criminal justice policy changes from 1840
to 1960. During the period of continuous long-term decline in
violent crime rates, most of the major institutions of modern
criminal justice systems were established: the penitentiary
(1820s through 1840s), the first probation programs (1850s
and 1860s onwards), the rehabilitation-premised reformatory
for young offenders (1870s), parole (1880s onwards), and the
juvenile court (1899 onwards). By 1930 nearly all of these
institutions existed in every American jurisdiction (Rothman
1980). The ubiquitous indeterminate sentencing systems that
characterized every jurisdiction, and survived everywhere

Figure 5.5 American violent-crime-rate pattern with policy changes
Sources: Gurr 1981, 1989.
little changed until 1975, had taken shape. In 1850 something

that resembled a combination of modern ”truth in sentenc-
ing“ and ”determinate sentencing“ existed throughout the
United States; by 1930 it existed nowhere.
One might not unreasonably suppose that the invention
of the modern criminal justice system is, or was widely seen
to be, the reason why crime rates declined so much and for
so long. Despite the shift from determinate to indeterminate
sentencing, however, and the invention of so many new and
unprecedented institutions, changes in criminal justice poli-
cies, practices, and processes are seldom mentioned as likely
or leading causes of the century-long decline in crime rates
(Rothman 1980; Friedman 1993; S. Walker 1998). Roger Lane
reported that other leading police historians “believe police
impact on non-‘order’ offenses was minimal and . . . [were]
almost contemptuous about the efficacy of patrol, Walker
[1977] pointing out that the thin blue line was too thin to
make much difference, Fogelson [1977] insisting that the duty
was more often shirked than shouldered” (Lane 1980, p. 38).
Lane’s “own more fully considered position would be that
the cops ‘worked’ only as part of the demand—created fun-
damentally by a changing economic order—for a more disci-
plined society and workforce. Acting alone, or out of phase
with this deeper economic need, as in an age of labor surplus,
their impact was (and still is) minimal” (Lane 1980, p. 38).
Samuel Walker, one of America’s leading criminal justice
historians, observes of the nineteenth-century police, “It is
unlikely that the police did much to prevent crime and disor-
der . . . there were too few officers spread too thin, spending
too much of their time avoiding work“ (1998, p. 66), and of
prisons that “the impact of the prisons and the other new
institutions on crime is a matter for much speculation. Crime
and disorder declined throughout the nineteenth century,
but it is unlikely that the prison was responsible for that,
any more than were the police” (1998, p. 109).
Instead, the primary explanation provided by Ted Robert
Gurr (1981, 1989) and Roger Lane (1982 and 1999) is a combi-
nation of a centuries-long decreasing tolerance for violence
in Western countries and the effects of the bureaucratization
of modern life. The first of these is consistent with Elias’s
“civilizing process,” and the second with Foucault’s ideas.
Since the early nineteenth century, individuals have been
socialized into conformity by public schools, factories, the
military, and other institutions. This is not unlike the Fou-
cauldian thesis in Discipline and Punish (1977) that prisons,
mental institutions, and the army exemplify a wider range
of ”disciplinary institutions“ that mold people into the roles
demanded by the modern state and economy.
Political scientist James Q. Wilson (1993; Wilson and
Herrnstein 1985) argued that a major part of the explanation
lay in the waves of religious revivalism of the nineteenth
century. These, he proposed, strengthened moral and ethical
standards of right behavior and personal responsibility,
which then became central features of the socialization of
children in families, schools, churches, and communities.
That in turn explains the decline in crime rates. Conversely,
in the latter half of the twentieth century, traditional norm-
creating processes broke down and the decline in ”the moral
sense“ is a major explanation for the rising crime rates of
the 1960s through the 1980s.
Wilson’s explanation of the effects of nineteenth-century
social changes is not very different from Durkheim’s views
or the Northern European notion of ”positive general preven-
tion” (Törnudd 1997; Lappi-Seppälä 2001). Most people ab-
stain from committing serious crimes because of internalized
ideas about right and wrong, which are the product of social-
ization by primary institutions such as the family, church,
and school, and not because they will be punished if they
are caught. In Wilson’s view, social changes of the last fifty
years broke down these socialization processes, but religious
reawakening in the nineteenth century strengthened them.
None of Gurr, Lane, and Wilson, it bears repeating, how-
ever, primarily attributes nineteenth-century crime trends
to the invention and implementation of the institutions of
individualized and indeterminate sentencing. This stands in
stark contrast to the last twenty years in the United States
when public figures and many academics have attributed
crime rate declines to criminal justice policy changes such
as increased use of imprisonment, three-strikes laws, and
zero-tolerance policing. I take this up later on in this chapter.
To anticipate: contemporary sensibilities being as they are,
many Americans are predisposed to believe that harsh anti-
crime policies are morally right and likely to work. As crime
rates fell, it was natural to believe anticrime policies were
the reason why, especially for political figures like California
Governor Pete Wilson and New York City Mayor Rudolph
Giuliani for whom harsh anticrime policies were part of their
political identities. Having their policies “work” was in their
political self-interest, and it takes no subtle psychology to
recognize that people tend to believe what is in their self-
interest to believe. People like Wilson and Giuliani, assuming

they really believed their policies worked, were almost cer-
tainly mistaken. Crime rates fell in every Western country in
the 1990s, whatever the governing crime control philosophy
and practices, and in every American city and state, whether or
not they adopted three-strikes laws or zero-tolerance policing.
Drug Use and Drug Policy
There is a regular interaction between declines in drug use,

enactment of harsh public policies, and demonization of mi-
nority groups. The pattern has recurred in the past twenty
years in the United States, with drug use peaking in 1979–80
for most drugs (and 1982–84 for cocaine), the harshest anti-
drug laws being passed in the late 1980s, and black inner-
city residents being portrayed as the enemy in the drug wars
(Tonry 1995, chaps. 1, 3). The same pattern holds for crime
in general: crime rates have fallen since 1990–91 (some would
say since the early 1980s), the harshest laws were passed in
the early 1990s, and blacks and Hispanics have been the
principal targets of the crime wars.
Since my interest here is in how secular trends affect peo-
ple’s thinking, the bases for the preceding assertions are only
sketched. The basic outline of the argument comes from histo-
rian David Musto’s (1973, 1987a) work on the history of
American drug policy, embellished a bit by evidence from
historian Joseph Gusfield’s (1963) work on the history of
alcohol policy.
During the past two centuries, the United States has experi-
enced three peaks of prohibitionistic moralism: the 1850s when
by 1856 eight states had enacted alcohol prohibition; 1915–35,
the era of national Prohibition; and 1980 to the present.
Nineteenth-Century Alcohol Prohibition

Antebellum prohibitionism culminated at least forty years
of temperance activism, but was truncated by the Civil War,
which distracted activists and policy makers. Three things,

however, stand out in Gusfield’s account of the history of
alcohol policy. First, the prohibition and temperance move-
ments of the nineteenth century were partly a product of
status conflicts between cultural groups—initially settled,
generally abstemious, American Protestants of British descent
against newly arrived, often Catholic, often bibulous Euro-
pean immigrants. Prohibition was a device through which a
“cultural group act[ed] to preserve, defend, or enhance the
dominance and prestige of its own style of living within the
total society” (1963, p. 3).
Second, particular minority groups, in the eyes of the domi-
nant settled Protestants, came to personify the dangers and
the lifestyle represented by alcohol:
The symbol of abstinence as a symbol of respectability
was enhanced when large numbers of Irish and German
immigrants entered the United States and made up the
unskilled labor forces of the growing urban centers during
the 1840s and 1850s. In the culture of the Irish and the
Germans, use of whiskey or beer was customary and often
a staple part of the diet. Both groups were at the bottom
of the class and status structure in American society. In
the evolution of class symbols, the groups at the lowest
rungs of the ladder affect the behavior of those above by
a process of depletion in which those traits originally
shared by both groups become progressively deprized
among the more prestigeful. The incoming group thus
widens the status gap between it and the natives. If the
lowly Irish and Germans were the drinkers and the drunk-
ards of the community, it was more necessary than ever
that the aspirant to middle-class membership not risk the
possibility that he might be classed with the immigrants.
(Gusfield 1963, p. 51)
Easily recognized lower status groups like the preponder-
antly Catholic Irish and German immigrants personified the
dangers and moral laxity of alcohol consumption. Con-
versely, alcohol use and abuse personified the irresponsibility
and lesser worthiness of the newer immigrants.

Third, when drinkers refused either to accept calls to ab-
stention or to acknowledge the wrongfulness of their behav-
ior, they became enemies:
When one group acts in a manner contradicting the other’s

belief in the legitimacy and domination of its own norms,
the situation becomes that of a conflict between enemies.
. . . The assumption that the norms-violator recognizes the
legitimacy or domination [of prevalent norms] . . . is con-
tradicted when the norms-violator is perceived as an en-
emy . . . someone who is hostile and must be approached
as an enemy; who must be forced to accept the dominance
of the reformer . . . [is] a hostile enemy who must be co-
erced through legislation. (Gusfield 1963, pp. 67–70)
All three of these themes—drug policy as an arena of cul-

tural conflict, identification of low-status minority groups
with immorality, and conversion of people who reject domi-
nant values into enemies—recur in both the alcohol and drug
wars of the twentieth centuries.
Twentieth-Century “Drug” Prohibition

David Musto has several times written of the seeming anom-
aly that prohibitionistic sentiments become strongest and
drug policies harshest after drug use has begun to decline.
Use of psychotropic and other mood-altering substances is
always a contested subject. Peaks in the prevalence of use
followed by drops are often accompanied by a hardening
of anti-drug sentiments in the population, a reduction in
perceived ambiguities about the issues drug use raises, and
a shift toward more vigorous, moralistic, and punitive poli-
cies and practices. Here I develop only his further point about
a tendency at such times of declining drug use for socially
marginal minority groups to be stereotyped as immoral and
irresponsible drug abusers.
In a revised version of his book published at a time when
the modern drug wars were well underway, Musto wrote:

“One cannot help but be concerned that the fear of drugs
will again transfer into a simple fear of the drug user and
will be accompanied by draconian sentences and specious
links between certain drugs and distrusted groups within
society, as was the case with cocaine and Southern blacks
early in the Twentieth Century” (1987a, p. 27).
Throughout U.S. history during periods of high intolerance
of drug use, minority group stereotypes have been associated
with deviant drug use. In the mid-nineteenth century, per
Gusfield, immigrant Irish and Germans were targeted. In the
late nineteenth and early twentieth centuries, even though
middle-class women were the modal category of opiate users,
Chinese opium smokers and opium dens were among the
images invoked by opponents of drug use and are part of
the backdrop to the Harrison Act of 1914, the first major
federal antinarcotics legislation. At about the same time and
in the 1920s, it was blacks and cocaine. In the 1930s, imagery
of Mexicans and marijuana was prominent in the anti-mari-
juana movements that culminated in the Marijuana Tax Act
of 1937 and in many state laws criminalizing marijuana use.
In the antidrug atmosphere of the 1980s, crack cocaine, the
emblematic drug of the latest drug war, became associated
in public imagery with disadvantaged black residents of the
inner cities (Musto 1987a, b; Tonry 1995, chap. 3; Mauer 1999).
Musto observes that excesses in overzealous law enforce-
ment, stigmatization of minority groups, and stereotyping of
drug users occur when drug use is declining, because “people
become so righteous and so zealous.” They become zealous
because drug use comes to be seen in terms of a contrast be-
tween moral right and immoral wrong. They become righ-
teous because right is on their side.
Cycles of Intolerance of Criminals?
The two preceding sections offered two propositions. Crime

rates change slowly, in response to long-term underlying
social and normative changes. At least in relation to drug use
policies, penal sensibilities vary depending on whether drug
use is increasing or decreasing, with the harshest sensibilities
developing shortly after usage peaks and has begun to fall.
Taken together, those propositions support a hypothesis that
penal sensibilities concerning crime parallel those concerning
drug use. Whether the three-decade increase in American
crime rates was a perturbation in the centuries-long decline
in violent crime, or not, rates have declined sharply since
at least the early 1990s. Applying Musto’s model to crime
generates hypotheses that attitudes would become harsher,
policies severer, law enforcement more vigorous, and minori-
ties more ensnared in the justice system. All of that happened.
The principal questions are whether the drug use, drug
policy pattern Musto first described in 1973 occurred during
the current drug wars, and, if so, whether the same pattern
characterizes crime rates and crime control policies. Yes, and
yes, are the answers. First, though, a look at recent crime
rate declines, and whether they are likelier to result from
long-term social forces or short-term policy innovations, is
in order. Then I turn to the central questions.
Short-Term Crime Trends

The recent fall in U.S. crime rates is much likelier to result
from underlying social and normative changes than from
recently adopted crime control policies such as three-strikes
laws, zero-tolerance policing, or mass incarceration. That so
many people believe, or purport to believe, the latter is a
product of contemporary penal sensibilities that predispose
such beliefs. There are two major, and several lesser, reasons
for so concluding.
The place to begin is to look at crime rates from the Uniform
Crime Reports (the “UCR”) for the last forty years. Figure
5.6 shows rates of crime reported to the police for murder,
rape, robbery, motor vehicle theft, and burglary from 1960
to 2000. In order to show rates for those five crimes on one
figure, rates for burglary and motor vehicle theft have been
divided by ten and those for rape and murder have been

Figure 5.6 Reported crime rates, United States 1960–2000 (per
100,000 population), murder, rape, robbery, burglary, and motor
vehicle theft
Source: Sourcebook of Criminal Justice Statistics Online, Table 3.120.
multiplied by ten. This does not distort the figure’s message

because its aims are to show trends for particular crimes and
allow comparisons between trends for different crimes.
The Uniform Crime Reports, which are compiled and pub-
lished annually by the FBI from data provided by police forces,
classify those five crimes plus aggravated assault, theft, and
arson as Index Crimes. I excluded arson from figure 5.6,
because of its relatively low incidence. Aggravated assault
is excluded, because we know that the percentages of assaults
reported to the police steadily increased during the past
twenty years, and rising rates partially reflect major changes
in citizen reporting rather than changes in the occurrence
of assaults. Rising rates of aggravated assault partly result
from declining tolerance of violence, which means that as-
saults previously seen as too minor to record are recorded,

and partly from reduced tolerance of domestic violence,
which means that acts of family violence once seen as primar-
ily private matters are recorded as crimes (Blumstein 1993).
Rape reporting patterns have also changed in recent decades
because alleged acquaintance, date, and marital rapes are
much more often reported and recorded than in earlier times.
I nonetheless include rape in figure 5.6 because of its serious-
ness and policy importance and because reporting and re-
cording patterns are more likely than those for aggravated
assault to have stabilized and less likely to create serious
distortions.
Several things stand out from figure 5.6. The trends for all
five crimes are broadly similar. Rates for all five rose steeply
from the 1960s to the early 1980s, reaching their highest
levels since 1960. Then they fell for five years. Beginning in
1986, they began another rise that lasted for five years and
topped out in 1990–91. After that, rates fell sharply and
consistently. Some people date the crime rate declines from
the first peak in the 1980s, others from the early 1990s highs.
It does not really matter. Rates may have begun a long-term
decline around 1980 and then temporarily changed direction
before resuming. Or the climb may have begun in earnest
in 1980 but temporarily been obscured by the effects of the
crack cocaine epidemic beginning in the mid-1980s. Alfred
Blumstein (1993) has argued that the rapid expansion of new
crack cocaine markets, at a time when youth gangs were
spreading rapidly and ever-more-lethal firearms were becom-
ing readily accessible, produced an explosion of violence and
related crime, which in a few years abated when the markets
matured and turf battles ended. Either way crime rates have
been falling substantially for a considerable period. Sooner
or later they will stabilize, and inevitably they will at least
for a time rise again. As yet though, there have been no signs
of steep upward movements.
The timing of the major shifts in direction is the same
whether the increases of the late 1980s are taken into account
or not. Rates for all five offenses peaked around 1980, and
have fallen since, or peaked twice, and each time fell sharply
thereafter, but in either case, rates for all five changed at
about the same times. Something fundamental was changing
in the United States and it affected each of these major crimes
in the same ways.
Two other sources of information about crime trends, the
U.S. National Crime Victimization Survey (the “NCVS”) and
the International Crime Victimization Survey (the “ICVS”),
tell the same story. Victimization surveys, which ask respon-
dents to answer questions about crimes they have suffered,
provide a crosscheck on official crime data. Victimization
surveys generate much higher estimates of the amount of
crime than do official data. That is because they include minor
and uncompleted crimes that people do not consider worth
reporting or about which it appears unlikely the police can
do anything. However, the proportions of each type of vic-
timization that are reported to the police are known, and
change little from year to year. Thus, though the absolute
levels of crime reported by the UCR and the NCVS should
be different, the trends they show should be the same.
The NCVS, conducted by the U.S. Bureau of the Census
for the U.S. Bureau of Justice Statistics, has since 1973 pub-
lished data from interviews conducted every six months with
members of 40,000 to 60,000 households. It is the world’s
largest, longest running, and most technically sophisticated
national victimization survey. Figure 5.7 shows reported vic-
timization rates for assault, rape, robbery, and theft per 1,000
persons aged 12 or over, burglary rates per 1,000 households,
and motor vehicle theft per 1,000 vehicles for the years 1980
to 2000. Rates for burglary and theft are divided by 10. These
are four of the five crimes shown in figure 5.6, plus theft
and assault. The victimization trends for rape, robbery, and
motor vehicle theft follow the same twin-peaks-followed-by-
steady-decline pattern as the official crime data shown in
figure 5.6. So does aggravated assault, which I have included
in this figure, since victims’ reports are not distorted by
official reporting and recording practices. Burglary and theft
show continuous declines.

Figure 5.7 U.S. victimization rate as measured by the NCVS,
1980–2000
Source: National Criminal Victimization Survey, 1980–2000.
The ICVS is a smaller, newer, and less frequent victimiza-

tion survey than the NCVS. It has been conducted four times
beginning in 1988–89 by varying numbers of countries, but
always including England, Holland, Germany, and the United
States. Its goals include cross-national comparisons of crime
trends and rates. To do this, and avoid the complicating
effects of national differences in how crimes are defined, re-
ported, and recorded, the ICVS employs standardized ques-
tionnaires and methods in each country. Rather than ask
respondents whether they have suffered particular crimes,
the definitions of which vary between countries, the ICVS
asks such things as, has anyone taken anything from you?
When the response is “yes,” follow-up questions pinpoint
exactly what happened. For that reason, and because in vari-

ous ways it is less sophisticated than the NCVS, the ICVS
differs substantially from the NCVS in many technical re-
spects. It also uses a much smaller sample of respondents.
And yet, as table 5.1 shows for robbery, assault, burglary,
and auto theft, the American victimization rate trends for
the ICVS parallel the victimization rates reported by the
NCVS and the crime rate trends of the Uniform Crime Reports.
U.S. crime rates, however measured, have been falling for
more than a decade. Year-to-year changes may result from
random variation or, in the Uniform Crime Reports, changes
from one year to another in which jurisdictions fail to submit
data on time or at all. Trend changes for a single crime such
as notably rape, aggravated assault, or auto theft may result
from changes in citizens’ reporting or police recording pat-
terns. But when all credible sources of data on crime trends
show the same trends for a decade, when the rate declines
are steep and nearly continuous, and when they are the same
for murder, the most reliably reported crime, as they are for
other crimes, something has happened and crime has fallen
as a result.
Why Have Crime Rates Fallen?

The most obvious explanations for why crime rates have
fallen, and the most widely believed—that sentencing poli-
Table 5.1 U.S. assault, auto theft, burglary,

and robbery rates per 1,000 Americans 16
and over, International Crime Victims Survey,
1989, 1996, 2000
1989 1996 2000
Assault 101 100 65

Auto theft 29 20 6
Burglary 56 39 33
Robbery 28 16 6
Source: John van Kesteren, Pat Mayhew, and Paul Nieuw-

beerta. 2000. Criminal Victimisation in Seventeen Industrialized
Countries—Key Findings from the 2000 International Crime Vic-
tims Survey. The Hague: Netherlands Ministry of Justice.

cies became tougher, police became more aggressive and more
efficient, and many more people were sent to prison—are
almost certainly wrong. These policy changes no doubt in-
fluenced the fall in crime rates, but they were not the major
explanation. This can be shown in two ways.
First, crime trends in cities and states that adopted espe-
cially celebrated crime control policies can be compared with
trends in comparable cities and states that did not. When such
comparisons are made, it becomes clear that comparable trends
occurred everywhere, including in cities that did not adopt
aggressive zero-tolerance policing styles, and in states that
adopted neither truth-in-sentencing nor three-strikes laws.
Second, crime trends in the United States in the 1990s can
be compared with those in other Western countries that did
not adopt harsh American-style crime control policies. Incar-
ceration rates and trends can be used to show that U.S. poli-
cies were more punitive than those elsewhere and that the
increase in punitiveness was unmatched elsewhere. The U.S.
imprisonment grew by half between 1990 and 2002, rising
from 450 per 100,000 to over 700. By comparison the highest
rates in other Western countries in 2000 were around 150
per 100,000 (Portugal, New Zealand) and for most were below
100 (France, Germany, Italy, all the Scandinavian countries)
(Barclay and Tavares 2002).
In addition, there were substantial country-to-country
variations in imprisonment rate patterns (Tonry 2001). They
went up about as rapidly in Holland as in the United States,
rose significantly in England, New Zealand, and Portugal,
fell steeply in Finland, were broadly stable in Germany and
Canada, and fluctuated widely in France. Yet crime rates fell
in the 1990s in every Western country. Once cross-national
comparisons are made, it is not obvious why harsh policies
should be seen as explaining the U.S. crime rate declines
when similar declines occurred in many countries that did
not adopt U.S.-style policies.
Zero-Tolerance Policing in New York City Not everyone knows

about historical crime trends or how U.S. trends compare
with those elsewhere. So it is not unreasonable that many
people wrongly believe that tougher crime control policies
are the primary reason American crime rates fell in the 1990s.
After all, New York City in 1994 introduced policy changes
popularly known as zero-tolerance policing and New York
City crime rates fell in 1995. California enacted its three-
strikes law in 1994 and crime rates fell in 1995. Nationally,
the number of people in state and federal prisons quadrupled
from 316,000 in 1980 to 1,300,000 in 2000 and the violent
crime rate in the Uniform Crime Reports fell substantially
during the same period. Unfortunately, all of those appear-
ances are misleading.
Consider, first, the effects of zero-tolerance policing in New
York and California’s three-strikes law. In each place, crime
rates fell the year after the innovation took effect, and New
York’s Mayor Rudolph Giuliani and California’s Governor
Pete Wilson were quick to take credit. Neither of them, how-
ever, nor many other observers, took account of the familiar
phenomenon of the pre-existing trend. If, for example, all
the 9-year-old children in a town are given a banana-a-day
as part of their regular diet for a year following the ninth
birthday, they will on average gain a measurable number of
pounds and inches. Anyone who claimed that eating bananas
was the principal reason they grew would be ridiculed. Ten-
year-olds are bigger and taller than 9-year-olds. Though
eating bananas might be part of the reason, it would be but
a small part.
It turns out that both New York City and California had
been experiencing crime rate declines for several years before
their self-heralded policy changes occurred. Continuation of
the pre-existing trends is as likely an explanation for the
crime rate drops as were zero-tolerance policing and three
strikes.
New York’s innovations go under other names—order main-
tenance policing, misdemeanor policing, broken windows po-
licing (Harcourt 2001). The core elements, whatever name is
used, were increased arrests of misdemeanants and aggressive
patrol. Since zero-tolerance policing is the term in widest

use, I use it here. Sharp reduction in the homicide rate was
the most often claimed evidence of the crime-reducing effects
of zero-tolerance policing, but in retrospect that evidence
is not very compelling. Franklin Zimring and Jeffrey Fagan
showed that non-gun homicides began a steady decline in
1987. Gun homicides increased rapidly in the late 1980s and
early 1990s and then began their steep decline. This contrast
is important because non-gun homicides are typically associ-
ated with powerful emotion, alcohol, domestic violence, and
the structures of everyday life. The social conditions that
give rise to these homicides change slowly. Gun homicides,
by contrast, are associated with street crime, gangs, and drug
markets, and as gangs and markets change, homicide patterns
change with them. Thus, if evidence is wanted of crime rate
changes associated with fundamental changes in values, non-
gun homicides are the place to look, and their decline pre-
dated zero-tolerance policing by almost a decade.
What needs explaining is why gun homicide rates went
up in the late 1980s rather than why they went down in
the mid-1990s. Benjamin Bowling (1999) showed, as Alfred
Blumstein (1993) had earlier proposed, that explosion of the
crack cocaine markets in the mid-1980s, and their stabiliza-
tion in the 1990s, parallels the national and New York City
crime trends of that period.
One way to assess whether New York homicide rates were
reduced as a result of zero-tolerance policing is to compare
New York’s homicide rate trends with those of other large
American cities during a comparable period. Here again, the
data urge caution about claiming too much from zero-toler-
ance policing. Figure 5.8 shows homicide rates for ten Ameri-
can cities from 1990 to 2000. The rates have been standardized
to the 1990 rates. That means that the 1990 rate, whatever
it was, is assigned the value 100 and subsequent changes are
shown as percentage changes from 100. This makes trend
comparisons easier to see.
The patterns shown in figure 5.8 vary in detail, partly
because the numbers of homicides in some cities are small
enough that random statistical variations can produce sig-
Figure 5.8 Homicide rates for ten American cities, 1990–2000,
standardized to 1990 = 100
Source: Sourcebook of Criminal Justice Statistics.
nificant year-to-year shifts. Nonetheless, the broad patterns

are similar. Homicide rates in every big city, including New
York City, began to drop before zero-tolerance policing was
initiated in 1994. New York’s 1990s decline in homicide rates
is among the steepest, but, perhaps ironically, it shares that
distinction with San Diego, a city whose police department
publicly rejected the zero-tolerance model and prides itself
on its community policing approach. Homicide rates in Hous-
ton, a city not notorious for innovativeness in policing, fell
just about as much. Judy Greene, analyzing the city-by-city
data in a different way, reached much the same conclusion
(Greene 1999). Bernard Harcourt, in the most exhaustive anal-
ysis of the evidence for what he calls broken windows polic-
ing, concludes: “After reviewing the social-scientific data,
replicating a key study, and closely scrutinizing the empirical
evidence in New York City, Chicago, and other cities, I find

there is no good evidence to support the broken windows
theory” (2001, p. 7). John Eck and Edward Maguire, in an
exhaustive survey of evidence on the effects of policing on
crime rates in the 1990s, concluded zero-tolerance policing
“cannot be given credit for the decline in homicides in New
York City” (2000, p. 235). So the evidence for the crime-
reducing effects of New York’s zero-tolerance policing is
much weaker than is generally recognized.
Three-Strikes in California Likewise the evidence for Califor-

nia’s three-strikes law. Figure 5.9 shows homicide, rape, rob-
bery, motor vehicle theft, and burglary rates for California
from 1980 to 2000, with some multiplied and others divided
by ten. Declines in rates for all those crimes began long before
the three-strikes law took effect and continued long after-
ward. The pattern is the same as figure 5.6 showed for the
entire country: rates for all offenses fell from 1980 through
Figure 5.9 Crime rates (per 100,000 population), selected offenses,

California 1980–2000

1985–86, rose through 1990–92, and fell steadily thereafter.
Figures 5.10 and 5.11 show homicide and robbery rates for
the five most populous states including California, standard-
ized to 100 in 1980, for the period 1980–2000. Those states
vary widely in their sentencing policies and only California
enacted a widely drawn three-strikes law. Compared with the
other four most populous states, figures 5.10 and 5.11 show
nothing special about California crime rates before or after
the three-strikes law took effect. In both figures, crime rates
declined in all five states with Illinois showing lesser declines
and the patterns for the other four being indistinguishable.
This conclusion that the three-strikes law had little effect
on crime rates in California is consistent with most assess-
ments. Experience through 2000 has been summarized by
Franklin Zimring and colleagues. Application of the three-
strikes law depended on discretionary decisions by prosecu-
tors to charge defendants under its provisions. In some coun-
ties, prosecutors seldom filed three-strikes charges. Others
Figure 5.10 Homicide rates, standardized to 1980 = 100, five most

populous states, 1980–2000

Figure 5.11 Robbery rates, standardized to 1980 = 100, five most
populous states, 1980–2000
did so often. The counties where charges were least often

filed had the largest crime rate drops, those where they were
most often filed had the smallest. Zimring and his colleagues
conclude: “The decline in [California] crime [rates] after the
effective date of the three-strikes law was not the result of
the statute” (2001, p. 101). A sizable number of other studies
have reached similar conclusions (Males and Macallair 1999;
Turner et al. 1999).
This conclusion that changes following adoption of major
new crime control initiatives are largely attributable to con-
tinuation of preexisting trends is not novel. One of the best-
known demonstrations of that phenomenon occurred in Cali-
fornia in connection with the 1976 adoption of its Uniform
Determinate Sentencing Law. In enacting this law, which
abolished parole release, California became the first large state
to adopt “determinate sentencing.” The earliest evaluations
compared sentences imposed during the year before and the

year after the law took effect, and reported that the effects
included an increase in the proportion of convicted felons re-
ceiving prison sentences (e.g., Lipson and Peterson 1980). This
is what conservative proponents of the law wanted (Messinger
and Johnson 1978), and it was hailed as a sign of success.
When another research team looked at sentencing patterns
for three years before and three years after the law took
effect, it became clear that the percentages of felons receiving
prison sentences had increased each of the six years. The
rate of increase was no greater after the law took effect than
before (Casper, Brereton, and Neal 1983). The law’s apparent
effects turned out instead to be merely the continuation of
preexisting trends. A few years later, evaluators of North
Carolina’s Fair Sentencing Act, which took effect in 1981,
concluded that apparent effects of that law’s implementation
were instead the continuation of preexisting trends (Clarke
1987).
The simplest, and therefore best and likeliest, explanation
for California’s experiences with determinate sentencing in
the 1970s and the three-strikes law in the 1990s is that broad-
based changes were underway in values and norms that
caused both the changes in laws and the behaviors that the
legal changes appeared to produce. A rising tide lifts all boats.
Here is what likely happened. Beginning in the 1970s, Cali-
fornians began to become less sympathetic to offenders and
more moralistic and judgmental about wrongful behavior.
As a consequence, prosecutors became more aggressive in
their handling of cases, judges became more severe in their
sentencing, sending increasing percentages of offenders to
prison, defense lawyers became less sympathetic to their cli-
ents, and probation officers became less likely to look the
other way when probationers violated conditions. Once these
changes started, they became a trend, and the percentage of
offenders receiving prison sentences grew each year. In the
midst of these normative and behavioral changes, both sens-
ing the shifts in their constituents’ attitudes, and feeling the
same way themselves, legislators enacted laws that were less

sympathetic to offenders’ interests. In the following years the
percentages of offenders sent to prison went up, seemingly as
a result of the law change but actually as a result of the
underlying normative changes and the associated behavioral
changes. The changes in sentencing patterns would have
happened even if the Determinate Sentencing Law had not
been adopted. That is what happened in the 1970s in many
other states that did not radically overhaul their sentencing
laws. Figures 5.10 and 5.11 tell the same story about the 1994
three-strikes law.
International Crime Trend Comparisons

In the preceding subsection I showed that there are good
reasons to doubt that harsh crime control policies caused
recent falls in the American crime rate. One way to test
whether that is right is to compare U.S. crime trends in the
1990s with those in other countries. This can be done in two
ways, and both suggest that U.S. crime trends correspond to
those of other countries that did not adopt U.S.-style policies.
Crime trends can be compared using trend data from the
International Crime Victimization Survey. Crime and impris-
onment rate data from the United States and Canada can
be compared. Both these comparisons support the same two
conclusions: U.S. crime rate trends closely parallel those of
other Western countries, and crime rates move indepen-
dently of imprisonment rates.
Figure 5.12 shows trends in total crime rates for five coun-
tries as shown by the International Crime Victimization Sur-
vey. Because the ICVS is repeated at three- and four-year
intervals, the only data available are for the four surveys to
date. Nonetheless, they show exactly the same trends as offi-
cial data from these countries show. For the United States,
rates rose from 1988 to 1991 and fell in 1995 and again in
1999. Other countries’ downturns as shown in official data
began later but by the mid-1990s crime rates were falling
almost everywhere. And so figure 5.12 shows.

Figure 5.12 Trends in crime as measured by the ICVS, 1988, 1991,
1995, 1999 (offenses per 100 inhabitants)
Source: International Crime Victimization Survey, 1989, 1992, 1996,
2000.
Figure 5.13 shows homicide, robbery, and imprisonment

rates per 100,000 population in Canada and the United States
since the early 1980s. A 2001 report of the Canadian govern-
ment statistics agency on “Crime Comparisons between Can-
ada and the United States” observes in its opening paragraph:
“Despite differences in rates, trends in crime between the
two countries have been quite similar over the past twenty
years” (Gannon 2001, p. 1). Violent crime rates in Canada
are generally lower and property crime rates higher. Figure
5.13 shows homicide and robbery trends because both are
serious offenses and trend data are not especially vulnerable
to distortion by changes in recording practices. The patterns
for the two offenses are the same in both countries: an early
1980s peak followed by a mid-1980s trough, followed by a
second peak around 1990 and steady declines afterwards.
Imprisonment rates, however, are starkly different. The Ca-
nadian rate was flat throughout the 1990s, falling to its lowest
Figure 5.13 Violent crime and imprisonment rates for the United
States and Canada (per 100,000 inhabitants)
Source: FBI: Uniform Crime Reports and Bureau of Justice Statistics;
Canadian Centre for Justice Statistics.
level of the decade in 2000. The U.S. rate rose continuously.

In 1990, the Canadian imprisonment rate was about a quarter
of the American rate. In 2000 it was a seventh. And yet both
countries experienced a decade-long decline in crime rates.
It makes you wonder.
I draw two lessons from this survey of evidence on the
effects of changes in crime control policy on crime rates and
patterns, both consistent with this book’s sensibilities theme.
The first is that policy changes, and their putative effects,
are both often the consequences of previous broad-based
changes in social norms and attitudes. Such underlying changes
in California in the 1970s explain why some people proposed
and others enacted particular policies, and they explain why
practices changed in particular ways before and after the law
took effect. The three-strikes law was adopted in California
and zero-tolerance policing in New York when they were

because shifts in attitudes and norms made their adoption
seem sensible and right and produced their effects.
The second lesson is that people were predisposed to be-
lieve the changes would produce their sought-after effects.
Even though the continuation of preexisting trends hypothe-
sis was at least as plausible as the hypothesis that the innova-
tions “worked,” only the latter seems to have occurred to
those involved.
Put differently, changes in relevant sensibilities made the
policy changes appear desirable, made achievement of their
intended effects likely, and made critical assessment of the
cause-and-effect processes unlikely.
Does Crime Policy Follow

Drug Policy Cycles?
Crime policies and practices are affected by changes in sensi-

bilities in much the way drug policies are. Because, however,
U.S. drug policy historians have documented three cycles in
the past two centuries and crime historians describe only a
single U- or reverse-J crime rate curve in the same period,
followed by the 1960–90 crime rate rise, the analysis must
be by analogy.
The validity of Musto’s drug policy cycles analysis, first
offered in the 1973 edition of his classic book The American
Disease, can be tested against the events that followed. Here,
again, are Musto’s predictions of what happens after drug
use peaks and begins to decline:
[1] Public opinion turns against drugs and their acceptabil-
ity begins to evaporate. [2] Gradually, drug use becomes
associated, truthfully or not, with the lower ranks of soci-
ety, and often with racial and ethnic groups that are feared
or despised by the middle class. [3] Drugs become seen
as deviant and dangerous and become a potent symbol
of evil. [4] Trailing behind this decline come large-scale
legislative and law enforcement efforts . . . aimed at cur-

tailing drug sales and use through energetic prohibition
and enforcement and ever-harsher punishments against
sellers and users. (Kagan 1989, p. 7)
Viewed as hypotheses, no informed person can seriously

disagree that the four numbered propositions were con-
firmed. Here I pithily summarize.
Hypothesis 1: After drug use in the United States began
a steep long-term decline in 1979–80, public opinion, evi-
denced by poll results and support for the War on Drugs,
turned decisively against drugs and their acceptability.
Hypothesis 2: The disproportionate focus of antidrug ef-
forts on racial and ethnic minority groups is shown by
targeting of street-level drug law enforcement on minor-
ity, principally black, inner-city dealers, by stereotyping
in the entertainment media, by a focus on crack cocaine
sold by blacks rather than powder cocaine sold by whites,
and by the disproportionate presence of blacks and His-
panics among imprisoned drug dealers.
Hypothesis 3: The hype and emotionalism of antidrug ef-
forts demonstrate how drug use increasingly was per-
ceived and portrayed as deviant and dangerous.
Hypothesis 4: Long after drug use began a steep decline
in 1979–80, the federal Anti-Drug Abuse Act of 1986 es-
tablished decades-long mandatory minimum sentences
and the 100-to-1 crack/powder cocaine rule, in 1988 the
first federal Drug Czar was authorized and the War on
Drugs was declared, and in 1989 the arrest rate for drug
crimes reached six times the 1980 level.
Figure 5.14 shows self-reported drug use data for people

aged 12 to 17, and 18-to 25-year-olds from annual surveys
sponsored by the federal government. The six curves repre-
sent the percentages of people in those two age groups who,
over a quarter of a century, said they had used heroin, co-
caine, or marijuana in the preceding year. For marijuana and
heroin (and also, though not shown here, for amphetamines,
Figure 5.14 Estimated drug use by age group, selected years,
1974–2000
alcohol, and nicotine) self-reported use peaked around 1980

and declined rapidly thereafter, eventually settling at much
lower plateaus. The cocaine pattern is similar though the
peak was a little later. Since the mid-1990s, rates have stabi-
lized well below the peaks, and there were slight downturns
in 2000.
That use of so many different addictive or dangerous sub-
stances peaked at about the same time is striking. Something
in American sensibilities changed in the late 1970s. The ef-
fects of that change affected consumption not only of dramat-
ically dangerous and deviant substances like cocaine but of
butter, eggs, and whole milk, all things traditionally associ-
ated in American popular imagery with health and sturdi-
ness. The sensibilities change certainly reflected heightened
concern about personal health. Probably also it included a
growing sense of the importance of personal responsibility

for one’s life and well-being, and growing judgmentalism
about the weakness or irresponsibility of people who are
dangerously self-indulgent and personally irresponsible.
All of these changing attitudes and beliefs fit Musto’s model
of changing sensibilities toward drug use, and they help ex-
plain both why hard drug use peaked when it did and began
to decline, and why the predicted changes in beliefs and
policies occurred. And they did occur. The harshest policies
were not adopted until 1986, three-to-six years after use
peaked, the first Drug Czar was appointed only in 1988, and
the highest arrest rates for drug offenses were reached only
in 1989.
If reduction in drug use, and its related social ills, were
the primary aims of drug policy, America turned the corner
in its drug war in 1979–80, nearly ten years before the Drug
Czar arrived and policies became harshest. All of this is as
Musto predicted, and it is not surprising if the drug war
is seen not in instrumental terms but in expressive terms.
Sensibilities affecting drug use changed, and the new policies
expressed, reinforced, and celebrated the changes.
But there is more to figure 5.14, also consistent with Mus-
to’s predictions. He predicted that, after a period of increas-
ing self-righteousness in public attitudes and zealousness in
law enforcement, the tension would ease. Laws would be
enforced less vigorously, policies become less stereotyped,
and voices expressing concern about the side effects of anti-
drug policies or about the issues raised by drug use become
louder.
Figure 5.14 shows this too. In 1988 Baltimore Mayor Kurt
Schmoke began openly to call for reconsideration of U.S.
drug policy and soon was followed by conservative intellec-
tuals like Milton Friedman and William F. Buckley. More
recently, and more vocally, conservative Republican New
Mexico Governor Garry Johnson repeatedly called for funda-
mental reconsideration of U.S. drug policies. In 1995 the U.S.
Sentencing Commission called for repeal of the 100-to-1 pol-
icy that punished mostly black sellers of 10 grams of crack

cocaine as severely as dealers of 1,000 grams of powder co-
caine. In 1994 conservative Utah Republican Senator Orrin
Hatch began to call for reconsideration of mandatory mini-
mum sentences, and in 1994 the U.S. Congress enacted
“safety-valve” legislation that reduced the severity of federal
drug law penalties for many first offenders; states such as
Washington enacted comparable laws.
Drug courts, premised on the ideas that treatment can work
and that drug dependence is a chronic relapsing condition,
were established in many states. Though most began as diver-
sion programs for non-violent first offenders, as the decade
progressed many began to work with people with lengthy
records or charged with serious crimes. From the mid-1990s
onwards, referenda were adopted, initially in conservative
states like Arizona and California, permitting medical use
of marijuana, and in 2000 California again by referendum
adopted policies preferring referral of first- and second-time
drug offenders to treatment rather than to prosecution. Con-
servative New York Republican Governor Pataki has called
for and proposed substantial weakening of New York’s noto-
rious Rockefeller Drug Law. The tide, as Musto predicted
thirty years ago would happen, is turning.
So much for drugs. The question is whether patterns like
those Musto foresaw for drug policy are likely also to affect
crime policy. The answer is “yes,” as figure 5.15 shows. Fig-
ure 5.15 presents the same data as figure 5.6 for homicide,
robbery, motor vehicle theft, and burglary rates from 1960
to 2000, except that it also shows high points in the toughen-
ing and softening of U.S. crime policies. It may be right to
consider U.S. crime rates to have peaked around 1980, per-
haps not coincidentally when drug use peaked, even though
there was a five-year reversal in the late 1980s. The harshest
policies were adopted in the fifteen years after 1980. The
prison population grew unremittingly, initially because more
drug offenders were sent to prison but also because the
chances that offenders received a prison sentence grew enor-
mously, and more recently because prison sentences became
much longer (Blumstein and Beck 1999). The federal manda-

Figure 5.15 American violent-crime-rate pattern with policy changes
tory sentencing guidelines took effect in 1987. Under federal

financial pressures more than half of the states adopted truth-
in-sentencing laws requiring many violent and drug offend-
ers to serve at least 85 percent of their sentences. In 1993
Washington State passed the first three-strikes-and-you’re-
out law. California’s much broader law followed in 1994 and
nearly half the remaining states followed within a year or
two. In 1994 the U.S. Congress made the federal death penalty
applicable to more than fifty additional offenses. Megan’s
laws requiring registration of sex offenders and providing
notification of their presence spread across the country in
the mid-1990s.
So the first part of Musto’s post-peak model appears also
to apply to crime. That is not surprising. It is easy to imagine
parallel psychological, social, and political developments and
comparable changes in sensibilities. Through the 1970s, for
crime as for drugs, there was spirited debate about causes
of unwanted behavior, about the comparative merits of treat-
ment or punishment, and about the competing values that
underlie policy choices, and there was deep pessimism about
the likely effectiveness of state action. I describe in chapter
7 how now-nearly-forgotten debates raged in the 1970s about
issues such as preventive detention and prosecutorial appeals
of sentence that seem in retrospect almost quaint. Today it
seems obvious to most people (though in both cases not to
me) that offenders may fairly be held before trial because of
additional crimes they might commit and that it is fair for
prosecutors to appeal sentences they do not like.
During the 1970s, due process liberals were still arguing
in the courts for greater procedural fairness for criminal de-
fendants and still winning respectably often. The prisoners’
rights movement was winning numerous victories in the
courts. Sentencing reform initiatives were concerned primar-
ily with equal and consistent treatment of offenders and pro-
ponents only occasionally claimed that changes would reduce
crime rates. In other words, people were spiritedly disagree-
ing about the timeless questions of how collective interests
in security and safety can be pursued while taking account
of the legitimate interests of alleged offenders. That does not
mean that due process liberals were winning most of the legis-
lative and court battles or that law-and-order conservatives
were mostly losing, but that disagreements were vocal and
vigorous and it often was not obvious which side would win.
By the mid-1980s and continuing into the mid-1990s, the
picture was very different. Open political debate on crime
issues stopped. “Soft on crime” became the most damaging
and dreaded epithet with which one electoral candidate could
tar another. Rather than risk the near-certain defeat that
label—if successfully attached—would bring, candidates
competed to assure voters of their toughness. President Bill
Clinton was one of many Democrats who eventually won
congratulations for deciding not to let the Republicans get
to his right on crime (S. Walker 1998; Friedman 1993). The
effect of all this was that voices were stilled and opposition to
policies that were seen as ham-fisted, cruel, or misconceived
moved behind the closed doors of legislative committee
rooms.
Capturing the ethos of the time, a senior U.S. Justice Depart-
ment official, whose position gave her influence over funding
of Crime and Justice, a scholarly journal Norval Morris and
I started, once told us, “People are either for victims or for
criminals,” and, chillingly at the time, “you’re for criminals.”
That forced choice is nonsense, but it was widely held to be
necessary in the 1980s and early 1990s.
A human mind can coherently and responsibly contain
the ideas that the fears and sufferings of victims should be
recognized and sympathetically addressed, that alleged of-
fenders should be treated fairly and sympathetically, and
that both groups are entitled to be treated with concern and
respect. Nonetheless, saying that in election campaigns and
on legislative floors, and acting as if it were true, exposed
politicians to the accusation that they were “for criminals,”
and few would take that risk.
The second half of Musto’s post-peak hypothesis seems also
to apply to crime. Just as is happening with drugs, monolithic
anticriminal views are breaking down, and competing values
are again being urged. In 1996 in Koon v. United States, 518
U.S. 81, the U.S. Supreme Court made application of federal
guidelines less rigid and returned discretion to federal judges.
In 1997 the California Supreme Court, all of its members
appointed by conservative Republican governors, in People
v. Alvarez, 14 Cal. 4th 968, fundamentally weakened Califor-
nia’s three-strikes law by giving judges authority to decide
it ought not to be applied in any particular case. In 2001 in
Andrade v. State, 270 Fed. 3rd 743, a federal court of appeals
held California’s three-strikes law unconstitutional when ap-
plied to the case of a man convicted of a $150 theft from
K-Mart. In July 2002 the California Supreme Court in People
v. Mower, 122 Cal. Rptr. 2d 326 (200), held that California’s
medical-use-of-marijuana law precluded prosecution for pos-
session or cultivation of marijuana by a qualified patient or
physician. Throughout the criminal justice system, rehabili-
tative programs, especially concerning drugs, rapidly gained
support and funding. Throughout the country, restorative
and community justice programs were established at grass-
roots levels and within state and county bureaucracies. Most
were concerned more with problem solving and dispute reso-
lution than with punishment and blaming. Sentencing and
sanctions systems are fragmenting in the United States.
Musto’s cycles model thus equally well predicted the drug
and crime policy progressions of the past century. If past be
prologue, the model is likely also to predict future policy
trends, if we let it. But, if we can learn from the distant and
recent pasts, and come to a better understanding of the ways
sensibilities are shaped, we may be able to do better and
avoid the excesses that inevitably future generations will
come to regret.
Moral panics had a role to play in the recent evolution of
crime control and penal policy. Overall, the principal back-
ground to policy changes was the interaction between cycles,
sensibilities, and policy. Against that backdrop, however,
moral panics had much greater power, and longer term ef-
fects, than otherwise they might. Prevailing sensibilities pro-
vided especially rich soil in which to plant the most emotion-
based crime control ideas of the past two decades.
In our time, cycles, sensibilities, and moral panics coincided
in ways that produced current crime control and penal poli-
cies. Crime rates rose, and steeply, for an extended period.
The world changed with globalization, economic restructur-
ing, fundamental social changes, and increased population
diversity. All these things raised fears and anxieties that were
in part displaced to people (criminals) and things (crime and
disorder) that were ready objects of hatred and derision.
How events would have unfolded had American crime rates
not begun their sharp rise in the 1960s and their sharp fall
more recently is unknowable. We do know, however, that
horrible crimes have been committed in all periods and that
moral panics and sustained periods of extraordinarily harsh
crime and punishment policies have not always followed.
Examples that come to mind in my no longer short life include
Richard Speck’s 1966 killings of eight nurses in Chicago,
Richard Gacy’s and Jeffrey Dahmer’s separate killings of tens
of vulnerable young men, the Tate-LaBianca killings by Charles
Manson and his followers. All of these were horrible crimes
that filled people with horror and revulsion, but none has
been memorialized in policy as Len Bias, Polly Klaas, and
Megan Kanka have been.
Whatever did or did not happen in other times, we know
what happened in our time and we have a good sense of
why. Knowing these things, we should be able to devise ways
to undo the excesses of recent crime control and punishment
policies.

6
Formerly Unthinkable Policies
Contemporary penal sensibilities have led public officials to

propose and support policies that would have been unthink-
able in other recent times and in other countries today, even
though they are commonplace in the United States, and have
been enthusiastically supported by public officials, policy
intellectuals, and the general public.
One way to show how much modern American policies
differ from those in earlier times and other places is to note
how often they call for imposition of disproportionate pun-
ishments. Most normative theories of punishment, and the
punishment jurisprudences of most Western countries, attach
importance to the ideas that more serious crimes warrant
harsher punishments than do less serious crimes and that
comparably serious crimes warrant comparably severe sanc-
tions. This is often in shorthand referred to as the proportion-
ality principle: punish like cases alike and unlike cases differ-
ently.
Proportionality in punishment is not a modern idea. From
retributive and utilitarian premises that are often seen as
nearly opposite, Immanuel Kant and Jeremy Bentham, the
archetypes of the two principal contending modern views
about punishment, both proposed that just punishments
should be proportionate to the seriousness of crimes. Kant
favored proportionality on a lex talionis principle. There is
a moral imperative to punish wrongdoers in proportion to
the nature of the wrong they have done. Bentham argued
for proportionality as a way to optimize the deterrent effects
of punishments. It was not only philosophers who thought
proportionality important. Thomas Jefferson in 1776 averred:
141
“Punishments I know are necessary and I would provide
them, strict and inflexible but proportioned to the crime,”
and proposed legislation to that end in the Virginia Assembly
(N. Walker 1999, p. 39). Blackstone in his Commentaries on
the Laws of England insisted that “punishment ought always
to be proportioned to the particular purpose it is meant to
serve, and by no means to exceed it” (N. Walker 1999, p. 40).
The principal disagreements in our time have been about
details. People such as Andrew von Hirsch (1993) think that
proportionality is the central moral value underlying any
just punishment system and that the seriousness of the crime
should be the primary determinant of punishment. Others
such as Norval Morris (1974) think it is one among several
important values and sets outer limits on justifiable punish-
ments. Still others such as H. L. A. Hart (1968) think that
proportionality ideas are important for practical reasons; a
system that regularly imposed disproportionate punishments
would defy conventional morality, create a sense of public
insecurity, and lose credibility.
Classical utilitarians such as Jeremy Bentham (1843) argued
that human rights claims based on moral ideas were “non-
sense on stilts,” that human happiness should be the measure
of good policies, and that pursuit of abstract ideals like pro-
portionality was not inherently important. Nonetheless, he
urged that more serious crimes merit harsher punishments,
to give offenders an incentive to choose to commit the less
serious of possible alternate crimes, and that punishments
should increase with the value or seriousness of the crime,
so that the penalties for crimes would outweigh the likely
benefits. He also urged a punishment-limiting principle of
“parsimony”: no punishment can be justified unless, taking
account of both the criminal harms it might avert and the
suffering it imposes on the offender, it produces net happi-
ness or utility (Morris 1974; Tonry 1994).
Yet modern American punishment policies regularly re-
quire imposition of disproportionate penalties, and many pol-
icy makers and some influential scholars (Wilson 1995; Ben-
nett, DiIulio, and Walters 1996) support them. Sentencing

laws and guidelines in the federal system and in many states
require harsher penalties for drug crimes than for many as-
saults, rapes, and homicides. California’s three-strikes law
demands twenty-five-year and longer sentences for some
people convicted of trivial third felonies even though first-
time rapists and robbers receive much shorter sentences.
Some people justify these policies on the rationale that
punishments may legitimately reflect popular demand for
vindictive and shaming punishments (e.g., Kahan 1998a, b),
others (Bennett, DiIulio, and Walters 1996) on the rationale
that public opinion demands harsh punishments and so be
it. Putting Kahan’s rationale differently, the important rela-
tionship is not between crime and punishment, or between
punishment and public safety, but between punishment and
public edification. This is a profoundly radical idea that re-
moves any considerations of proportionality, fairness, or jus-
tice to the offender from the calculus. The public opinion
argument is little less radical because, as the cycles and moral
panic literatures discussed earlier make clear, public opinion
can be impulsive, ill-considered, and cruel. Philosopher Ron-
ald Dworkin argues that justice requires all people to be
treated with “equal concern and respect” (Dworkin 1986).
Punishment policies premised either on public appetites for
debasement of offenders or on knee-jerk responses to opinion
polls do not do that. Nor do policies that ignore proportional-
ity concerns.
There is nothing inherently wrong with laws that punish
repeat violent offenders more severely than first-timers, mon-
itor the whereabouts of sex offenders, or treat one kind of
drug trafficking more harshly than another. Reasonable peo-
ple can differ about their desirability and their detailed provi-
sions. However, during moral panics and the downswings
in cycles of intolerance, many people tend to exaggerate the
scale of problems, become self-righteous and intolerant, and
support views and policies that in less hysterical times they
would reject. Many people come to believe things that at
other times they would not believe, and to say things that
at other times they would not say.
FORMERLY UNTHINKABLE POLICIES 143

We live in a time when a series of moral panics about sexual
and violent crimes and about drugs have coincided with
downturns in crime rates and drug use, which has meant
that the short-term effects of moral panics have interacted
with the effects of intolerant attitudes associated with periods
of long-term decline in drug use and crime. Repressive con-
temporary policies are the result. The hitherto unthinkable
became not only thinkable but also acceptable.
Law professor Dan Kahan, for example, has offered a num-
ber of policy proposals that would have been close to un-
thinkable thirty years ago. Below I discuss one of these—
deliberate humiliation and debasement of offenders serving
community penalties—at some length. Here I sketch another.
Kahan has argued that the ex post facto principle, the idea
that people should not be subject to criminal prosecution and
punishment for actions that were legal when they occurred, is
wrong and ethically unnecessary. The ex post facto principle
has been uncontroversially accepted in Western countries as
an ethical necessity since the Enlightenment. It has been
embedded in every major human rights protocol, including
the U.S. Constitution, most American state constitutions, the
European Convention on Human Rights, and the United Na-
tions Declaration on Human Rights. Kahan (1997) proposes
that judges should be able to reinterpret existing laws so as
to punish behaviors that were not criminal when they were
committed. The logic is that a person who does something
morally wrongful knows it, and if a judge retroactively de-
cides to treat that act as a crime, so be it; the defendant
should have known better and deserves what happens to him.
Kahan uses sexual exploitation—“sleep with me and I will
support you and your children; otherwise, leave”—to illus-
trate his argument. Such stories strike emotional chords. It
is not fair for people to manipulate others to obtain sexual
intimacies, even though legislatures may not have made it
criminal. And it is easy to empathize with vulnerable people.
Kahan’s logic, however, applies equally to retroactive crimi-
nalization of homosexuality, adultery, drug use, blasphemy,
or sedition. Just as one judge could decide that use of eco-
nomic pressures to win sexual favors is wrong, that the per-
son exerting pressure should know it is wrong, and that laws
should retrospectively be reinterpreted to make it criminal,
another judge could decide that homosexual acts are wrong,
known to be wrong, and should be treated as crimes. Few
people would want to live in a society where judges held
such powers.
The realms of right, ethical, moral, and legal conduct have
never been perfectly congruent. If legislatures want to crimi-
nalize sexual intimacies with people subject to unusual psy-
chological or economic pressures, that can be done. Legisla-
tive definition of wrongful conduct as criminal is the critical
moment only after which citizens can fairly be presumed to
know what actions constitute crimes.
Why have proposals such as Kahan’s taken shape now?
Because the sensibilities of our times have made it easier to
be judgmental, to stereotype wrongdoers, and to look for
punitive and authoritarian solutions to difficult problems.
The past few decades have been a time of harsh penal sensibil-
ities associated with declines in drug use and crime. With
the passage of time, such proposals will again become un-
thinkable.
Changing sensibilities affect the ways we see and react to
things, and how we think about them, but that does not
necessarily mean that things we do under their influence will
be bad or inhumane. Contemporary penal sensibilities may
foster excessively vindictive and uncaring responses to crime
and criminals, but they may also have made more construc-
tive policy changes possible. Contemporary programs to ad-
dress victims’ interests and needs, for one example, and the
move to criminalize domestic and marital violence, for an-
other, may be products of contemporary penal sensibilities.
I return to these more positive effects in the final paragraphs
of this chapter. Here, however, my focus is on vindictive
and inhumane policies.
What is “unthinkable” to one person may not be to another,
so I offer examples in this chapter of modern policies that
reject conventional views in the United States at other times

and in other countries today. The first is the federal sentenc-
ing guidelines’ “real offense sentencing” provision under
which people are punished for crimes of which a jury acquit-
ted them. Few people in other times and places would ap-
prove of punishing people for crimes of which they had been
found “not guilty.”
The second is a provision of the Virginia sentencing guide-
lines that, all else being equal, punishes young offenders,
for example, 18-year-olds, more harshly than 35-year-olds.
In all other Western countries today, and in every U.S. juris-
diction until recently, comparatively gentler handling of
young offenders is or was generally believed to be wise and
just.
The third is Professor Kahan’s “punishment incommensu-
rability” theory of punishment as debasement (Kahan 1997,
1998a, b). Kahan argues that community penalties cannot
substitute for prison sentences because they are insufficiently
debasing. Few informed people in other times and places
would find such a proposition congenial.
The three examples are merely illustrative and were picked
because they are largely unknown except among specialists.
In discussing them I first describe them and then summarize
the relevant scholarly literatures that might arguably support
them. Lest it be thought that the three examples are aberra-
tions or anomalies, of which every era has some, do not forget
that they have occurred in a time when the United States is
awash in overbroad and indiscriminate three-strikes, sexual
psychopath, and truth-in-sentencing laws, innovations that
have been widely emulated in no other countries (Tonry and
Frase 2001).
Real Offense Sentencing
The real offense sentencing provisions of the federal sentenc-

ing guidelines direct judges to calculate sentences on the
basis not only of offenses of which the defendant was con-
victed but also of offenses with which the defendant was

not charged, for which charges were dropped or dismissed,
and of which the defendant was acquitted. All these other
offenses are to be taken into account if the judge decides
the defendant committed them. In making this decision, the
judge is not bound by the general criminal law standard that
people can be found guilty only on the basis of proof “beyond
a reasonable doubt.” The decision need be made only on the
basis of the much lower standard that applies to civil law
matters, “more probably than not.” This means the judge
must punish a person even if the judge believes there is
only a slightly better than fifty–fifty chance that he or she
committed a crime at all.
Eighth Circuit Federal Court of Appeals Judge Philip
Heaney described how real offense sentencing works:
Under the guidelines, however, sentencing judges are rou-
tinely required to sentence offenders for “relevant con-
duct” which was not charged in an indictment or informa-
tion and which was not admitted in a guilty plea or proved
at trial. Indeed a court may also increase an offender’s
sentence for acts of which the offender was acquitted.
Uncharged conduct need not be proved beyond a reason-
able doubt, but only by a preponderance of the evidence.
(Heaney 1991, p. 209)
Judge Heaney slightly understated the provision’s force. Un-
der the relevant guidelines, the judge, not “may,” but “shall”
take account of uncharged and acquitted conduct in setting
the sentence.
Three things about this policy stand out. First, real offense
sentencing involves a radical rejection of basic ideas of fair-
ness. Affected offenders lose the benefit of the criminal law’s
requirement of proof beyond a reasonable doubt, the rules
of evidence, and constitutional procedural protections. These
are all devices created and intended to protect against the
possibility, or at least reduce the likelihood, that innocent
people will be convicted of crimes and made vulnerable to
the consequent stigmatization and punishment. The safe-
guards for defendants in federal sentencing hearings, in

which loss of liberty is in issue, are fewer and weaker than
those for defendants in lawsuits about minor car accidents
where, at most, money is at stake.
Second, a real offense sentencing policy exists in the federal
sentencing guidelines and nowhere else in the United States
or any other Western country. Every other American sen-
tencing commission that considered adopting such a policy
rejected it, sometimes as in Washington State (Boerner and
Lieb 2001), expressly forbidding judges to take account of
such alleged but unproven crimes. In other countries, includ-
ing Australia (Freiberg 2001), appellate courts have explicitly
forbidden sentencing judges to take such information into
account.
This is not to say that judges do not sometimes take account
of unproven criminal conduct when they impose sentences.
Under the indeterminate sentencing systems that were com-
mon in the United States before 1980, judges were supposed
to individualize sentences, and doing that was seen to require
consideration of any information about the offender that the
judge thought relevant. And it would be hard not to expect
that a judge who thought one defendant had assaulted a
spouse just once under great stress and another had done so
many times would take that difference into account in decid-
ing what to do. The critical difference is that the federal
guidelines require judges to do so. Indeterminate sentencing
allowed them to do so.
Third, despite real offense sentencing’s radical nature, nei-
ther the U.S. Sentencing Commission nor the federal appellate
courts have shown much ambivalence about it. Objections
to real offense sentencing have been frequent and vocal since
the U.S. Sentencing Commission issued its first proposed draft
guidelines, but the commission has never indicated that the
policy might or would be reconsidered. Federal appellate
courts, including the Supreme Court, have squarely upheld
it, even in relation to offenses of which defendants had been
acquitted.
Section 1B1.3 of the federal guidelines provides that sen-
tences “shall” be calculated on the basis of the offender’s

“relevant conduct,” or as the commission called it in explain-
ing the provision, his “actual offense behavior.” The rationale
was that the relevant conduct approach would offset efforts
by prosecutors to manipulate the guidelines. The federal
guidelines are expressed in a two-axis grid. One axis is a
scale of offense severity, the other of past criminal record.
The sentence should be selected from within a narrow range
indicated in the cell at which the two scales intersect. The
prosecutor’s ability to select and dismiss charges can deter-
mine which cell in the grid governs a particular case and
thereby can greatly narrow the judge’s choices. If, however,
because of the real offense sentencing policy, neither filing
nor dismissing charges necessarily affects the sentence that may
be imposed, the prosecutor in theory will have less influence.
In practice, the commission’s sacrifice of defendants’ protec-
tions did not work; prosecutors gained power under the fed-
eral guidelines (Heaney 1991; Schulhofer and Nagel 1997).
Statutory sentence maxima are the only absolute constraint
on the relevant conduct provision. If, for example, the statu-
tory sentence maximum for the charge of conviction is sixty
months, the applicable guideline range for that offense is
forty to forty-six months, and the applicable guideline range
for all relevant conduct is 108 to 110 months, no sentence
longer than sixty months may be imposed.
The U.S. Supreme Court has upheld the constitutionality
of real offense sentencing in general (Witte v. United States,
515 U.S. 389 [1995]) and in the case of offenses of which the
defendant was acquitted (United States v. Watts, 519 U.S.
148 [1997]). The rationale in Witte v. United States, on the
basis of Williams v. New York, 337 U.S. 241 (1949), is that
it was always so. In the rehabilitation-premised era of indeter-
minate sentencing, judges could take account of anything
they deemed relevant in making an individualized decision
in a particular case. That was the rationale of Williams v.
New York. The Supreme Court, however, seemed untroubled
by the segue from the court may take account in order to
individualize, to the court shall take account even though it
may not individualize.
Every state that adopted sentencing guidelines, and the
American Bar Association’s Standards on Sentencing, re-
jected real offense sentencing. There were two reasons: the
Caesar’s wife notion that justice should not only be done but
be seen to be done and the practical objection that it would
not reduce the power of prosecutors—plea bargaining prose-
cutors and defense lawyers will devise ways to defeat it
(Tonry 1996, chap. 3). The evidence on the latter point is
clear. From the guidelines’ earliest days (Federal Courts Study
Committee 1990) through their early maturity (Nagel and
Schulhofer 1992) to their maturity (Schulhofer and Nagel
1997), observers have agreed that the federal guidelines
shifted power to, not from, prosecutors. Most cases are dis-
posed of by negotiated plea agreements and various features
of federal sentencing law and policies have strengthened the
prosecutors’ hand. In other words, real offense sentencing
was unlikely to work and has not worked. When I describe
federal real offense sentencing to U.S. lawyers who are un-
aware of it, or to lawyers and judges outside the United States,
the reaction is always a combination of disbelief, disapproval,
and disdain.
Youth as an Aggravating Circumstance
In Virginia, youth is an aggravating circumstance that in-

creases the likelihood that a convicted person will be sen-
tenced to imprisonment. This stands on its head the main-
stream view for most of this century in the United States and
most Western countries that youth is a mitigating circum-
stance. The juvenile court, however beleaguered in our time
by changes in age eligibility and waiver-to-adult-court laws,
still exists in every American jurisdiction and is premised
on notions that kids are different and should be treated differ-
ently (Feld 1999). In Sweden, Finland, and much of Northern
Europe, the age of criminal responsibility, the age at which
acts can be treated as criminal, is 15 (Janson 2004; Kyvsgaard
2004). Older teenagers are dealt with by the regular criminal

courts, but most are sentenced to special programs for young
offenders and all are entitled to have the severity of their
sentences reduced because of their age. Social welfare agen-
cies deal with what would otherwise be crimes by youngsters
under 15; as a legal matter, their wrongful acts are not crimes.
In New Zealand, most juvenile cases must be handled by
means of consensual restorative processes (Morris 2004). In
Scotland, youth justice remains a branch of the social welfare
system (Bottoms and Dignan 2004). Virginia’s approach is
unique.
Rationales for why children should be treated more sympa-
thetically and gently than adults have varied over time and
space and between people. Sometimes the argument is that
young people are morally immature and, for that reason, less
fully responsible for their actions. Sometimes it is that young
people are developmentally immature and thus more mallea-
ble and “amenable to treatment.” Sometimes it is that many
people do things as teenagers, because they are teenagers,
that they will later regret, and that aging is all that is needed
to turn most into law-abiding citizens. Criminal records
would damage them and their life chances for no good reason.
Whatever the rationale, even the most aggressive attacks on
the juvenile court—in laws that make it easier for juvenile
court judges or prosecutors to waive young offenders to the
adult courts, laws that remove some offenses from juvenile
court jurisdiction altogether whatever the defendant’s age,
and laws that reduce the maximum age jurisdiction of the
juvenile court from 17 to something younger—leave a large
majority of cases traditionally handled by juvenile courts
within the courts’ jurisdiction (Zimring 1982, 1989; Bishop
2000). Virginia’s sentencing commission, however, rejected
the view that young people should receive different and
gentler handling.
Explaining what Virginia policy makers did requires a close
look at Virginia’s guidelines. The guidelines, like all U.S.
sentencing guidelines, specify whether a prison sentence
should be imposed and, if so, for how long. Like the federal
guidelines, the Virginia guidelines can be set out in a grid.

The recommended sentence is shown in the cell at which the
applicable offense severity row and criminal record column
intersect.
However, for people sentenced for drug offenses (14,193
in 1997), larceny (8,817), and fraud (4,725), a separate scoring
sheet indicated who among them should be sentenced to a
community punishment like probation or community service
rather than to a prison sentence. This is an important provi-
sion, since those three offenses constituted nearly three-
fourths (71 percent) of sentenced cases in 1997 (Virginia Crim-
inal Sentencing Commission 1997, p. 22). The scoring sheet
assigned points to various of the offender’s characteristics.
Some, such as prior arrests, convictions, and incarcerations,
relate to the offender’s past criminality. Others, such as sex,
age, marital status, and employment record, relate to personal
characteristics. A number of Virginia policy choices might be
questioned. Being male, unmarried, and unemployed warrant
“points” that increase the likelihood of imprisonment. Many
people might believe it unfair to punish someone more
harshly for those reasons.
Age counts against a defendant’s chances of avoiding
prison in two ways. It is important to know that in 1997
“nine points” was the cut line above which the guidelines
specify offenders should be sent to prison. First, the younger
the offender, the more points. An offender who is 19 or under
receives six points, and the number of points then declines
with advancing age: 20 to 27 (four points), 28 to 33 (three
points), 34 or older (no points). Second, any prior juvenile
commitment or incarceration counts four points. Thus, a 17-
year-old (six points) who previously was committed to a
juvenile institution (four points) receives ten points and
should go to prison. Being unemployed (one point), unmar-
ried (one point), and male (one point) also add points, so the
17-year-old described in the preceding sentence with those
characteristics (most will have them), receives thirteen
points, well above the nine-point cutoff. Even without a prior
juvenile commitment or incarceration, most 17-year-old drug
or larceny defendants will be unemployed, unmarried, male,
and pointed by the guidelines toward prison. So much for
any ideas in Virginia that young offenders deserve special
tolerance, solicitude, or the benefit of the doubt.
The juvenile aggravators are even more striking when com-
pared with other factors that get scored. No other factor is
weighted more heavily than being under age 28 and few
as heavily as a prior juvenile court record. The four points
assigned for a single juvenile commitment dwarfs those for
more serious records. Having four or more felony convictions,
for example, counts for three points. Having been incarcer-
ated as an adult five or more times counts as three points.
Compare two hypothetical offenders. An unemployed (one
point), male (one point), 18-year-old (six points) drug of-
fender who served a one-week juvenile commitment when
15 (four points), totals twelve points, and should go to prison.
An unemployed (one point), male (one point), 35-year-old
(no points), drug offender with four prior felony convictions
(three points) and five prior adult incarcerations (three
points) totals eight points and should be diverted from prison.
Most people, knowing no more, would doubt that most 18-
year-olds with a minor juvenile record should be sent to an
adult prison, but acknowledge that some possibly should.
My guess, however, is that few people would have equal
ambivalence about a 35-year-old with the record described.
There is a logic to the Virginia policies that derives from
the Virginia guidelines’ “selective incapacitation” rationale
(Virginia Criminal Sentencing Commission 1997, pp. 41–52).
If the only concern at sentencing were prediction of future
offending, if matters of fairness, justice, and youth policy
were deemed not relevant, Virginia’s antiyouth policies
might make sense. At least three robust research findings are
relevant.
First, criminal careers research instructs that most active
offenders age out of criminality by their mid-thirties (Blumstein
et al. 1986). Thus, 35-year-olds are less likely than younger
offenders to reoffend. This “justifies” the seeming anomaly
that older offenders with lengthy records receive no aggra-
vating points.
Second, developmental research in the United States and
other countries instructs that there are distinct age-crime
curves, with the prevalence of offending peaking in the mid-
teens for property offending and in the late teens for violent
offenders, with rapid drop-offs after the peaks (Farrington
1986). Thus, the older the (young) offender, the more likely
that he or she will soon desist from crime. This “justifies”
the anomaly that the aggravating points are inversely related
to the offender’s age.
Third, developmental research also shows that early onset
of serious delinquency is a strong predictor of both serious-
ness and continuation of offending (Tremblay and Craig
1995). The specialized vocabulary of this literature distin-
guishes between “adolescence-limited” offenders and “life-
course persisters” (Howell and Hawkins 1998). The former
are wild-oats teenagers for whom delinquency is a short-
term developmental phase and the latter are more persistent
offenders whose delinquency extends into adulthood. For
young offenders believed to be “adolescence-limited,” the
smartest course often is to let them alone, turn them over to
their parents, or deal with them informally. They will most
likely grow up to be law-abiding. Stigmatizing them with a
juvenile or criminal record may lessen the odds that will
happen. An early onset of delinquency, however, is one char-
acteristic that often distinguishes life-course persisters from
adolescence-limited offenders, even though many people
who commit offenses when young soon stop. This, arguably,
“justifies” the heavy weight given to a juvenile record.
Many objections could be raised to the Virginia scheme.
Punishing people more severely because they are male vio-
lates gender-neutrality norms. Punishing people more se-
verely because they are unmarried penalizes people as a con-
sequence of an inherently personal choice that most people
would ordinarily insist is not the state’s business. Punishing
people because they are unemployed may injure those al-
ready disadvantaged by lack of marketable skills, mental
disabilities, or high unemployment rates. Giving greater pu-

nitive weight to offenders’ ages than to their past criminality
violates basic and widely shared culpability notions.
But most strikingly, counting peoples’ youth against them
rather than for them sacrifices individual lives and futures
in the interest of aggregate statistical predictions. None of
the three bodies of knowledge on which Virginia’s policy
ostensibly rests supports it. First, criminal careers research
shows that the vast majority of offenders desist from crime
by their mid-twenties, so Virginia’s guidelines will result in
many offenders being held long after they would have ceased
offending anyway.
Second, the age-crime curves tell us that the highest offend-
ing rates are in the mid- to upper teenage years and that most
young offenders soon desist.
Third, the distinction between adolescence-limited offend-
ers (the large majority of young offenders) and life-course
persisters (a handful) is real, but based on hindsight. Exami-
nations of the records of 25-year-olds who committed crimes
as 16- and 17-year-olds make it clear that most desisted from
offending and a few did not. Looked at from the other way
round, we can not know who among current 16- and 17-
year-old offenders will persist and who will not. We can
make predictions, based on individuals’ personal characteris-
tics, but those predictions are not very good. Among those
predicted at 16 to be persisters, most will not be (Blumstein
et al. 1986).
It is important to remember that Virginia’s distinctive
youth-as-an-aggravating-circumstance sentencing policy has
nothing to do with moral ideas about offenders’ culpability
or the seriousness of their crimes. All else equal, including
the crimes people commit and their past records, the Virginia
policy treats younger people more harshly. In adopting it,
the Virginia commission was interested only in incapacita-
tion, and considerations of just treatment of offenders were
beside the point. The policy premise for punishing teenagers
more harshly than middle-aged adults was “selective incapac-
itation.”

Selective incapacitation was an idea that was promoted in
the early 1980s on the basis of RAND Corporation research
headed by Peter Greenwood (Greenwood with Abrahamse
1982) that claimed to show that researchers could successfully
identify high-rate serious offenders. If true, adoption of selec-
tive incapacitation policies would have permitted jurisdic-
tions to increase sentence lengths for high-rate offenders
while reducing sentence lengths for other offenders and,
overall, reducing prison populations and costs. The only dif-
ficulty, as the National Academy of Sciences (Blumstein et
al. 1986) soon showed, was that his predictions were based,
like those concerning “life-course persisters,” on hindsight.
Used prospectively, the predictions were too inaccurate to
be relied upon and would have resulted in lengthy confine-
ment of many people who would not have committed serious
crimes in the future.
How could Virginia policy makers have adopted so regres-
sive a policy and one so unjustifiable on the basis of available
knowledge? Part of the answer is political ideology. The com-
mission’s initial chair was former U.S. Attorney General Wil-
liam Barr, an outspoken proponent of harsh punishment poli-
cies. But why would such a person have been made chair of
a state sentencing commission, and why would other mem-
bers of the Virginia Sentencing Commission have allowed
such policies to be adopted? The temper of the times, the
prevailing penal sensibilities, permitted Virginia policy mak-
ers to ignore the demonstrated failures of selective incapacita-
tion as a policy option, to ignore traditional notions about
appropriate handling of young offenders, and to adopt poli-
cies that treat children more severely than adults.
Disintegrative Shaming
Professor Kahan’s “shaming” (1998a, pp. 1639–43) or “pun-

ishment incommensurability” (1998b, pp. 691–708) argu-
ment is the strangest of the three examples. Twenty years
ago, Kahan’s arguments on disintegrative shaming, like those

mentioned earlier on ex post facto laws, would have been
widely considered bizarre. In our time, they won him ap-
pointments at two of America’s pre-eminent law schools, the
University of Chicago and Yale. Though he has written about
shaming repeatedly and at great length, the basic argument
can be briefly summarized. Punishment is not, as most people
who write about punishment theory suppose, primarily
about attributions of culpability and imposition of deserved
punishments, or primarily about crime prevention, but about
shaming. Most writers about punishment do not understand
this, he says, and if they hope ever to influence policy must
accept that punishments should ”unambiguously express dis-
gust“ of the offender (Kahan 1998a, p. 1656). Explaining, for
example, that community penalties will not win public favor
until they are made more debasing, he compares them unfa-
vorably with imprisonment:
Prison, in contrast, does unequivocally evince disgust.
. . . By stripping individuals of liberty—a venerated sym-
bol of individual worth in our culture—and by inflicting
countless other indignities—from exposure to the view
of others when urinating and defecating to rape at the
hand of other inmates—prison unambiguously marks the
lowness of those we consign to it. (Kahan 1998, p. 1642)
Elsewhere, he writes that offenders should be subjected to
“intrinsically repulsive,” “degrading,” or at least “effectively
stigmatizing” punishments and that, for example, community
service should be renamed “shameful service” (1998a, p. 706).
So far as I can tell, these ideas derive from two intellectual
developments of recent decades, both of which Kahan misun-
derstands or misinterprets. The first is increased attention to
the norm-reinforcing, moral-educative, and expressive ef-
fects of punishment that in earlier chapters I associated with
Émile Durkheim. Until the last ten years, most scholarly writ-
ing on punishment was by lawyers and philosophers, and
they tended to concentrate on procedures and policies as they
affected the convicted offender standing before the judge, or
on the possible crime-reductive effects of particular sentences

or policies. Since then, precipitated by sociologist David Gar-
land’s (1990) writing on the sociology of punishment and
psychologist Tom Tyler’s work on procedural justice (1990,
2003), much more attention has been given to the broader
normative and social-psychological effects of punishment.
The second development Kahan apparently misconceives
is the growing and increasingly subtle literature on public
understanding and opinion about punishment (Roberts et al.
2002). This shows broadly that the general public believes
sentences are too lenient, but it also shows that public opinion
is based on misconceptions of crime and punishment attribut-
able to media concentration on exceptional cases, that judges’
sentences are harsher than the public realizes, that the sen-
tences citizens say they would prefer are less severe than are
actually imposed, and that there is widespread support for
rehabilitative programs and community penalties.
Kahan is an appropriate figure on whom to focus because,
more than any other contemporary American writer, he offers
views about criminal justice policy that seem best understood
in terms of prevailing penal sensibilities. It is hard to imagine
how else they could be explained. Besides his proposals con-
cerning abrogation of the ex post facto doctrine and degrada-
tion as a penal strategy, he has offered two other equally
radical proposals. First, he proposed elimination of the cen-
turies-old doctrine of strict construction of penal statutes
(Kahan 1997). That doctrine, because of the importance of
individual liberty, requires criminal statutes to be narrowly
construed, with all interpretive ambiguities construed in fa-
vor of the individual. The maxim that it is better that ten
guilty people go free than that one innocent be convicted is
another emanation of the same underlying idea. Second, Ka-
han has vigorously and outspokenly supported efforts to give
police authority to forbid poor young inner-city blacks from
gathering peacefully on the streets (Kahan and Meares 1998;
Meares and Kahan 1997).
I discuss the intellectual backdrop to Kahan’s disintegrative
shaming analysis in some detail to make a general point and
a specific point. The specific point is that, looked at with any

care, existing systematic knowledge and research findings
provide no credible evidentiary base for Kahan’s proposals.
The general point, however, is that proposals such as Kahan’s
make perfect sense, of a sort, in the context of the penal
sensibilities of the last ten years.
Expressive Punishments
Three major literatures—in philosophy, criminal law theory,
and sociology—have revived interest in expressive punish-
ments. Philosophers have developed communicative theories
in which a central aim of punishment is to express to the
offender the wrongfulness of his or her behavior. Criminal
law theorists have emphasized the role of criminal punish-
ments in reinforcing or undermining social norms. Sociolo-
gists have investigated ways in which official responses to
crime shape the ways offenders perceive the integrity of the
system and have argued that more respectful, nurturing, and
holistic approaches may increase offenders’ prospects for
achieving law-abiding lives.
Communicative Theories of Punishment The first is the devel-

opment in the philosophical literature, exemplified by the
writings of Antony Duff (2000), Joel Feinberg (1970), Jean
Hampton (1984), and Jeffrie Murphy (1985), of “communica-
tive” theories of punishment. All of these are moral theories,
premised on respect for the moral autonomy of the offender,
in Dworkin’s (1977, p. 227) terms showing ”equal respect
and concern“ for each offender, and call in various ways for
punishment to express to the offender the wrongfulness of
his acts.
The prevailing image is of a judge, looking deeply into an
offender’s eyes, more in sorrow than in anger, and wanting
to help her understand why what she has done is wrong. If
it works, the offender will come to understand that she was
temporarily distracted from right values by egoism, emotion,
or impulse, and be ready to reclaim her place among right-
thinking people. This will not happen because she was bul-
lied, threatened, or brainwashed, but because as a morally
autonomous person she now understands that, and why, her
actions were wrong.
There are differences in view among philosophers working
within the communicative punishment tradition as to whether
the aim is solely to express norms to the offender, so that
she as a morally responsible actor can come to understand
the wrongfulness of her acts, or whether in addition there
is a collateral aim through punishment to express norms to
the general community as bystander. There are differences
in view as to whether her regret, repentance, contrition, or
remorse is a sufficient outcome or whether, for her sake or
the larger society’s, some material punishment should also
be imposed or borne. There is also a question of how offenders
who are insensitive, defiant, or impervious to moral reasoning
should be handled. However, nowhere in this literature are
there indications that the aim of punishment should be to
express disgust of the offender or to debase her in order to
placate public opinion. That, in the conventional Kantian
language, would be to use the offender merely as a means,
and that is something no mainstream moral theory would
allow (N. Walker 1991).
Moral-Educative Theories of Punishment A second relevant

punishment literature derives from the Durkheimian func-
tionalist notion that the criminal law serves to identify and
reinforce basic social ideas about right and wrong, that
”crime brings together upright consciences and concentrates
them.“ The Norwegian writer Johannes Andenaes (1974),
who wrote about the ”moral-educative functions” of the crimi-
nal law, was influential in reviving interest in such views.
The criminal law is seen as performing a dramaturgical func-
tion, with punishment directed primarily at the community
and not at the offender. Crime is a part of every human
society, possibly a necessary part, a functional mechanism
that helps set and then illuminate and reinforce the bound-
aries of acceptable behavior. The primary aim of punishment
is to restate and reinforce prevailing norms: “[Punishment]
does not serve, or else serves only secondarily, in correcting
the culpable, or in intimidating possible followers. From this
point of view, its efficacy is justly doubtful and, in any case,
mediocre. Its true function is to maintain social cohesion in-
tact” (Durkheim 1933 [orig. 1893], p. 108).
There has been a revival of interest in Durkheimian ideas
about punishment in Germany and Scandinavia, as an alter-
native to retributive and utilitarian ideas. Influential elabora-
tions of such views have been offered by Finnish writers
Patrik Törnudd (1997) and Tapio Lappi-Seppälä (2001) under
the name “general prevention,” and in Germany by Bernd
Schünemann under the name “positive general prevention“
(Schünemann 1998). Both traditions distinguish their neo-
Durkheimian ideas from Anglo-American “(negative) general
prevention,” which operates through the utilitarian pro-
cesses of deterrence, incapacitation, and rehabilitation (Hart
1968).
The German and Scandinavian ideas emphasize that pun-
ishment has an important role to play in setting and reinforc-
ing norms, but it is a secondary role. For the most part,
extraordinary circumstances aside, people do or do not com-
mit crimes because of the socialization they do or do not
receive from primary institutions such as the home, the fam-
ily, the church, the school, and the community. The primary
work of crime prevention must be done in those places. It
is important that the criminal law confirm basic behavioral
norms, and be seen to do so, but the primary work must be
done elsewhere. This has important, perhaps to most Ameri-
cans surprising, implications.
First, because the law should perform its back-up role, it
is important that criminal acts have penal consequences. As
a result, although the Finns, Swedes, and other Scandinavians
have among the lowest imprisonment rates in the world,
expressed in terms of people in prison on an average day
per 100,000 residents, they have among the highest prison
admission rates in the developed world, expressed in terms
of the number of people per 100,000 admitted to prison in
a year (Young and Brown 1993; Kommer 1994).
Second, because the law should perform its back-up role,
it is important that punishments be commensurate with the
gravity of the offenses for which they are imposed. Put differ-
ently, proportionality is a first principle; unless more serious
crimes receive harsher penalties and less serious crimes lesser
ones, the law’s normative messages will be morally incoher-
ent and contradict the primary norm-setting processes (Lappi-
Seppälä 2001).
Third, there is no reason to expect changes in the severity
of penalties to be effective or desirable. If primary institutions
play the major socialization roles, and punishment can only
marginally reinforce or undermine prevailing norms, it would
be unrealistic to imagine that changes in the severity of sanc-
tions can have much effect. Patrik Törnudd (1997, p. 190)
thus observed, “A strong belief in general prevention as the
guiding rationale of the criminal justice system thus does not
imply that changes in policy, such as increases in the severity
of punishment, would widely be seen as an appropriate or
cost-effective means of controlling crime.”
Once again, there is nothing here, express or implied, about
debasement, disgust, or stigmatization. Indeed, since North-
ern European countries take seriously the European Torture
Convention’s prohibition of inhumane and degrading pun-
ishments (Evans and Morgan 1998; Morgan 2001), it would
be surprising if there were.
Reintegrative Shaming “Reintegrative shaming” is a recon-

ceptualization of punishment proposed by Australian crimi-
nologist John Braithwaite (Braithwaite 1989, 2001). The pro-
posal is simultaneously a hypothesis that restorative responses
to criminal offenders might have greater crime-reducing ef-
fects than criminal justice responses and a normative argu-
ment that reintegrative processes are more humane and re-
spectful of human dignity.
Braithwaite’s notion is that reactions to crime should simul-
taneously express disapprobation and support, in much the
same way as parents communicate to children that they have
misbehaved but that they are still loved. The ”shaming“

communicates through disapproval the importance of the
norms or expectations that were violated but in a way that
also communicates respect for the individual and concern
for her well-being, and is therefore ”reintegrative.” This is
contrasted with the destructive shaming of the traditional
criminal justice system that ostracizes, alienates, and often
breeds defiance or leads to rejection of pro-social norms and
attachment to antisocial ones. Braithwaite (1999, 2001) has
proposed a number of reasons why traditional criminal justice
approaches are likely to be less effective at socializing offend-
ers and preventing crime than are reintegrative approaches.
Even this brief summary should make it clear that Braith-
waite’s affirmative theories of shaming have nothing in com-
mon with Kahan’s proposals which, to the contrary, embody,
even celebrate, the kind of destructive processes that Braith-
waite decries.
Public Knowledge and Opinion

The second growing literature that might support Kahan’s
ideas, but does not, concerns public knowledge and opinion
about punishment. In particular, Kahan cites a study by Cana-
dian psychologists Anthony Doob and Voula Marinos (1995)
of Canadians’ support for the use of fines as punishments for
various crimes. A majority of respondents indicated that, for
many crimes, a fine would not be a normatively appropriate
substitute for imprisonment. Tom Tyler and Robert Boeck-
mann’s (1997) study of reasons (primarily “expressive”) why
ordinary people support three-strikes laws even though there
are sound reasons to doubt their instrumental effectiveness
offers comparable findings: imposing a lengthy prison sen-
tence for a third violent crime sends a message people believe
should be sent. Hans Boutellier (2000) and Joseph Kennedy
more recently (2000) argue that punishment and crime control
policy should be understood in our time as having primarily
expressive functions. Kahan (1998b), drawing on a hodgepodge
of newspaper clippings and letters to the editor as authority,
makes similar arguments about community service.
There are serious problems with Kahan’s claim that public
support for punishment practices requires disintegrative
shaming. It is poorly informed. A large literature shows that
public attitudes are much more complex and less single-
mindedly vengeful than he suggests (Roberts et al. 2002). It
is overbroad. Even if nonincarcerative punishments for some
very serious crimes would, in Model Penal Code language,
“unduly depreciate the seriousness of the offense,” most
crimes are not that serious. It is parochial. In many Western
countries, fines and community service are commonly used
as sanctions for quite serious, including violent and sexual,
crimes (Albrecht 2001; Lappi-Seppälä 2001; Tak 2001).
Under-Informed The large multinational public opinion liter-

ature offers much more complicated findings than Kahan sug-
gests (Roberts 1992; Hough and Roberts 1997; Roberts et
al. 2002). Surveys consistently show that the general public
believes that the average crime is more serious than it is,
substantially underestimates the severity of punishments im-
posed, and generally supports punishments less severe than
are now imposed. The public has ambivalent views about
punishment, wanting offenders to be punished for their
crimes but, believing that social disadvantage and drug de-
pendence are primary causes of offending, also wanting (and
being willing to pay for) offenders to be rehabilitated. Finally,
for all but the most serious crimes and the most incorrigible
criminals, the public is willing to have criminals sentenced
to community penalties in place of prison, so long as the
community penalty is burdensome or restitutive (community
service or work release coupled with restitution are okay,
house arrest without more is not). Thus, while there is some-
thing to be said for Kahan’s claim that the public wants
symbolically appropriate punishments, that by itself is a par-
tial and misleading summary of the evidence. None of the
major scholars of public opinions and attitudes about crime
interprets the preference for symbolically appropriate sen-
tences as a demand for shaming punishments.

There is substantial evidence of public support for wider
use of community services in place of incarceration for many
kinds of offenders because it satisfies symbolic demands for
punishments that are burdensome and express right values,
and is in an important sense restitutive (Doble 1997; Doble
and Immerwahr 1997; Farkas 1997; Begasse 1997). To the
contrary, community service is a sanction that Kahan argues
must be made “degrading,” “stigmatizing,” “shameful,” and
“intrinsically repulsive” if it is to win public support (1998b,
p. 706).
Overbreadth Even if Kahan were right that the public insists

on only prison sentences for violent crimes, most people now
admitted to prisons have not been convicted of violence. In
1996, for example, of people committed to state prisons, 29.3
percent had been convicted of violent crimes, 29 percent of
property crimes, 30.2 percent of drug crimes, and 11.3 per-
cent of something else. In federal prisons the percentages,
for the same categories in 1997 were 11.9 percent, 5.6 percent,
60.1 percent, and 22.4 percent. Among convicted jail inmates
in 1996, the corresponding percentages were 21.8 percent,
28.6 percent, 23.7 percent, and 25.9 percent. The public opin-
ion evidence summarized in the preceding paragraph shows
that for the nearly 80 percent of people sentenced to prison
for nonviolent crimes, there is broad support for increased
use of nonincarcerative sanctions that do not debase.
Parochialism Experience in other countries suggests that the

public would be willing to accept fines and community ser-
vice as prison substitutes for all but the most serious crimes.
Throughout Scandinavia and Germany, for example, day
fines are the modal sanction. Day fines are penalties that
take into account some measure of the offender’s earnings
or wealth, usually a day’s net pay, and the seriousness of
his crime. A moderately severe crime might generate a fine
of thirty days’ net pay. Day fines are used in these countries
as punishments for most crimes, including many violent and

sexual crimes and especially property crimes (Weigend 1997;
Jareborg 2001; Albrecht 2001). In England, Scotland, and
Holland, community service was established to serve as an
alternative to prison sentences for moderately severe crimes
and is used in that way (Pease 1985; McIvor 1995; Tak 2001).
Kahan’s arguments for disintegrative shaming are not
deducible from any of the important recent developments
concerning expressive punishments or research on public
opinion about crime and punishment. What is left is an idio-
syncratic argument that offenders should be subjected to
debasing, degrading punishments because, Kahan apparently
believes, offenders have “failed to internalize society’s moral
norms” and populist appetites, however well- or ill-informed,
would thereby be appeased or gratified.
All the consequences of penal sensibilities such as character-

ized the 1980s and 1990s need not be bad. Heightened will-
ingness to adopt moralistic postures, harnessed properly, can
be used to achieve positive social goals. The feminist move-
ment has greatly sensitized Americans to issues of sexual
exploitation and urged greater use of the law to address gen-
der-related abuses of power and violence. Before the 1970s,
people who were social policy liberals, and hence likely to
be concerned with gender equity issues, were also likely to be
due process liberals, and hence uncomfortable with criminal
justice solutions to social policy problems that treated offend-
ers harshly. By the 1980s, criminalization of domestic vio-
lence became a central plank in the feminist platform (Zimring
1989) and led to calls for mandatory arrest policies for alleged
domestic assault misdemeanors. Reasonable people can differ,
of course, as to when the criminal law should be invoked
and for what kinds of cases. Since contemporary penal sensi-
bilities make overreliance on the criminal law a recurring
risk, it is possible that criminal law is being overused in our
time as a response to domestic violence and that other, more
nuanced, approaches in many cases would do more good and
less harm.

Likewise, the modern victims’ movement may owe much
of its success to prevailing penal sensibilities. No informed
person doubts that the victims’ movement has enhanced at-
tention to victims’ interests and mobilized new resources to
address victims’ needs. There is no good reason why concern
for fair handling of offenders implies lack of concern for
victims, but it is clear that the contemporary victims’ move-
ment dates only from the 1970s. Before that, little program-
matic or policy attention was given to victims issues. The
overreactions associated with contemporary penal sensibili-
ties predisposed advocates to the nonsensical view that they
must be both “for” victims and “against” criminals. That
dualism is not inexorable and may be moderating. As with
the domestic violence movement, contemporary sensibilities
appear to have mobilized and energized the victims’ move-
ment.
On balance, though, the most dramatic products of contem-
porary sensibilities have been negative. None of the three
policies and proposals principally discussed in this chapter
could have been seriously put forward in the 1950s or the
1960s. The penal sensibilities of those decades were, as the
next chapter demonstrates, radically different. These three
policy ideas, however, stand as a stark warning about what
can happen when, during a cyclical period of heightened
intolerance of offenders and drug users, policy makers lose
their senses of humility and proportion and lose sight of
timeless values.

7
Unthought Thoughts
Just as there are bad ideas and regrettable policies in contem-

porary America that are best understood as underconsidered
outgrowths of contemporary penal sensibilities, there are
ideas and policies, formerly so mainstream as not in their
own times seriously to be questioned, that have all but disap-
peared. There are also ideas that are so much a part of contem-
porary sensibilities as seldom to be questioned, yet they seem
nearly absent in other times.
Lots of examples could be offered of issues and ways of
thinking that are common in particular times and conspicu-
ously absent in others. Two issues, preventive detention and
prosecutorial sentence appeals, which due process liberals
and crime control conservatives fought over in the 1970s,
provide examples. The preventive detention question was
whether people charged with crimes could be held in jail
until trial because they were believed to be dangerous and
likely to commit an offense while free on bail. Until the 1970s,
the conventional view was that, with one legitimate and one
illegitimate exception, the only reason to deny bail was be-
cause the defendant was believed likely otherwise to run
away. The legitimate exception was for capital cases, cases
in which capital punishment was a possible punishment. The
illegitimate exception concerned cases in which the judge
wanted the defendant held in jail and, lacking a valid reason
for doing that, purposely set the bail higher than the defen-
dant could pay. The sentence appeal question was whether
prosecutors should or constitutionally could be allowed to
appeal sentences they did not like.
169
Most academic observers at the beginning of the 1970s
would have expected due process liberals to win both de-
bates. Both were issues in which individuals’ interests in
personal liberty conflicted with collective interests in public
safety. Consistent with the penal sensibilities of that time,
most academic observers believed that the nebulous and dif-
fuse collective public safety interest should take second place
to the concrete and particular liberty interests of individuals.
The due process liberals lost both arguments. More to the
point, though, few people in the early years of this new
century would consider either issue particularly contentious.
Penal sensibilities have changed and to most observers it is
self-evident that dangerous offenders should be held before
trial (and in some cases, if we believe they are dangerous, after
their prison sentences have been completed), lest prospective
victims unnecessarily suffer. Likewise to most people it seems
self-evident that prosecutors should be able to appeal sen-
tences they believe are not severe enough, lest offenders get
off easy.
The rest of this chapter devotes more sustained attention
to three other subjects—the Model Penal Code’s treatment
of purposes of punishment, how recidivism rates are mea-
sured, and how the costs and benefits of punishment should
be measured. The first illustrates concerns whose centrality
today is taken for granted but just a few decades earlier
received little attention. Punitive ideas expressed in terms
of just deserts, commensurability, retribution, and propor-
tionality are predominant in modern discussions of punish-
ment. Assertions that public opinion should or does guide
setting of penal policies are common. Yet, during the decade-
long development of the Model Penal Code, the most ambi-
tious and successful criminal law reform ever undertaken in
the United States, only passing mention was made of just
deserts, proportionality, retribution, or placating the public
mood as important aims of punishment or sentencing.
The second illustrates an empirical and policy issue that
forty years ago seemed central and today is nearly forgotten:
the curious disappearance from criminological research of

interest in the recidivism of people released after serving a
first prison term. Of people released from prison the first time,
typically only about a third reoffend. Of all people released
from prison, typically two-thirds reoffend. The former find-
ing suggests either that prison sentences rehabilitate offend-
ers, or that many people are sent there who need not be.
Both were common beliefs in the 1960s. Conversely, the latter
finding is consistent with widespread attitudes in more recent
decades, when few believed prison sentences rehabilitate and
many believed most offenders are incorrigible. The curiosity
is not that each period seemingly unthinkingly employed
recidivism measures that matched prevailing sensibilities,
but that the earlier measure seems to have been forgotten
entirely.
The third illustrates the collective amnesia or blindness
that changing penal sensibilities can cause and, perhaps more
optimistically, that there may be signs that the amnesia is
wearing off as sensibilities become less harsh. Jeremy Ben-
tham, the first great analyst of the costs and benefits of pun-
ishment, was clear that all costs and benefits counted—crimes,
losses, and victim suffering prevented; costs to the state of
administering punishments; and the suffering of punished
offenders and their loved ones. Bentham’s analyses and calcu-
lations were theoretical. Only in recent decades have special-
ized skills been available that permit sophisticated quantita-
tive calculations of the costs and benefits of punishments.
And in these decades, nearly everybody left the offender,
and his family, and his community out of the equations.
Costs of running the courts, jails, probation, and prison were
calculated, and costs of law enforcement, and victims’ prop-
erty losses, lost earnings and productivity, medical costs, and
pain and suffering. Few analysts thought about offenders’
lost earnings and productivity, reduced life chances and life
expectancies, or pain and suffering, much less the similar
burdens borne by their partners, children, and other loved
ones. It should not be too surprising, in an era when people
were expected to be “for” victims or offenders, that politi-
cians overlooked the effects of penal policies on offenders
UNTHOUGHT THOUGHTS 171

and their families. Were it not for the force of prevailing
sensibilities, however, it would seem astonishing that aca-
demics were equally blind. In the last few years, analysts
have remembered offenders (e.g., Cook and Ludwig 2001).
All three illustrations are like Holmes’s non-barking dog.
The absence of issues from an era offers important clues to
the governing penal sensibility. And, if those times were
blind to some things and acutely aware of others, in each
case opposite to our own times, we are likely to be just as
vulnerable to collective selective awareness.
Preventive Detention
and Prosecutorial Appeals
In 2002 battles over preventive detention and prosecutorial

sentence appeals are distant memories. Few people seem in
principle opposed to either or to recognize the issues as in
any way contentious. Due process liberals, judges, and just
about everyone else has been socialized by the ethos of our
time into accepting the public safety cases for both as self-
evident. It was not so before and likely some day will not
be again.
In the 1970s, there were heated and extended political
fights over preventive detention. Due process champions for
years successfully opposed federal adoption of legislation
that would permit persons accused of crimes to be confined
before trial because of dangers they might present to public
safety. Defendants are presumed innocent, the argument
went, and pretrial confinement is allowable only to assure
defendants’ presence at trial. Detention for any other reason
except in the case of capital crimes is an unconstitutional
invasion of citizens’ liberty interests.
Crime-control conservatives by contrast argued that citi-
zens’ interest in public safety justified restrictions on the
liberties of dangerous alleged offenders. Moreover, they ar-
gued, judges in practice sometimes set bail so high that people
they believed to be dangerous offenders could not raise the
money to pay it and accordingly had to remain in jail until
trial. This is hypocritical, they argued, and probably resulted
in judges detaining dangerous offenders less often than they
should. The battle raged for years but finally the conserva-
tives won and in 1973 Congress enacted the District of Colum-
bia preventive detention law (D.C. sect. 23–1322-23–1331).
Preventive detention today is commonplace, authorized by
federal law and the laws of most states, and repeatedly sanc-
tioned by state and federal appellate courts. The Supreme
Court long ago, in Bell v. Wolfish, 441 U.S. 520 (1979), held
that the presumption of innocence is merely an evidentiary
rule and does not by itself imply anything about pretrial
commitment. In its most extreme form, sexual psychopath
laws provide that states may hold convicted sex offenders
after the expiration of their prison sentences if a court de-
cides, by a civil law more-probable-than-not standard, that
they are dangerous. The constitutionality of such laws was
upheld by the U.S. Supreme Court in Kansas v. Hendricks, 521
U.S. 346 (1997). The federal real offense sentencing provision,
discussed in chapter 6, also allows people to be sentenced
to longer prison sentences on the basis of more-probable-
than-not findings. The rationale is not preventive detention,
but the practice resembles preventive detention in the sense
that people’s liberty is taken on the basis of scant evidence
to achieve larger public purposes.
Here is another example of a long-forgotten but sharply
contested issue of principle. Before the first sentencing guide-
lines were developed in the 1970s and 1980s, appellate review
of sentences in the United States was close to nonexistent.
This was because indeterminate sentencing, ubiquitous in
the United States before 1975, accorded judges and other
officials great discretion to individualize decisions. There
were, in effect, no sentencing rules or standards whose cor-
rect application appellate courts could assess. The creation
of sentencing guidelines, however, implied the right of defen-
dants to appeal sentences (Frankel 1972). Guidelines created
standards or presumptions that, if improperly or incorrectly
interpreted or applied by the judge, could result in wrongful

deprivation of an offender’s liberty. Surely a defendant could
seek appellate review of a trial judge’s sentence when he or
she reasonably believed the sentence was longer or otherwise
harsher than applicable guidelines indicated it should be?
But if defendants were to be allowed to appeal sentences
they believed too harsh, should not prosecutors be allowed
to appeal sentences they believed too soft?
There were two principal arguments for the prosecution:
procedural symmetry, what is fair for the defendant is fair
for the prosecution; and substantive symmetry, if the defen-
dant has a legally enforceable interest in avoiding too much
punishment, the state has a reciprocal interest in avoiding
too little.
There was one principal argument against: the state should
not be able to get on appeal what it could not get in court.
Just as, under double jeopardy doctrine, an acquittal absolves
a factually guilty defendant of the cost, stigma, and emotional
turmoil of a second prosecution, so an anomalously slight
sentence should insulate a defendant from the risk of a harsh-
er one. The state gets but one bite at the defendant’s apple,
and the presumptions all run in favor of the defendant’s
autonomy and liberty interests. Such a position is also consis-
tent with the principle of strict construction of penal statutes
mentioned in chapter six. Just as ties in baseball go to the
runner, ambiguities in criminal court benefit the defendant.
There was extended and spirited disagreement. Most crimi-
nal procedure scholars predicted that prosecutorial appeal
would not withstand constitutional scrutiny. In United States
v. DiFrancesco, 449 U.S. 117 (1980), regarding a “dangerous
special offender,” it did, and the issue has disappeared. Prose-
cution appeals of sentence are now commonplace in every
American jurisdiction that allows defendants to appeal sen-
tences.
One might suggest that the two preceding examples merely
illustrate the legal doctrine of stare decisis in action. Litigants
have accepted that the courts have ruled, and life has moved
on. That could be, but not necessarily. When courts decide

essentially arbitrary questions in civil law disputes, it should
be no surprise that the choice is typically accepted. In con-
tract law, for example, common law courts long ago were
faced with the problem of acceptances of offers of sale that
are lost in the mail. When the acceptance genuinely is lost,
either party might be damaged. If the price of the goods goes
up, the would-be buyer, believing he has already bought at
a good price, will stand pat. If his lost acceptance prevents
formation of a contract, he will have to pay more to buy the
same goods elsewhere (or from the original seller). Con-
versely, the seller who did not receive the acceptance may
sell the goods to another buyer and, if he is held to the
original contract, may have to sell at a loss if the price of
goods goes up, or pay damages. In the long run, it did not
matter which way the courts resolved the issue of the accep-
tance lost in the mail. If the rule is clear, parties can plan
around it by buying insurance or negotiating over who bears
what risk of loss.
Disputes over civil liberties or other normative issues often
are another matter. There are many normatively contested
subjects about which people do not lambishly accept authori-
tative court decisions. Abortion, gun control, and capital
punishment are examples. The difference may be that these
are such deeply contested normative issues that proponents
of polar positions are less likely to be worn down by the
general penal sensibilities of an era, or those issues may be
so normatively contested that there are no general penal sen-
sibilities.
Issues imbued with less primal emotions may be more sus-
ceptible to the social and cognitive forces Musto (1973), Gus-
field (1963), Gilmore (1974), Johnson (1976), and many other
historians describe. Preventive detention and prosecutorial
appeals, because they squarely involve deprivations of indi-
viduals’ liberty for collective purposes, might have been ex-
pected to remain galvanizing issues akin to gun control or
abortion. The sensibilities of our time, however, are such
that they did not.

Desert and the Model Penal Code
It can be eye-glazingly tedious to encounter an nth recital

of the variety of mainstream but often overlapping sets of
punishment theories—retributive, utilitarian, consequential,
hybrid, communicative, expressive—and the various ways
they can be subpartitioned. Precursors, foreshadowers, and
full statements of core ideas can be found in third-century
Athens, and they have echoed and accreted throughout the
centuries. This is a subject about which little that is new is
likely to be found under the sun. Every post-Enlightenment
period has its matched pair of preventive and moralizing
debaters—Jeremy Bentham and Immanuel Kant, John Stuart
Mill and James Fitzjames Stephen, Herbert Hart and Lord
Patrick Devlin, Norval Morris and Andrew von Hirsch.
For the past quarter century, retributive “just deserts”
ideas have been in the ascendant among philosophers, politi-
cians, and practitioners. Philosophers and just deserts theo-
rists generally put forward subtle and principled accounts
that tie punishment to offenders’ culpability and take their
liberty interests seriously. In practical politics, the ascen-
dancy of desert ideas is cruder and reduces to little more
than an idea that offenders deserve to suffer. Often this is
expressed in slogans—“Do the crime, do the time;” “Three
strikes and you’re out;” “Hard time for hard crime.” And yet,
the Model Penal Code, completed forty years ago, a decade in
the making, rarely mentions desert, retribution, or propor-
tionality in its text or original commentary, and echoes of
such ideas only occasionally appear in the transcripts of the
American Law Institute (“ALI”) meetings at which successive
drafts were discussed.
This wafting of ideas, even of particular ideas, in and out
of eras, is nothing new. Nigel Walker (1999, p. xi) similarly
observes of an earlier era in which utilitarianism was ascen-
dant:
In 1877 Edward Cox, Recorder of Falmouth, published

the first English textbook on sentencing, The Principles of

Punishment. It preached a Benthamite version of utilitari-
anism, in which the only important consideration was the
need for deterrence in some kinds of case, and the absence
of this need in other kinds. Reformation was treated with
scepticism. . . . Most remarkable, was the absence of any
attention to retribution, let alone “just deserts.”
The Model Penal Code’s draftsmen began work early in

the 1950s, and the final version was formally approved and
published in 1962. The following paragraphs describe and
quote from American Law Institute discussions of such things
as purposes of punishment, presumptions against imprison-
ment and for parole release, and related subjects. The Code
went through a series of tentative drafts. I mostly refer to
these rather than to the final version, since my interest here
is in the proposals and how they were discussed rather than
in the Code per se.
The main themes are that offenders’ prospects for law-
abidingness are nearly always relevant to decision making,
that officials should have broad discretion in making deci-
sions, and that prison should be used as little as possible. A
fourth theme could be offered—that public safety considera-
tions be taken into account, in effect a co-equal goal with
offender rehabilitation. An offender who is believed not to
have been rehabilitated is an offender who may need incapac-
itation. In transcripts of debates, however, allusions to public
safety often feel as if they are offered by rote and the Code’s
provisions clearly give primacy to the first three themes.
The ALI is an establishmentarian organization of lawyers,
judges, and law professors. It has a self-perpetuating nomi-
nated membership and for much of its life has been prepon-
derantly composed of members of major commercial law
firms. At least in legal circles, one might suppose that “presti-
gious” were part of the organizational name, as in “X is a
member of the prestigious American Law Institute.” It was
established in the 1920s with a law reform purpose and spe-
cial emphasis on systematizing and rationalizing state laws
(Goodrich and Wolkin 1961; Hull 1990; American Law Insti-

tute 1998; Elson 1998). Each American state and the federal
government has its own legal system, and this led to the
practical problem for businesses operating in national mar-
kets that legal provisions concerning basic features of con-
tract, tort, insurance, and personal property law varied sig-
nificantly from state to state. Initially ALI concentrated on
descriptive summaries or “restatements” of existing law, as
in the Restatement of the Law of Contracts, or on more pre-
scriptive synthetic proposals for improvement, as in the Uni-
form Commercial Code. Particular ALI projects are carried
out under the aegis of committees, and under the direction
of “reporters,” but proposals receive official ALI imprimatur
only upon positive votes of the membership at annual meet-
ings.
The Model Penal Code exemplified later and more prescrip-
tive attempts to develop model codes which, if successful,
might serve as the bases for comprehensive refashioning of
state laws. Columbia University Law School professor Her-
bert Wechsler, in retrospect the most influential American
criminal law scholar of his century, was the Chief Reporter
for the Model Penal Code. The key drafters of the Code’s
sentencing and corrections provisions, and the oversight
committee, included leading law professors, judges, prosecu-
tors, psychiatrists, mental health specialists, and corrections
professionals. Sanford Bates and James V. Bennett, for exam-
ple, successive reformist heads of the U.S. Bureau of Prisons,
and Paul W. Tappan, sometime head of the U.S. Parole Board,
and comparable state officials, were active participants. Over
a ten-year period, from 1953 to 1962, drafts of various pro-
posed sections were discussed at ALI annual meetings and
final approvals were given in 1961 (sentencing and correc-
tions) and 1962 (the rest).
I describe ALI and its processes to highlight that it is about
as establishmentarian as an organization can be. Its center
of gravity in the elite commercial bar implies institutional
conservatism. Its deliberative processes are ponderous and
unwieldy and, at day’s end, require that the text of a pro-
posed model criminal law pass muster with an ALI member-

ship composed largely of judges and business lawyers. In
other words, it is not an organization predisposed to cutting-
edge ideas or radical innovation in the criminal law.
Curiously, though, the provisions of its Model Penal Code
are so radical by contemporary standards as nearly to be
unimaginable. University of Chicago law professor Albert
Alschuler (1978, p. 552) expressed this in 1978, commenting
in close retrospect on the indeterminate sentencing systems
the Model Penal Code proposed, “That I and many other
academics adhered in large part to this reformative viewpoint
only a decade or so ago seems almost incredible to most of
us today.”
With hindsight, it is striking how much attention the draft-
ers gave to the perceived need to accommodate offenders’
treatment needs and prospects and how little to notions of
“deserved punishment,” “just deserts,” or public opinion.
The following paragraphs set out the principal sentencing,
parole release, and good time provisions (ALI 1954, 1956,
1960; ALI Proceedings 1954, 1956, 1960, 1961) and the pri-
mary justifications offered in their support. I do this with
apologies to readers who are not sentencing specialists. For
those who are, the reminder of the distance between the
sensibilities of the Code’s drafters, and those of our own era
that the Code’s provisions illustrate, may be enlightening.
In reading the next few paragraphs, bear in mind that modern
sensibilities are hostile to judicial and parole board discretion,
often opposed to parole release itself, enamored of the idea
that punishments should relate primarily to the crime and not
the criminal, supportive of the idea that the public opinion
should be taken into account, and suspicious of rehabilitation
as a purpose of punishment. The generalizations in the preced-
ing sentence apply about equally to conservatives and liberals.
Purposes of Sentencing
The first official draft, from 1962, lists eight “general pur-
poses of the provisions governing the sentencing and treat-
ment of the offender.” The first three are “To prevent the

commission of offenses; To promote the correction and reha-
bilitation of offenders; To safeguard offenders against exces-
sive, disproportionate or arbitrary punishment” (ALI 1962,
pp. 2–3). “Safeguarding against disproportionate punish-
ment,” please observe, is not at all the same thing as “assuring
proportionate punishment.” To argue that an offender should
not be punished more harshly than he deserves implies nei-
ther that he should be punished as much as he deserves nor
that he may not be punished less than he deserves. Norval
Morris’s (1974) “limiting retributivism” and Finnish “asym-
metrical proportionality” (Törnudd 1997; Lappi-Seppälä
2001), for example, both allow for fixed upper limits to pun-
ishment, based on proportionality ideas, but impose few or
no lower limits. Nowhere in the Model Penal Code is mention
made of “imposing deserved punishment,” “acknowledging
the seriousness of the crime,” “expressing public outrage,”
or anything similar.
Authorized Prison Sentences

The first proposed draft divided all felonies into three classes.
Judges set minimum sentences and maximums were pre-
scribed by statute. For first-degree felonies, the minimum
was one to twenty years and the maximum was life. For
second-degree felonies, the minimum was one to three years
and the maximum ten. For third-degree felonies, the mini-
mum was one to two years and the maximum five (tentative
draft 2, ALI 1954). In other words, a judge could impose a
one-year minimum sentence, no matter how serious the of-
fense. This is flatly inconsistent with modern attitudes that
underlay mandatory minimums, truth-in-sentencing laws,
and three-strikes laws. The rationale was that, no matter
how horrible the crime, particular circumstances may make
a lengthy sentence unwise or unjust.
Authorized Probation Sentences

The first proposed draft authorized judges to sentence any
person to probation, including those convicted of murder

and other first-degree felonies potentially punishable by a
life sentence, when the judge “deems that his imprisonment
is unnecessary for protection of the public.” To make certain
no one could misunderstand, the commentary explains that
the draft “is based upon the view that suspension of sentence
or probation may be appropriate dispositions on conviction
of any offense” (tentative draft 2, ALI 1954, p. 34; emphasis
added) unless a mandatory sentence of death or life imprison-
ment is prescribed. No matter how horrible the crime, the
judge may conclude that a prison sentence is unwarranted
under all the circumstances, and the provision allowed for
a sentence of probation in those instances.
Reconsideration of Sentences
The first proposed draft made every prison sentence “tenta-
tive” for the first year and authorized the corrections commis-
sioner to petition for resentencing. The commentary explains
that judges have limited opportunity to study the offender
and that corrections officials may decide that the judge “pro-
ceeded on the basis of misapprehension as to the history,
character or physical or mental condition of the defendant”
(ALI 1954, p. 57). Herbert Wechsler explained during the
ALI deliberations, “This is a really fundamental provision of
the draft. One of the great arguments against judicial sentenc-
ing has always been that the judge must decide too much,
too soon. . . . Even with a good pre-sentence report there
has been a limited opportunity to study the offender” (ALI
Proceedings 1954, p. 143). Even Judge Alger Fee from Ore-
gon, one of the ALI’s few skeptics about the Code’s deference
to the expertise of experts, was happy to authorize judges
to reconsider sentences: “[A]s many times as the correctional
authority desires to bring the case before the judge again it
should be done” (ALI Proceedings 1954).
Mitigation of Sentences
The first proposed draft empowered judges, when they be-
lieved the offense of conviction authorized sentences that were

too harsh under all the circumstances, to sentence an offender
convicted of any felony as if he or she had been convicted of
a lower degree felony or a misdemeanor (tentative draft 2, ALI
1954). Wechsler explained that the provision was “really a
kind of introduction of equity, the basic Aristotelian idea of
equity into law that every generalization—and law must be
general—has unstatable qualifications that must be acknowl-
edged when the case arises” (ALI Proceedings 1954, p. 92).
New York County Prosecutor Sarafite agreed that “one of
the most important functions of their [the prosecutors’] office
is to cooperate in the process of mitigating the rigor of the
law in appropriate cases” (ALI Proceedings 1954, p. 92). Sara-
fite was firmly “in favor of anything that would mitigate an
unduly harsh sentence that the judge finds he has to impose
after a jury conviction” (ibid., p. 92), but not in the case of
guilty pleas. Wechsler responded that “the whole reason for
doing it this way was to transfer a certain amount of authority
in this area from the District Attorney to the Courts” (ibid.,
p. 93). In our time, to the contrary, legislators have purposely
transferred discretion from judges to prosecutors to reduce
chances that judges will mitigate sentences. That is a primary
purpose of mandatory minimum sentence and three-strikes
laws.
Good Time
The first proposed draft directed that prisoners receive six
days’ time off for good behavior for each month satisfactorily
served, and that corrections officials could award another
six days per month for “especially meritorious behavior or
exceptional performance of his duties” (tentative draft 5, ALI
1956, p. 81). The good time credits would apply to (and thus
advance) both the minimum to be served before parole release
eligibility and the maximum to be served before mandatory
release. Modern truth-in-sentencing laws, with their require-
ment that at least 85 percent of the sentence be served, are
premised on ideas that sentence reductions for good behavior
are bad things.

Parole Release
The first proposed draft made prisoners eligible for parole
release on completion of their minimum sentences less any
applicable good time and created a presumption in favor of
release when prisoners first became eligible. Prisoners were
required to be released when they had served their maximum
sentences (net of good time) (tentative draft 5, ALI 1956). In
our time, sixteen states have abolished parole release alto-
gether and, contrary to the presumption that release should
occur at the earliest possible time, many modern sentencing
laws are designed to prevent or minimize early release of any
kind.
Public Sentiments
The Model Penal Code created no mandatory sentences or
probation ineligibility provisions (except concerning life sen-
tences and the death penalty [on which it took no position]).
The final draft, however, almost as an afterthought that had
not arisen during a decade’s discussions, acknowledged that
public opinion might sometimes be relevant. In a provision
creating a presumption against imprisonment and for proba-
tion in every case, the final draft added a new criterion, “a
lesser sentence will depreciate the seriousness of the defen-
dant’s crime,” among the reasons for disregarding the pre-
sumption and ordering a term of imprisonment (ALI 1962,
p. 106). A similar provision was added to a list of considera-
tions that might justify disregarding the presumption that
prisoners be released at first parole eligibility.
Wechsler explained in 1961, however, that the new lan-
guage was not meant to direct a judge to respond to public
sentiment about a particular crime, but instead to consider
deterrent processes: “[W]hat really is of concern to the court
in relation to deterrence is not the stiffness of the disposition
of this particular man, but that the disposition not have a
general effect on the community which tends to depreciate
the gravity of the crime and thus imply a license to commit
it” (ALI Proceedings 1961, p. 340).

The tone of the discussions on the relevance of public opin-
ion was set by Corrections Commissioner (and former judge)
Anna M. Kross of New York. Her concern was that public
opinion might sometimes influence judges’ decisions. She
worried that “sentencing power [would] . . . remain in the
hands of a judge who at the time is in the first place influenced
by the temper of the people. He is worried about what they
are going to say about him. He is worried about what the
newspapers are going to print. His judgment, no matter how
well adjusted and how sound and how sane and how honest
he is, necessarily is affected. He is only human” (ALI Proceed-
ings 1954, p. 75).
Only one speaker spoke of public opinion in modern (ca.
2003) terms. Judge James Alger Fee of California, after telling
of his father’s life as a judge in frontier Oregon, observed:
There is one place where I do not agree at all with some
of the sentiments that have been expressed here. I think
that local thought on the subject of penalty is one thing
you should recognize because it is to the community even-
tually and the local community, not some United Nations
or Federal community—that you are really responsible.
It is the people that make and enforce the law, and it is
community sentiment that backs up a law.” (ALI Proceed-
ings 1954, p. 78)
Others acknowledged the existence and force of public

opinion, but as something to be managed or evaded, not as
something to be respected, and thought judges were better
placed than administrators to resist popular pressures and
public opinion. Judge Gerald F. Flood of Pennsylvania, for
example, preferred to place powers in judges’ than adminis-
trators’ hands because judges were better able to withstand
the force of public opinion:
[S]omeone has to take up the shock of the community
outrage and community pressure. And I think that Dr.
[Thorsten] Sellin [about to go to California for ALI to study
California practices] will find (it was a visit to California

that changed my mind, really) the community pressure is
less able to be resisted by a Board than it is by a Judge.
A Judge is elected for a long period, and he can much
more easily resist pressure than can a Board which can be
thrown out of office by a Governor the next morning.
(ALI Proceedings 1954, p. 79)
Wechsler noted the problem, but in an oddly circumscribed

way. The original proposals on minimum sentences allowed
judges to set minimums as long as twenty years for first-
degree felonies such as murder. Wechsler explained this as
a response to public passions:
It often happens that crimes of violence are shocking
crimes and are perpetrated by individuals who three or
four years later do not, to the people who study them,
seem to present the kind of threat for the future that the
particular act that they committed suggests. But the public
remembers the act. Therefore we thought that in this area
where by hypothesis public indignation as to the offense
in the sense of need for a general deterrent penalty would
be very strong, it would be prudent to give the Judge
power at the time of sentence to satisfy the community
that this person would be taken out of currency for a
substantial period of time. (ALI Proceedings 1954, p. 74)
And so the Code provided (though, as eventually adopted,

the twenty years was reduced to ten). It strikes me as note-
worthy that Wechsler described the public demand in terms
of “a general need for a deterrent sentence” rather than in
terms of retributive or expressive needs. It also strikes me
as noteworthy that judges could (but need not) impose up
to a lengthy minimum for a first-degree felony (but only up
to three years for any other), but could also impose probation
in the same case, and could reconsider any prison sentence
at any time within a year of imposition.
This is written in a time when powerful ideas like “truth
in sentencing,” “do the crime, do the time,” and “just des-
erts” have great influence. They are not new ideas. Cesare

Beccaria would have found the concepts though not the
phrases familiar. Herbert Wechsler, and many of the judges,
lawyers, and law professors with whom he worked, would
have been familiar with the writings not only of Beccaria
but also of Kant, Hegel, Stephen, and many other proponents
of ideas that are only very weakly discernible in the Model
Penal Code. That so talented a group of law reformers, work-
ing within so inherently conservative and establishmentarian
a setting as the American Law Institute, produced and won
approval of a document so uninfluenced by powerful and
influential ideas, should make us wonder what similarly pow-
erful ideas are being ignored in our time.
“Prison Works”: Recidivism

of First-Time Prisoners
One striking sign of the differences in penal sensibilities be-

tween our and earlier times is in the way we think, speak,
and write about prisoner recidivism. Two robust recidivism
findings were evident in the 1960s and today. I overstate
them for effect, because inevitably methodological qualifica-
tions warrant mention—for example, results vary with the
measure of recidivism (self-report, rearrest, reconviction, re-
imprisonment), what counts (any offense, only felonies, only
particular types of offense), the follow-up period (one year,
two, three), and the reliability of the data sources—but they
nonetheless express an important contrast.
First, answering a question researchers often asked in the
1950s and 1960s, of people released from a first prison sentence,
two-thirds do not return to prison. I simplify here for rhetori-
cal purposes and do so again when I discuss recidivism rates
based on all releasees. I could have said 60 percent or given
a range, say 60–70 percent, or said approximately two-thirds.
Prison works in the sense that most people sent there do not
want to go back and organize their future lives so as not
to go back. Or, alternately, prison can be said often to be
unnecessary and the low return rates to demonstrate that
many people are sent to prison who present little public
safety risk and could safely, responsibly, and more cost-effec-
tively be sanctioned in other ways.
Second, answering the question researchers asked in the
1980s and 1990s, of people released from prison in a given
year or as part of a representative release sample or cohort,
considerably more than half, often two-thirds, return to
prison. Prison does not work in the sense that most people
released from prison do not manage thereafter to conduct
law-abiding lives and to refrain from the kinds of unlawful
behavior that lead to apprehension, conviction, and impris-
onment.
Both statements can be true because they are statements
about different groups of people. The first statement refers
to people in prison for the first time. Some have done things,
including violent things, that are the product of unusual
circumstances and out-of-character, and they would be sub-
sequently law-abiding whether or not they were sent to
prison. Some may be essentially conformist people with pro-
social values who behaved criminally and are shamed or
shocked by their imprisonment into law-abidingness. Some
may be young offenders who committed crimes but were not
too badly damaged by their prison experiences and after
release found love, work, or God, or just aged out of their
offending years, as most young offenders do. For some, prison
may have provided a short sharp shock that captured their
attention and made them resolve to obey the law. And some,
a minority, are so developmentally disadvantaged, or so pre-
disposed to deviance, or so committed to a life of crime, that
they return to crime and in due course to court and prison.
The second statement refers to all people released from
prison in a representative sample of releasees. This means
they include a small fraction who are being released from
prison for the first time, and have a one-third likelihood of
returning to prison, and a much larger fraction who have
been released from prison before, and have a much higher
likelihood of returning to prison. The clearest example of this
is found in data on return to prison for all persons released

from Florida prisons between July 1, 1986 and February 1992.
Of those with no prior incarcerations, only 31.6 percent re-
turned to prison, compared with 42 percent of those with one
prior incarceration and 52 percent of those with four or more
(Florida Legislative Office of Program Policy Analysis and Gov-
ernment Accountability 1995, exhibit 9).
Because recidivism and reincarceration of first-time prison-
ers have not received much attention for three decades, data
on the subject are not readily available, but other data sup-
port strong inferences. For example, a U.S. Bureau of Justice
Statistics survey of arrest data for all prisoners released in
1983 from prisons in eleven states found that, within three
years, 62.5 percent were rearrested for a felony or serious
misdemeanor, 46.8 percent were reconvicted, and 41.8 per-
cent were returned to prison or jail (Beck 1989). However,
of those releasees with only one prior arrest at the time of
the study (that which led to the incarceration), 38.1 percent
had been rearrested within three years compared with 63.2
percent for the entire sample and in excess of 75 percent for
the half of the group who had been arrested six or more
times (Beck 1989, table 11). Similarly, of all adult Minnesota
prisoners released in 1992, 59 percent were rearrested within
three years, 45 percent were reconvicted, and 28 percent
were reimprisoned for a new offense (Minnesota Legislative
Auditor 1997, figure 3.2). (These numbers are comparatively
low because only state prison sentences are counted.) How-
ever, of those with no criminal history (no prior adult felony
convictions and not more than one minor petty misde-
meanor), 39 percent were rearrested and only 24 percent
were reconvicted (ibid., table 3.6).
None of this should be surprising. Research on age and
crime shows that crime participation rates for property crime
peak in the mid-teens and for violence in the late teens, and
that most criminally active people desist by the early twenties
(Farrington 1986, pp. 189–250). Research on criminal careers
shows that the probability of rearrest given one arrest is not
high but that the probability rises monotonically with each
subsequent arrest, until it stabilizes around 80 percent by
the sixth arrest. Putting the two findings together should
make it apparent that most people sent to prison, and hence
released, will be persistent offenders, and accordingly that
on average people released from prison will present relatively
high risks for recidivism.
Is it not odd though, that in our time, we have forgotten
the lower risk first-timers, and do not even think in our
research to ask separately about them? Social scientists who
self-select to study criminals and prisoners are not especially
known for their conservatism or for voting Republican and
yet, in our time, they have forgotten that the question they
ask—What percentage of prisoners fail?—provides an an-
swer that supports the penal policies of our time, while a
different question about first-time prisoners would not.
The Costs and Benefits of Crime
Equally blind has been the failure to take offenders into

account in calculating the costs and benefits of crime control
policies. There was not much writing on this subject until
Chicago School economists Gary Becker (1968) and Richard
Posner (1977, 1980) began to offer economic analyses of the
likely effects of alternate punishment policies. Their work,
however, was primarily theoretical, built on hypotheses
about individuals’ rational calculations of likely gains and
losses, and the likely general deterrent effects of possible
policies. In this, they resembled Bentham, who offered a vari-
ety of rational actor proposals:
Rule 1. The value of the punishment must not be less
in any case than what is sufficient to outweigh that of the
profit of the offence.
Rule 2. The greater the mischief of the offence, the
greater is the expence, which it may be worthwhile to be
at, in the way of punishment.
Rule 3. Where two offences come in competition, the
punishment for the greater must be sufficient to induce a
man to prefer the less.

Rule 4. The punishment should be adjusted in such man-
ner to each particular offence, that for every part of the
mischief there may be a motive for the offender from giv-
ing birth to it.
Rule 5. The punishment ought in no case be more than
what is necessary to bring it into conformity with the
rules here given.
Rule 6. That the quantity actually inflicted on each indi-
vidual offender may correspond to the quantity intended
for similar offenders in general, the several circumstances
influencing sensibility ought always to be taken into ac-
count. (Bentham 1938–43; quoted in von Hirsch and Ash-
worth 1998, pp. 54–56)
Though written in prose of less than crystalline clarity,

with pondering, Bentham’s meaning becomes clear. The first
four rules are simple deterrence. The penalty should be
greater than the likely gain. Lesser crimes should receive
lesser punishments than more serious crimes to encourage
offenders to commit the lesser. Requiring lesser penalties for
attempted than for completed crimes gives offenders an in-
centive to stop part way through.
The fifth and six rules, however, focus on the offender.
The fifth, the parsimony rule, forbids punishments harsher
than the minimum required to achieve the purpose. The sixth
requires that punishments be tailored to offenders’ sensibili-
ties, the ways that punishments might affect them differently
than other people. Finally, Bentham describes circumstances
when punishment would not be justified, including “where
it is unprofitable, or too expensive; where the mischief it
would produce would be greater than what was prevented”
(von Hirsch and Ashworth 1998, p. 57). In other words,
punishment cannot be justified when, taking all the “mis-
chief” (synonyms used by Bentham elsewhere include pain
and unhappiness) into account, including to the offender and
his family, punishment would cause more mischief than it
would prevent.

Where Becker, Posner, and most others who have written
about costs and benefits of punishment differed from Bentham
was that their emphasis was on comparing only the financial
costs of punishing offenders with benefits that might be
gained through deterrence. Bentham, by contrast, was inter-
ested in measuring society’s, victims’, offenders’, and others’
happiness or unhappiness (or pain and pleasure, utility and
disutility).
For Bentham, punishment was justifiable only when it
served to maximize human happiness, or minimize human
suffering, taking everybody, including offenders, into ac-
count. Punishment, Bentham argued “itself is an evil and
should be used as sparingly as possible . . . it ought only to
be admitted in as far as it promises to exclude some greater
evil” (Tonry 1994, p. 63, quoting Bentham 1789).
A new cost-benefit literature began to emerge in the 1980s,
focused primarily on incapacitation rather than on deterrence
(Zedlewski 1987; Cohen 1988; DiIulio 1990; Kleiman and Cav-
anagh 1990). Much imagination was invested in calculating
the costs of crime and the criminal justice system, but no
one thought to take the offender and his costs and suffering
into account. The principal aim was to calculate whether
“prison works” in the sense that the economic value of crimes
prevented through locking people up outweighed the costs of
operating the criminal justice system, including the prisons.
The earliest article, by Edwin Zedlewski of the U.S. Na-
tional Institute of Justice, concluded that, for one average
prisoner, “a year in prison [at] total social costs of $25,000”
would produce a saving of “$430,000 in [reduced] crime
costs” (Zedlewski 1987, pp. 3–4).It provoked a fierce response
(e.g., Zimring and Hawkins 1988). Should the costs of the police
be included in the calculation, since every society has a police
force? Should crime prevention costs from burglar alarms to
private security firms’ fees? Were the estimates of the numbers
of crimes prevented per locked-up offender accurate or even
remotely plausible? Should both incapacitative and deterrent
effects be predicted or only incapacitative effects?

Gradually, over time, the estimation techniques improved
and general agreements were reached about what should be
taken into account. The most striking development, and the
most influential, was the development by economist Mark
Cohen of estimates of the victims’ direct out-of-pocket ex-
penses, their pain and suffering, and their risk of death (Co-
hen 1988, p. 546). For most crimes, the intangible costs of
pain and suffering and risk of death far outweighed out-of-
pocket costs. For rape, direct costs totaled $4,617, intangible
costs $46,441. For personal robbery, direct costs totaled
$1,114, intangible $11,480; for assault, $422 and $11,606.
Cohen developed these estimates by adapting data from jury
awards in tort (accident) cases, and generalizing from them
to crime.
This was a cockamamie thing to do for a number of reasons.
Only very serious accidents result in jury awards (most are
settled) and he was thus using data on the most serious acci-
dents to estimate the costs of average crimes. Jury awards
are inflated to include attorneys’ fees (a third to a half of
the total). Cohen’s inflated estimates, however, soon became
widely used. This is probably because they were the only
ones around.
Whether or not Cohen’s estimates are plausible, however,
is not the point. What is important is what is missing. No-
where in this literature, including even in vigorous critiques
of its quality and value (Zimring and Hawkins 1995, chap.
7), did anyone take account of offenders. A prison sentence
entails pain and suffering by the offender. When he is re-
leased, he will be stigmatized and face numerous obstacles
to resuming a normal life. These range from ineligibility for
many professions, inability to vote, and employer resistance
to employing ex-cons. He will after prison have lower life
chances, life expectancy, and lifetime earnings than if he had
not spent time in prison. His family also will have suffered,
economically, socially, and psychologically, if it survived
intact.
Clever economists can devise estimates of the economic
value of almost anything. They could have calculated the

costs of punishment policies generally and imprisonment in
particular to prisoners and their families. Jeremy Bentham
understood all this and would have taken all of these things
into account. They are absent from almost all of the cost-
benefit literature on punishment. A recent book by econo-
mists Philip J. Cook and Jens Ludwig (2001) on the costs of
gun violence is a notable, but recent exception.
Paralleling social scientists’ neglect of recidivism by first-
time offenders, social scientists did little better in relation to
the effects of punishment on offenders. My 1995 Malign Ne-
glect was one of the first calls for attention to collateral effects
of imprisonment on prisoners, their families, and their com-
munities. When, in 1997, I commissioned an essay from John
Hagan and Robin Dinovitzer (1999) on the collateral effects
of punishment, they had to report that the literature was
fragmentary and fugitive, though they did their best to pull
it together. On many important subjects—for example, the
effects of a parent’s imprisonment on children’s development
and well-being or the effects of imprisonment on offenders’
later health and life expectancy—there was no respectable
social science literature. The inattention is beginning to end
and work is beginning to appear on the effects of imprison-
ment on prisoners’ communities (Clear Rose, and Ryder 2001)
and on collateral costs more generally (Mauer and Chesney-
Lind 2002).
Under the influence of the penal sensibilities of the final
decades of the twentieth century, even politically liberal so-
cial scientists forgot something that was obvious to Jeremy
Bentham two centuries earlier—punishment hurts offenders,
and ethical policy makers and analysts should factor that
reality into their thought. A half-century ago, the draftsmen
of the Model Penal Code knew about the retributive moral
arguments of Kant and Hegel, but took no account of them.
They thought them antiquarian and vindictive and no longer
relevant to the concerns of a more humane age. Bentham
thought and wrote about the overall costs and benefits of
punishment but, in our time, those who were “for victims”
did not care about the costs to offenders and those who re-

jected the for victims/for offenders dichotomy somehow
overlooked the subject altogether.
That is the last of the non-barking dogs described in this

chapter. Taken together, the various forgotten and over-
looked issues demonstrate a collective partial amnesia, which
is odd, given the universality of the underlying issues and
principles. They demonstrate the power of penal sensibilities
over men’s and women’s minds.

8
Better Understanding, People, and Policies
The United States has adopted criminal justice policies that

reflective people should abhor and that informed observers
from other Western countries do abhor. An anecdote makes
the point. I spoke recently in Noordwijk, the Netherlands,
at a European Union meeting on crime prevention. My assign-
ment was to survey current research findings on “what
works,” and in doing so I described Swedish educational
programs that had successfully and substantially reduced
criminal activities of immigrant children. It has been known
for a century, popular beliefs to the contrary notwithstand-
ing, that first-generation immigrants tend to have lower crime
rates than the resident population, but that second-genera-
tion immigrants typically have higher rates than their parents
or the resident population. Sweden had accomplished some-
thing noteworthy that furthered both crime prevention and
social welfare goals.
As I headed toward my seat, the moderator, Hans Dijkstal,
then the head of the Dutch Interior Ministry (popularly
called, in English, the “cops agency”) and a member of the
Liberal Party, the most conservative party in the coalition
government, thanked me and said that he found my remarks
on minority groups particularly interesting. “It was good,”
he continued, “to hear on such a topic from a member of a
U.S. minority group.” My mind raced as I wondered what he
could possibly mean. Few who meet me would be surprised to
learn that my immediate background is middle-class white
American and that my ancestors are a mix of Germans, Irish,
Scots, and English. After I sat down, the minister went on,
“It is good to have with us one of that minority of American
195
crime experts who hold humane and progressive views.” I
had not met the minister before and no one had seen a draft
of my talk. He could have had no idea I was going to mention
immigrants or Sweden. His quip was impersonal and spur-
of-the-moment.
Another anecdote, this one semantic, illustrates the same
point. In European English, as opposed to American or Brit-
ish, it is common to contrast preventive and repressive poli-
cies. In that usage, “repressive” is a synonym for “law en-
forcement” and “preventive” encompasses a wide array of
social, architectural, and community measures aimed at pre-
venting crime. Once I urged an American friend not to be
defensive about a reference in a Dutch government report
to repressive U.S. drug policies, because, I explained, “re-
pressive” merely meant law enforcement. “No, my friend,”
a Dutch colleague intervened, “When we refer to ‘repressive
U.S. drug policies,’ we mean it in both senses; they rely heav-
ily on law enforcement agencies and they are cruel.” I often
edit scholarly articles written by Europeans for whom En-
glish is not a first language, and for years smugly translated
Europeans’ repressives into words that would not be misin-
terpreted by Americans as polemical. Now I ask every time
what is meant.
As vigorous objections should be made, however, to the
processes by which American crime control policies were set
as to the policies themselves. For a quarter century, American
policy makers year-by-year made policies and punishments
harsher. Different policy makers endorsed different policies
for different reasons, but no one doubts that political self-
interest and cynicism played major roles. By the late 1980s,
Bill Clinton and the Democratic Leadership Council that he
shaped decided never to be less tough than their Republican
opponents, no matter how unjust, wasteful, or ineffective
the policies under consideration.
California’s overbroad three-strikes law was enacted, accord-
ing to Franklin Zimring and colleagues (2001, p. 6), because
Democratic legislators agreed to pass any proposal Governor
Pete Wilson supported, in hopes “that he would back down

from an unqualified ‘get tough’ stand or be politically neutral-
ized.” Wilson did not blink and the law was passed because
both sides were “unwilling to concede the ground on ‘getting
tough’ to the other side in the political campaign to come.”
Similar stories can be told of the passage of most high-
profile crime legislation in the 1980s and 1990s. Rising crime
rates in the 1960s and 1970s influenced policy makers, as no
doubt did public anxieties and fears. Those factors explain
why crime control policy was on legislative and executive
branch agendas, however, but not why policy makers took
so little account of whether or how new policies would work,
what they would cost, or what damage they would do to
offenders, their families, and their communities.
NYU sociologist David Garland explains that policy mak-
ers in the 1980s and 1990s ceased being interested in evi-
dence—of what works, of foreseeable effects—because they
were not primarily interested in crime reduction per se. The
state “abandon[ed] reasoned, instrumental action and re-
treat[ed] into an expressive mode that we might . . . describe
as acting out—a mode that is concerned not so much with
controlling crime as with expressing the anger and outrage
that crime provokes” (2001a, p. 110; emphasis in original).
Garland’s explanation for this is more benign than mine.
He argues that American governments adopted needlessly
harsh policies because politicians doubted any policies would
be effective but believed they would win favor with voters
by at least acting tough. I think policy makers responded
unreflectively to prevailing sensibilities about crime and dis-
order and that cynicism and self-interest played a large role.
Except for white-collar criminals, most offenders come from
poor disadvantaged backgrounds and no powerful interest
groups exist to promote or protect their interests. If policies
promoted to win elections do unnecessary damage to offend-
ers and their families, I suspect many conservative politicians
thought, so be it. That account, however, treats recent poli-
cies as the outcome primarily of selfish politicians and clearly
more than that was going on. Many harsh policies were con-
sistent with prevailing sensibilities. Does that justify them?
BETTER UNDERSTANDING, PEOPLE, AND POLICIES 197

That strong social and psychological forces of a moment
or a year can cause many people to act intemperately and
cruelly may, in some ways, be understandable. To under-
stand is not necessarily to forgive, but it is to understand.
Individuals sometimes lose control and so do collectivities.
The latter demonstrates the power of mob psychology, and
sometimes it takes over. We are at best only partly rational
animals and, regrettable though it is and often in retrospect
regretted, extreme circumstances can beget bad decisions.
For nearly two decades, however, U.S. crime policy has been
driven more by ideology, emotion, and opportunism than
by humane analyses of options and reasoned discussion, and
that is much too long for reason and compassion to have been
held in abeyance.
Most people have a doppelgänger who sits invisibly on a
shoulder and speaks up when we have misbehaved. Freud’s
super-ego and poets’ inner voices perform similar functions.
For the past twenty years those quiet reminders have seldom
been heard or acted upon in relation to crime policy. The
cyclical effects of steeply rising and then sharply falling
crime rates on prevailing sensibilities, and through them on
American penal culture and policies, produced the criminal
justice system Americans now have. That combination of
factors created conditions resembling a continuous moral
panic, and many American policy makers acted as panicked
people do.
How harshly should policy makers of the past quarter cen-
tury be judged? I have no doubt our successors will be con-
demnatory of the severity of American crime and drug poli-
cies, of the lives they damaged and ruined, and of the
foreseeable and disproportionate burdens they placed on
black Americans. And I believe many contemporary politi-
cians will be condemned for the cynical ways they manipu-
lated public anxieties and racial tensions to achieve short-
term political goals. Beyond that it gets blurrier.
Prevailing sensibilities do shape what people think and
believe. No doubt some in public life believed one must be
for criminals or for victims, that harsh policies can reduce

crime, and that harsh policies did reduce crime. In their sim-
plest forms all those claims are wrong. In complex forms the
first is nonsense and the latter two are partly true, but the
full truth is vastly more complicated. What moral judgments
should be made of policy makers who should have known
better but did not? Of policy makers who suspended their
critical judgments for longer than they should have? Of pol-
icy makers who came to realize their policies were not work-
ing, or were doing harm, but continued to support them for
other reasons?
Those are questions for posterity to answer. What needs
to be done now is to figure out how to undo unnecessary
damage and lessen the chances that similar excesses will re-
cur. It helps me think about this to liken it to the ways
individuals try to remake themselves. Imagine an occasion-
ally violent alcoholic who desperately wants to remake his
or her life, reduce chances of relapse, and make up for damage
that was done. Remaking a life requires self-understanding.
It is an ethical enterprise. For an alcoholic man, it includes
recognizing what alcohol does to him, what he does under
its influence, and why that behavior is wrongful and destruc-
tive. Reducing chances of relapse requires self-aware efforts
to understand the dangers posed by various situations and
people and to devise ways to lessen those dangers. Making
up for damage requires working to understand what damage
has been done, devising a plan to undo it or make amends,
and then carrying it out.
The corresponding three tasks at a collective level are ethi-
cal introspection, institutional change, and remediation.
Breaking loose from contemporary American crime control
politics will require us, like the metaphorical drunkard, to
address all three tasks. We need to learn to restrain our collec-
tive emotions in public life, to change governmental institu-
tions to provide greater insulation from the passions of the
moment, and to devise structural arrangements that will force
greater reflectiveness on policy makers. Changing collective
emotions and governmental arrangements are daunting ambi-
tions, so though I address them and suggest what needs to

be done, closer and more detailed attention is given to more
modest policy proposals that can undo damage and lessen
its future likelihood.
Ethical Introspection
U.S. crime policy is at a place where no one would have

wanted it to be for a combination of historical and structural
reasons. Because of an interaction between changes in sensi-
bilities about crime and criminals, and a series of moral panics
about crime and drugs, harsh and simplistic policies were
proposed at a time when they found uncritical support. Be-
cause American institutions of government do not buffer pol-
icy from the effects of short-term emotionalism, as do institu-
tions in other Western countries, harsh and simplistic policies
were adopted. Politicians’ risk aversion makes change diffi-
cult. The first step is to admit what has happened and why
and then to set about the task of learning to restrain our
collective emotions.
By that pompous-sounding phrase, “learning to restrain
our collective emotions,” I mean trying to incorporate into
public life some of the ethical folk wisdom of private life:
“Don’t take it out on your child.” “Sit down and count to
ten.” “Never strike in anger.” “Write the angry letter today
but don’t send it until tomorrow.” Aphorisms like these are
commonly part of our personal ethics because we know that
things that make us angry or depressed often make us irrita-
ble and at danger of overreacting or taking out our upset on
whoever is close at hand.
The ways we react privately to people who unfairly or
unreasonably take out their anger or frustration on others
provide a useful analogy for thinking about crime control
policy. We become progressively less empathetic as the harm
caused increases. Some expressions or displacements of anger
are understandable and tolerable, some are less so, and some
are not at all.

We understand and sympathize when someone swears or
pounds a fist on the desk or throws something to the floor
in the face of great unfairness or disappointment. It has hap-
pened to everyone. We know how they feel. They are letting
off steam, and not hurting anyone else, so why not? If it
helps, more power to them.
We understand but sympathize less when the upset pro-
duces unkind or hurtful words directed at a serendipitous
object whose misfortune is to be in the wrong place at the
wrong time. This might be thought the level at which “sit
down and count to ten” or “write the angry letter but . . .”
are apposite. Yes, we understand how you feel but that is
no justification for acting unfairly or hurtfully toward some-
one else. If we had done what you did, we would feel guilty,
and should feel guilty. Pounding a fist on a table is an expres-
sive action that communicates unhappiness. Screaming at
someone also expresses unhappiness, but it hurts the recipi-
ent or makes him uncomfortable. That is not fair. Expression
of the unhappiness can be justified. Hurting the bystander
cannot. A good person would be sufficiently self-aware to
recognize his or her emotional state and find a way to avoid
hurting someone else because of it. If not, a good person
would soon realize they had behaved badly, feel guilty, apol-
ogize, and try to make up for it.
Sympathy disappears altogether when the upset produces
physical violence toward a serendipitous victim, or emotional
abuse of a child, spouse, or employee or someone who is
handicapped or otherwise vulnerable. This is the “never
strike in anger” or “don’t take it out on your child” level.
There can be no justification. Hurting someone for no reason
that relates to them is wrong, and that is that.
These are simple but timeless ideas. Anger is the emotion
that underlies the populist punitivist impulse to hurt wrong-
doers, and that gives rise in turbulent times to calls for more
executions and harsher punishments. The style and content
of recent policies, David Garland wrote, are shaped by “the
sense of a fearful, angry public” (2001a, p. 10).

And yet, most people understand, as Pietro Aretino, the
Renaissance poet and critic, wrote, “Angry men are blind
and foolish, for reason at such times takes flight, and in her
absence anger plunders all the riches of the intellect” (Gray-
ling 2001). Seneca, the Stoic philosopher, called anger “the
most hideous and frenzied of all the emotions,” and urged
resistance to its demands. Failing that, he continued, “There
are two rules: avoid anger if you can, and if you cannot, in
your anger do no wrong” (Grayling 2001).
Anger is humanly understandable, the ancients acknowl-
edged, but not if it leads to behaviors that are inappropriate
or disproportionate. “It is not easy,” Aristotle wrote, “to
define how, with whom, at what, and for how long one should
be angry, and at what point right action ceases and wrong
begins. . . . But so much is plain, that the middle state is
praiseworthy—that in virtue of which we are angry with
the right people, at the right things, in the right way, and so
on, while the excesses and defects are blameworthy” (Barnes
1984, p. 1777).
“In your anger, do no wrong.” That does not mean that
we should not be angry with people who commit crimes or
that they should not be punished. It does mean that their
interests should be fairly considered and that they should
not be punished more severely than can in principle and
humanity be justified.
Precisely how severely criminals may justly be punished
depends on why punishments are imposed, and reasonable
people differ in their views about this. Retributivists believe
that punishment should be tied closely to offenders’ blame-
worthiness, and that the amount of punishment should be
scaled to the relative seriousness of their crimes. The more
serious the crime, the severer the deserved punishment, and
vice versa.
Consequentialists believe that punishment can be justified
only in terms of its good effects. Consequentialists also be-
lieve that pain and unhappiness, including those of offenders,
are undesirable. Punishments accordingly may be imposed
only to such an extent as can be justified on the basis of

evidence or plausible hypotheses that their crime prevention
effects will prevent more pain and unhappiness than they
cause.
The ethical frameworks that retributivists and consequen-
tialists consider relevant are different, but they agree that a
punishment that cannot be justified in principle is unjust.
Consider, for example, the notorious case of Jerry Dewayne
Williams, the man who was sentenced to twenty-five years’
imprisonment under California’s three-strikes law for taking
several slices of pizza. Retributivists would consider that pen-
alty unjustly severe on proportionality grounds, since people
convicted of much more serious, violent crimes receive much
less severe sentences. If a rape or a robbery results in a prison
sentence of a few years, how can twenty-five years possibly
be justified for such a minor offense? That violates the princi-
ple that just punishments should be scaled to offenders’ moral
culpability and accordingly that more serious crimes should
receive severer punishments than less serious crimes. Conse-
quentialists would consider that penalty unjustly severe be-
cause no plausible case can be made that its crime prevention
effects could be sufficiently large to justify depriving the
offender of most of his life. The current state of knowledge
concerning rehabilitative, deterrent, and incapacitative ef-
fects of punishment cannot justify twenty-five-year sen-
tences for people who take property of nominal value.
The pizza slice case is extreme, but the analysis would
apply to, and condemn, many contemporary American sen-
tencing laws as unjustifiably severe and insufficiently re-
spectful of offenders’ interests. Ten- and twenty-year sen-
tences for routine drug offenses cannot, for example, be
squared with substantially shorter sentences for most as-
saults, rapes, robberies, manslaughters, and some murders.
The laws requiring such harsh penalties for drug offenses
reflected the sensibilities of the moment of their adoption
and expressed understandable anger at criminals and crimes,
but that does not justify them. They are in this respect the
collective equivalent of the understandably upset woman
who in her unhappiness inexcusably strikes her child. The

drug law penalties are wrong. They violate the injunction
never to strike in anger and they do great and unnecessary
damage to peoples’ lives.
What is unjustifiable in private lives is unjustifiable in
public life. Moral panics, cyclical patterns of moral intoler-
ance, and anxieties associated with fundamental social changes
all increase pressures for adoption of harsh, inhumane, and
ill-considered crime control policies. Knowing that, policy
makers should try all the harder to separate the ephemeral
and emotional from the lasting and reflective, and to resist
those pressures. Most elected officials in their private lives
no doubt do try to manage their outbursts of emotion and
their displacements of anger. If they did so in their public
lives, contemporary American crime policies would look very
different.
Institutional Change
Exhortations to resist anger and excess and to embrace mod-

eration and principle are unlikely to provide strong protec-
tions against the surging passions of a moment. Members of
the Congressional Black Caucus generally favor preventive
programs over punitive policies, and generally oppose man-
datory penalty laws because of their disproportionate effects
on black defendants. Under the influence, however, of the
moral panic associated with the appearance of crack cocaine
in inner cities and the crack overdose death of Len Bias, a
University of Maryland basketball star who was expected
to become an NBA superstar, eleven of twenty-one black
congressmen voted for the 1986 law creating the federal
crack/powder 100-to-1 sentencing policy (R. Kennedy 1997,
p. 370). Nearly all later changed their minds. The Caucus
has been an outspoken opponent of the 100-to-1 rule and
repeatedly sought its repeal, but at the time, for a while,
some of its members got carried away and supported a law
that, on cooler reflection, they realized is unwise and unjust.

A parallel story, this time of personal tragedy and horror,
is part of the backdrop to passage of California’s three-strikes
law. Shortly after the murder of 12-year-old Polly Klaas, kid-
napped from a slumber party in her bedroom by Richard
Allen Davis, a sex offender recently released from California’s
prisons, her father, Marc Klaas, was asked to endorse the
proposals for a California three-strikes law. “I was very dis-
traught,” he later recalled, and endorsed it without much
thought, according to former U.S. News & World Report re-
porter Ted Gest (2001, p. 194). After a while,
Marc Klaas was having second thoughts. He had endorsed
three strikes in a moment of anguish. Now he and his
family decided that Reynolds’s plan was too broad. In
some ways, it made little sense. It was fine to treat repeat
criminals harshly, but the possibility that just about any
felony could count as the third offense meant that a third-
time bicycle thief could get a prison sentence five times
as long as a first-time murderer. Klaas and his allies . . .
withdrew their support. (Gest 2001, p. 195)
It was too late. Too much momentum had built up. The pro-
posed law was enacted by the legislature and later approved
by voters in a referendum.
Why is it that only the United States among Western coun-
tries experienced skyrocketing prison populations or adopted
policies such as three strikes, life sentences without possibil-
ity of parole, or increased use of capital punishment? Horrify-
ing crimes occasionally happen in all countries, and crime
rates rose sharply in all Western countries from mid-century
through the early or mid-1990s. One reason the United States
adopts extreme policies, as other chapters show, is that the
beliefs and opinions shaped by long-term sensibility cycles
interacted with the passions generated by moral panics to
foster willingness in the United States to take too extreme
measures. Those provoked by the deaths of Polly Klaas and
Len Bias are but two of many. But why do overreactions like
those occur often in the United States and seldom elsewhere?

Judges and Prosecutors
as Career Civil Servants
The answer relates to differences in legal culture and the
structure of government between the United States and other
countries with which it is ordinarily compared. The institu-
tional organization of the American criminal justice system
and its close connection to electoral politics conduce to puni-
tive excess. In every other Western country I know, the
prevalent view among public officials and criminal justice
practitioners is that sentencing and punishment should be
insulated from short-term emotionalism and swings in public
opinion. This means that discretion over individual cases
should be left in the hands of judges and corrections officials
who are close enough to the facts of the individual crime
and the circumstances of the individual criminal to deal with
them sensibly and justly. That is why three-strikes, lengthy
mandatory minimum, and life-without-possibility-of-parole
laws are rarely adopted in other English-speaking countries
and never in continental Europe. Other countries’ legislatures
have the power to enact such laws. They just do not use it
because they do not think such laws are a good idea.
Consistent with the idea that just disposition of individual
cases should be insulated from political influence and short-
term public opinion is the practice in other Western countries
that judicial system officials are appointed, not elected. In
much of continental Europe, people who wish to become
judges or prosecutors receive specialized training from law
school onwards and become career civil servants. In England,
Canada, and Australia, they are appointed on the basis of
merit after professional screening. They are substantially re-
moved from partisan political influence and the temptation
to do things solely because they are politically popular. Most
senior Justice and Interior Ministry officials are career civil
servants. And crime policy is commonly seen, including by
elected officials, as something that should be insulated from
short-term political pressures and that should be based to a
large extent on systematic evidence and accumulated experi-

ence. American political values and institutional arrange-
ments provide few buffers.
Achievement of moderate, humane practices and policies
is easier in countries where judges and prosecutors are non-
political, career civil servants, or where they are appointed
under circumstances in which partisan considerations are
generally regarded as inappropriate. Career officials are more
likely than politically selected officials to decide individual
cases on the merits of their distinctive circumstances and
to consider policy proposals from long-term perspectives of
whether they will improve the quality of justice or the effec-
tiveness of administration. Commitment to abstract principles
of justice is part of the professionalism and professional self-
esteem of career officials, and buffers individual decisions
and policy choices from raw emotions and officials’ self-
interest.
Politically selected officials, by contrast, are more likely
than career officials to be influenced by short-term or self-
interested considerations. At the extreme, many American
prosecuting attorneys are continuously campaigning and for-
mulate their law reform proposals, their general policies, and
their tactics in individual cases on the basis of how they will
be reported on the evening news programs, and how the
reports will be received by the general public. Elected judges
are not immune from influence by consideration of the effects
of their decisions on their re-election prospects. Honest, hon-
orable judges I have known have privately admitted to me
that an impending election makes them less likely than at
other times to impose a sentence others might see as unduly
lenient.
American officials, conversely, are less likely to be influ-
enced by long-term considerations, institutional history, and
abstract principles of justice. Many judges come from civil
law practice backgrounds that give them little knowledge
of criminal law, criminal courts, or criminal justice policy.
American judges receive no special training in law school,
often take up those roles in middle age, and typically have
shorter careers than their counterparts in other Western
countries. This means that they often lack awareness of the
past experiences of their offices or of their profession, and
lack a personal or professional interest in long-term effects
of changed policies on the justice system.
Where turnover is high, institutional memories are short
and commitment to professional values is weak. Many elected
state prosecutors and appointed U.S. attorneys serve only
for a few years and aspire to be appointed to a judgeship,
to be elected to higher office, or to enter a lucrative private
law practice. Assistant state and federal prosecutors typically
start immediately after law school and plan to stay only for
a few years while they get litigation experience. Many then
enter private law practice and earn much more money.
Changing institutional arrangements so that judges and
prosecutors are less likely to act out of anger, cupidity, igno-
rance, or self-interest will not prevent bad decisions and
adoption of bad policies, but it will help. The goal should be
to shift from the current American model of criminal justice as
an instrument of politics, in which prosecutors and judges
are elected, to a professional model of criminal law as an
instrument of justice, in which they are career civil servants.
Replacing elected judges and prosecutors with career civil
servants would, over time, change the legal culture and re-
vive support for the once self-evident idea that judicial and
other proceedings affecting individuals’ liberties should be
insulated as much as possible from political influence and
public emotion. As chapter 7 shows, as recently as the 1950s
those were uncontroversial mainstream ideas in the United
States.
A legal system in which judges and prosecutors are career
civil servants would improve penal policies and practices
indirectly and directly. The more powerful but indirect ef-
fects would derive from recognition that penal policies that
fundamentally affect and diminish offenders’ lives should
not be the subject of partisan and self-interested politics.
Moving toward a civil service model would signal that recog-
nition. In addition, once a professional judiciary and prosecu-
tion system was in place, judges and prosecutors would be-
come sources of professional advice and informed experience
rather than active politicians with personal agendas to pur-
sue. They would also likely act as points of resistance to
penal policy proposals that were ill-considered or motivated
primarily by elected officials’ pursuit of partisan or ideologi-
cal advantage.
Their direct influence would come from their day-to-day
work. They would be less likely than elected officials in
dealing with individual cases to be influenced by political
considerations, media attention, or public clamor. Personal
promotion would be influenced by talent, hard work, and
demonstrated observance of professional ethics and norms,
rather than by editorial pages, public opinion polls, and focus
groups.
Professionalizing the judiciary and prosecution, and re-
moving them from direct partisan political pressures, would
not, however, mean that public attitudes and beliefs would
not influence the courts and sentencing. A sizable political
science literature shows that courts have distinctive, broadly
shared, legal cultures that shape perceptions of justice, that
change over time, and that reflect community values. The
ways lawyers behave and interact, and the decisions judges
and prosecutors make, are constrained by the local legal cul-
ture. Most courts have “going rates” for sentencing. Few
people like to be accused of behaving unreasonably or un-
justly and the going rates provide the criteria by which as-
sessments of injustice are made. That is one reason why prac-
titioners circumvent mandatory minimum and three-strikes
laws. Harsh contemporary laws often prescribe penalties that
are just too severe according to local standards. Circumven-
tion is a way to avoid imposing sentences that everyone in-
volved considers unjust.
Judges and prosecutors are parts of the communities in
which they live and work, however, and are not immune
from influence of changes in community norms and values.
In the United States in the 1970s and 1980s, for example, the
day-to-day decisions of judges and prosecutors toughened in
parallel to toughening of public attitudes about crime. Alfred
Blumstein and Allen Beck (1999) showed that the principal
driver of the rising prison population in the 1970s and 1980s
was increases in the percentages of convicted offenders being
sent to prison. It is often the case that changes in sentencing
patterns spelled out in new legislation begin before the legis-
lation was enacted. That is why, as chapter 5 shows, the
apparent effects of many legal changes are often better under-
stood as the effects of changes in sensibilities. As public
attitudes toward the seriousness of family violence have
changed, for example, so have the attitudes and decisions of
judges and prosecutors. That happens as a result of organic
changes in community values.
There is no good reason why such changes should not
affect how the justice system operates, and they will occur
whether judges and prosecutors are politically selected or are
career civil servants. Organic changes in community values,
however, are different from raw, short-term emotions and
partisan and self-interested politics. The former are an inevi-
table and desirable influence on the legal system. The latter
are the reason why American penal policies are as they regret-
tably are.
Delegated Policy Making

If making judges and prosecutors into career civil servants
seems too radical or large a step to take, another way to
depoliticize and de-emotionalize penal policy is to delegate
rule-making authority to specialized administrative agencies.
Experience with existing sentencing commissions shows that
delegation can work. That experience could be extended to
parole and pardons.
Judge Marvin Frankel’s initial proposals for creation of
sentencing commissions and guidelines had two premises.
First, administrative agencies would accumulate specialized
expertise that would enable them to do a better job at formu-
lating and overseeing sentencing policy than could legisla-
tures. Second, sentencing policy needs to be insulated from
partisan politics and short-term expediencies. Delegation of

rule-making authority to an administrative agency is a good
way to do that (Frankel 1972). The irony is that, while some
commissions’ guidelines showed they could achieve their
functional goals more effectively and justly than legislatures
had, some commissions abandoned the effort to insulate pol-
icy from politics (Tonry 1996, chap. 3).
Several states’ commissions had notable successes at achiev-
ing the goals set for them. Sentencing disparities were re-
duced generally and in relation to ethnicity and gender.
Greater consistency made sentencing fairer by increasing the
chances like-situated offenders would receive comparable
sentences. Greater consistency made sentencing more pre-
dictable and thereby enabled states to improve their correc-
tions planning and budgeting. Publication and use of guide-
lines and adoption of requirements that judges explain their
decisions brought greater transparency to sentencing. That
enabled offenders, victims, the media, and the general public
to know what was happening and why. Guidelines and rea-
sons requirements made appellate sentence review possible
and made judges more accountable for their decisions.
In some states that created sentencing commissions, includ-
ing Minnesota, Washington, Oregon, and North Carolina, the
commissions for a time did significantly insulate sentencing
policy from political partisanship and self-interest. The citi-
zens of those states were spared some of the excesses that
swept the country. All of those states had less steep increases
in prison populations and budgets than the national average
and all managed to tie their sentencing policies to their correc-
tional resources. In other jurisdictions that established sen-
tencing commissions, including the federal government, Vir-
ginia, and Florida, sentencing commissions abandoned their
insulating functions and competed with elected officials to
show who was tougher. In Oregon and Washington, sentenc-
ing policy changes enacted by legislation and referendums
eventually overrode the commissions’ more moderate and
rationalistic policies.
It is not too late to revive the idea of the sentencing commis-
sion as an administrative body with specialized knowledge,
an institutional memory, and some distance from electoral
and legislative politics. Franklin Zimring and colleagues (2001)
show how criminal justice policy in California in the 1990s
was distorted by its centrality in partisan and personal poli-
tics, and argue that something like a sentencing commission
is needed to lessen the odds of that happening again. The
Federal Reserve Board is an example they give of a successful
administrative body that has been partially removed from
partisan politics because of the importance of its function,
the need for specialized expertise to carry it out, and the need
for insulation from short-term political pressures. Policies
concerning deprivation of citizens’ liberty, they argue, de-
serve similar protection.
The considerations that support creation of sentencing
commissions also apply to parole and pardon. More than half
of American states retain parole boards with authority to
decide when prisoners are released. And if recommendations
offered below for establishing systems for reviewing long
sentences imposed in recent years are adopted, more states
will establish parole-like agencies or processes. Most of the
aims of sentencing guidelines—consistency, fairness, pre-
dictability, transparency, accountability—also apply to pa-
role. Parole commissions, like sentencing commissions, set
general policies, and those policies guide decisions in individ-
ual cases. The public interest in having decisions to release
people from prison made consistently, fairly, predictably,
openly, and accountably is no less than or different from the
public interest in having decisions to lock people up satisfy
those criteria.
The notoriety surrounding President Clinton’s late-night
pardons his last day in office in January 2001 demonstrate
why such decisions should also satisfy those criteria. Differ-
ent people for different reasons objected to various of the
Clinton pardons. The controversy resulted mostly from ap-
pearances that pardons went to personal cronies, campaign
contributors, and people whose causes were promoted by
Clinton family members and politically connected lawyers,

some of whom were paid large sums to exercise their personal
influence.
A pardon is an act of executive grace, but to be seen as
legitimate it needs to be handled openly and fairly. The tradi-
tional reasons for pardons, such as the belief that a miscar-
riage of justice has occurred or that a prisoner’s ill-health,
age, or other special circumstances justify early release or
exoneration, are different from those for why sentences are
imposed or why routine parole release is granted. There is
no reason, though, why pardon decisions should not be made
fairly, openly, and accountably. That is why many states
have pardon boards and why the U.S. Department of Justice
includes the Office of the Pardon Attorney.
The sentencing, parole, and pardon commission proposals
in the preceding few paragraphs can be adopted through
ordinary legislation. Those in the preceding section calling
for radical changes in the selection of judges and prosecutors
and in the organization of public prosecution and the judi-
ciary can be accomplished in most states and the federal
system only by means of constitutional amendments. That
would be difficult and time-consuming even were there wide-
spread support for depoliticizing American crime control and
sentencing policy. That support does not now exist every-
where, though it may in some states.
Remediation
Calling for moderation in the national tendency toward emo-

tional excess and for fundamental changes in political culture,
constitutional arrangements, and governmental structure is
to call for a lot. The call should be made nonetheless and
efforts should be made to act on it. Less heroic steps, however,
can move the United States from its current policies and undo
some of the damage they have caused. The two overriding
goals for crime policy change should be to refashion policies
to make them less severe, costly, and destructive, and to

diminish the disproportionate burdens they impose on mi-
nority communities and people.
The emphasis on minority groups warrants explanation.
Current crime control policies cause avoidable collateral dam-
age to offenders, their families, and their communities. Pris-
oners lose years, sometimes decades, of their lives and are
exposed to destructive influences in prison, but they also
often lose their families and their livelihoods. They exit
prison with the likelihood of reduced employment prospects,
average earnings, lifetime earnings, and life expectancies.
Their partners lose loved ones and income support. Their
children lose parents and the experience of family life and
suffer from reduced living standards. Their communities of-
ten become places characterized by single-parent families,
welfare dependence, abnormal age composition, and too few
young men to work or become partners to young women.
That damage is disproportionately done to minority groups
and communities and interferes with achievement of policy
goals relating to civil rights and social inclusion. The as-yet-
incomplete accomplishment of the goals of the Civil Rights
Movement is one of the great successes of modern America,
but the time when race and ethnicity do not limit individuals’
life chances is yet to come. Educational, housing, public
health, and other policies aim to further the goals of the Civil
Rights Movement. Contemporary penal policies undermine
those efforts.
The disproportionate burden of contemporary crime con-
trol policies on members of some minority groups results
from several things. Black and Hispanic Americans are more
extensively involved than whites in the kinds of drug and
violent crimes that provoke lengthy prison sentences. The
extreme example is the federal 100-to-1 crack/powder co-
caine sentencing differential: More than 90 percent of those
sentenced in federal courts for crack offenses are black.
Blacks and Hispanics are also more extensively involved than
whites in violent crimes, particularly murder, rape, and rob-
bery. To a real but unmeasured extent, members of minority

groups are adversely affected by officials’ conscious bias and
unconscious stereotyping.
Members of some minority groups are affected by unequal
distributions of social capital and life chances, which explains
why they become disproportionately involved in inner-city
drug crimes and violent crimes. Relative to other available
opportunities, dealing in drugs and participating in gangs
often seem sensible things to do. People barred from achiev-
ing self-esteem from conventional school and work sources
seek it from deviant peers, gangs, and the satisfactions of
illegal incomes. People with little to lose take risks that others
do not.
Finally, minority defendants are adversely affected by the
implementation of policies and practices that are not meant
to treat minority citizens especially harshly but work out
that way in practice. These are usually referred to as problems
of “disparate impact.” An example is decisions about pretrial
release on bail or “personal recognizance,” which are usually
based on whether criminal defendants have a job, a good
education, a permanent home, and a stable home life. Poor
people are less likely to have these things, and disproportion-
ate numbers of blacks and Hispanics are poor. Standard bail
release policies thus result in relatively fewer black and His-
panic defendants being released before trial and thus produce
a “disparate impact.”
Changes in crime control policy can not address all those
problems but can address some and lessen their adverse ef-
fects. Only a few of the following proposals expressly address
minorities’ issues per se. If policies are too harsh, simplistic,
costly, and destructive, they are those things for everyone
they affect. Changing them, however, will disproportionately
benefit people they now disproportionately damage.
First, enact “safety valve” laws authorizing corrections offi-
cials or specialized administrative agencies to award early re-
lease, according to established criteria, to anyone sentenced to
more than ten years’ imprisonment who has served the lesser
of one-third of the sentence or five years. Those not released

then should be reconsidered for release not less than every two
years afterwards. In most Western countries, fewer than 1
percent of prisoners serve sentences of five years or longer,
so five years seems a reasonable maximum time before release
consideration. Being eligible for release after serving five
years does not mean prisoners will be released. Some crimes
are so serious or notorious that imprisonment for longer than
five years is justifiable. Many offenders who committed very
serious crimes would no doubt continue to be confined.
Guidelines could be established for such decisions. The de-
tails and the criteria would vary from state to state.
U.S. prisons contain huge numbers of people serving de-
cades-long sentences for non-violent drug and property
crimes and of otherwise law-abiding people serving long sen-
tences for violent crimes committed under circumstances of
great emotion or stress. Many of these people should be re-
leased long before their current sentences expire. Individual-
ized consideration, case-by-case, of whether their sentences
are simply too long is the only way to assure that happens.
Second, repeal all three-strikes laws and any laws requiring
judges to impose mandatory minimum sentences. Nearly every
nonpartisan examination in the last fifty years of the desir-
ability of mandatory minimum and similar sentencing laws
has concluded that they are unwise, and for nearly always
the same reasons. They are unnecessary for really serious
cases that would command long sentences anyway. They
have either no discernible deterrent effects or short-term ef-
fects that soon waste away. They result in unjustly harsh
punishments for some offenders. They foster cynical circum-
vention by judges and lawyers. The unpredictable interac-
tions of the previous two deficiencies produce stark dispari-
ties in sentencing of like-situated offenders. Because they
typically apply to drug and violent crimes, for which compar-
atively more blacks and Hispanics are arrested than whites,
they worsen racial disparities in prison.
Repealing three-strikes and mandatory minimum sentence
laws does not mean that people who commit serious crimes
will not receive appropriately severe sentences. That can be
assured by adoption of systems of sentencing guidelines that
direct judges to impose particular sentences unless they give
good reasons for ordering some other sentence. If prosecutors
disagree with the judge’s reasons, appeals can be filed and
appellate courts can decide whether the trial judge’s reasons,
and the sentence imposed, are sound. A dozen or more states
now operate systems of sentencing guidelines that operate
in the way I just described. In their current forms, mandatory
minimums prescribe sentences of ten, twenty, and thirty
years, and in some cases life, for serious crimes.
Third, if three-strikes and other mandatory minimum sentence
laws are not repealed, they should be amended to authorize
judges to impose a lesser sentence if in their judgment the pre-
scribed sentence would be unjustly severe. The principal reason
why mandatory penalty laws are typically applied inconsis-
tently, and why judges and lawyers often devise ways to
circumvent their application, is that they mandate harsher
sentences than local practitioners consider just. England and
Australia, the only two Western countries to adopt manda-
tory sentence laws based on U.S. models, almost always in-
cluded provided-however clauses that authorize judges to
pull the laws’ punches in cases in which they believed that
to be appropriate. The United States should do likewise.
Fourth, reconsider sentencing guidelines and laws with a view
to establishing whether less severe penalties could achieve the
same or comparable crime prevention effects without unduly
depreciating the seriousness of the crimes to which they apply.
This is happening to a limited extent in some states already,
usually indirectly by creating diversionary policies that ex-
empt classes of offenders from general rules, rather than by
explicitly reducing the severity of punishments. Washington
State, for example, introduced drug treatment and boot camp
options into its sentencing guidelines as a way to substitute
shorter and more constructive sentences for offenders other-
wise subject to lengthy mandatory minimums. Even the U.S.
Congress in 1994 enacted a “safety valve” exception to the
mandatory minimums for some first-time drug offenders. In
addition, led by a California policy adopted as a result of a

citizen referendum, some states’ policies now prescribe diver-
sion into drug treatment of first- and second-time offenders
charged with nonviolent crimes. Drug courts, established in
hundreds of American counties, can also be seen as diversion
programs, since many receive their clientele as the result of
prosecutorial decisions to redirect cases from regular process-
ing or judicial decisions to sentence to the drug court rather
than to some other penalty.
However, there is no good reason not to decrease the sever-
ity of sentences across-the-board. No credible evidence sup-
ports the belief that a five-year prison sentence is a better
deterrent than three years, that three years are better than
two, or that incremental changes in punishments have any
deterrent effect. Certainty of punishment is what matters,
not severity. Nor does existing evidence on the incapacitative
or rehabilitative effects of prison sentence support current
sentence lengths.
Fifth, amend all existing sentencing laws to add sunset clauses
that provide for automatic expiration after ten years and require
inclusion of such clauses in all new sentencing laws. If the
“don’t-take-it-out on your . . .” adjuration were extended to
legislation, sunset laws would be ubiquitous. Sunset provi-
sions force legislators to reconsider whether punishment poli-
cies that once seemed to make sense still do. The English
government, for example, acknowledging civil liberties con-
cerns, included a five-year sunset clause in recent legislation
that strengthened law enforcement powers concerning ter-
rorism. The U.S. Congress likewise included sunset clauses
in antiterrorist legislation enacted in 2001.
The risk aversion of U.S. politicians in relation to crime
control policies makes the case for sunset clauses especially
strong. Very few people off-camera any longer defend the
federal 100-to-1 crack/powder sentencing differential. Crack
and powder are, after all, pharmacologically indistinguish-
able. Overwhelming and uncontested evidence shows that
the 100-to-1 rule is a major cause of racial disparities among
federal prisoners, but it remained on the books seventeen

years after it was enacted and few elected officials dared
openly oppose it.
In 1993 the U.S. Sentencing Commission recommended that
the differential be eliminated altogether, a conclusion that
Attorney General Janet Reno initially endorsed. Ted Gest
(2001, p. 126) reports, however, that “it seemed that hardly
any legislator ever wanted to be associated with lowering a
crime penalty,” and President Clinton and the Congress did
no more than agree on a bill to override the commission’s
proposal. In 1997, the commission tried again and proposed
retaining some but a lesser differential. Attorney General
Reno and “drug czar” Barry McCaffrey proposed a 10-to-1
differential. Nothing happened. In 2002 the U.S. Sentencing
Commission issued a major report on the racial disparities
caused by the rule. This time they did not propose eliminating
a crack–powder distinction but did propose narrowing it. In
March 2002 Bush administration spokesmen unequivocally
expressed their support for maintaining the 100-to-1 differ-
ential. Had the 1986 legislation contained a sunset clause,
however, the 100-to-1 rule that few people endorse on the
merits would almost undoubtedly have lapsed.
Sixth, require all proposals for sentencing legislation to be
accompanied by or subjected to impact projections that make
their resource implications clear. This should not be controver-
sial. Some states with competent sentencing commissions do
it as a matter of routine. The first sentencing guidelines com-
mission, in Minnesota, was the pioneer (Knapp 1984; Parent
1988). Its enabling legislation contained broad language di-
recting the commission to take available resources into ac-
count in developing guidelines. The commission interpreted
the language as a legislative mandate that any guidelines
developed must satisfy a resource constraint policy. The Min-
nesota commission interpreted this to mean that the projected
implementation of the guidelines should not produce a prison
population in excess of 95 percent of the rated capacity of
the state’s prisons. If impact projections showed the limit
would be exceeded, the commission would have to, and sev-

eral times did, change the guidelines to meet the target. Or,
if proposed penalty increases for a particular crime would
cause the prison population to exceed the limit, either the
proposal had to be changed or penalties for some other crimes
be reduced to offset the forecast increase. Oregon, Washing-
ton, North Carolina, and other state commissions followed
suit and adopted similar population constraint policies. Not
all commissions did. Pennsylvania, for example, expressly
chose not to be guided by resource considerations and the
U.S. Sentencing Commission simply ignored a statutory direc-
tive to take account of available resources.
Experience, however, makes clear that states that want to
manage their prison populations and corrections resources
can. Policy histories of the North Carolina (Wright 2002),
Minnesota (Frase and Dailey 2004), and Washington (Boerner
and Lieb 2001) sentencing commissions all tell of proposed
legislation that was abandoned or substantially altered when
its resource and financial implications became clear. Other
states also, generally in less sophisticated and effective ways,
require fiscal notes or correctional impact projections.
Seventh, require all proposals for sentencing legislation that
is projected to require new resources to contain provisions au-
thorizing whatever additional resources are needed and appro-
priating funds to pay for them. In the 1970s, state legislatures
routinely passed laws requiring more and longer prison sen-
tences while treating corrections resources as if they cost
nothing and somehow would become available when need
arose. Severe overcrowding resulted with prisons in some
states operating at 150–200 percent of capacity. At various
times during the 1980s, individual prisons and sometimes
entire prison systems in as many as forty states were subject
to federal court orders based on findings of overcrowding
and other unconstitutional conditions. Some states, including
Florida, North Carolina, and Texas, operated under federal
court orders or consent decrees establishing population caps
that required that prisoners be released whenever the cap
was exceeded.

The federal courts’ intervention and prison riots eventually
captured policy makers’ attention. A few states, including
North Carolina and Kansas, adopted sentencing guidelines
and brought prison numbers under control that way. A siz-
able number of other states established sentencing commis-
sions and directed them to develop guidelines that would take
account of available prison resources. Some states—Texas,
Oklahoma, and Florida are the extreme cases—undertook
massive programs of prison building. Future programs of
prison and criminal justice system expansion should, in fair-
ness to taxpayers and to offenders, be pay as you go.
Eighth, require all proposals for sentencing legislation to be
accompanied by or subjected to impact analyses that project
their differential effects for women and for nationality and eth-
nic groups. This does not feel like a radical proposal, but no
state or country does it. Some disadvantaged ethnic or na-
tional minority in every country is heavily overrepresented
among crime victims, offenders, and prisoners (Tonry 1997).
Generally, this is widely seen as regrettable and by some it
is seen as the result primarily of either intentional or “institu-
tional” racism.
The causes of racial and ethnic disparities invariably in-
clude policies that in principle apply to everybody but in
practice disproportionately affect particular groups. The fed-
eral 100-to-1 rule that punishes sale of five grams of crack
cocaine as harshly as the sale of 500 grams of powder cocaine
is facially neutral. Because, however, poor blacks typically
sell small amounts of crack and whites typically sell powder
in varying quantities, in practice the distinction produces
longer sentences for black than for white drug dealers and
worsens racial disparities in imprisonment. The disparate ef-
fects were foreseeable in 1986 when the 100-to-1 rule was
enacted.
If disparate impact projections and analyses were an obliga-
tory part of the legislative process, policy makers would have
had explicitly to weigh wanted crime prevention effects
against presumably unwanted but foreseeable disparate im-

pacts on white and black defendants. This would have had
the salutary effect of forcing realistic assessments of what
preventive effects were likely and how confidently they
could be expected to occur. Perhaps after such deliberations,
Congress might still have enacted the 100-to-1 rule, but then
the decision would have been explicitly made that the likely
gains outweighed the undesirable side effect of punishing
blacks more severely. Or, the likely drug-abuse reduction
gains might have been deemed so speculative that they did
not justify the disparate effects at all or that the differential
should have been less. Or even, possibly, the decision might
have been made that, given a choice between higher drug
abuse or crime rates and worsened racial disparities, the latter
was the greater evil and the rule should not be adopted. In
any case, such analyses would force explicit choices to be
made between crime control policy goals, whether expressive
or substantive, and race relations goals.
Policy makers shy away from directly addressing racial
issues in the criminal justice system. One reason for this is
a combination of bad motive and hypocrisy. Some people
argue that many features of the federal war on drugs that
disproportionately affect blacks are best explained this way
(Tonry 1995; Wacquant 2001).
Two other reasons for avoiding disparate impact analyses
are that the issues they raise are difficult and sensitive. The
difficulty is illustrated by the U.S. Supreme Court’s decision
in McCleskey v. Kemp, 481 U.S. 279 (1997), in which the Court
accepted the validity of empirical evidence that, all else equal,
capital punishment in Georgia was much more often ordered
for blacks who killed whites than for black-on-black, white-
on-white, or white-on-black killings. The issue was devil-
ishly difficult. The Court had to decide whether to allow
McCleskey to be executed in the face of evidence that people
like him are more likely to be executed because of their and
their victim’s race, or to accept the inferences raised by the
statistical evidence and effectively end capital punishment
in Georgia (and by implication, the rest of the country). In
the face of evidence that race matters, the Court nonetheless

held that the statistical evidence was immaterial and that
only proof of racial bias in the individual case was relevant.
McCleskey’s lawyers could not show this—few biased judges
or lawyers openly admit that they are acting discriminatorily,
and much discrimination in any event is unconscious and
based on insensitivity and stereotypes. McCleskey in due
course was killed by the State of Georgia.
In explaining the Court’s decision, an obviously troubled
Justice Lewis Powell pointed out that the logic of Mc-
Cleskey’s claim, that statistical evidence of inexplicable racial
disparities should raise a presumption of bias that the state
must refute, might apply throughout the criminal justice sys-
tem. The Court could have avoided that result by asserting
that “death is different,” but in principle Powell was right.
Some would argue that statistical evidence of unaccountable
disparities should raise presumptions of bias, and if that bene-
fits minority defendants, so be it.
Reasonable people differ about the question whether courts
should create such presumptions, but no one doubts that
legislatures can (and in some contexts such as employment
opportunity and college athletics, have). A Martian might
find it a bit ironic that an employer who hires a lower percent-
age of blacks than reside in the local community, according
to statistical evidence, is legally presumed to have acted dis-
criminatorily and must justify its hiring practices or be ad-
judged guilty of employment discrimination, while equiva-
lent but much stronger evidence about capital punishment
raises no legal presumptions at all. The seeming inference is
that the law cares more about peoples’ jobs than their lives.
That was a difficult issue. Here is a sensitive one: Should the
race of offenders be taken into account in deciding whether to
increase penalties for homicide? The analysis would apply
also to three-strikes laws and to mandatory minimum and
truth-in-sentencing laws covering violent crimes. The thresh-
old problem is that blacks are arrested for violent crimes much
more often relative to population than are whites. Thirteen
percent of the U.S. population in 2000 was black. Of people
arrested in that year for homicide, rape, and robbery, 27.9
percent, 34.1 percent, and 53.9 percent, respectively, were
black.
Assume for purposes of argument that arrest percentages
are an accurate indicator of racial patterns of serious violent
crime (the best evidence is that for serious violent crime,
arrest proportions are reasonably valid indicators). A pro-
posed law increasing the minimum prison sentence for ho-
micide from ten to thirty years would, because of the racial
offending pattern, disproportionately affect blacks. Over time,
racial disparities in U.S. prisons would worsen. Is that a good
thing or a bad thing?
Preventing crime is a good thing, but so is preventing racial
disparities. If we knew that the increased lengths of sentence
would through deterrent and incapacitative effects decrease
the homicide rate by x percent or save the lives of y plus or
minus ten people, the trade-off could be discussed. Does an
estimated crime reduction effect of a particular amount justify
an increase from, say, 46 to 50 percent of the prison popula-
tion who are black? What if no crime reduction effect could
credibly be estimated? Could any increase in racial dispropor-
tions be justified, or any increase in penalties that would
disproportionately affect blacks? Whatever the crime reduc-
tion estimate, does it matter that violent crimes, especially
homicide and rape, are heavily intraracial, and thus that vio-
lent black offenders mostly have black victims? Punishing
violent offenders more harshly will, if that has crime preven-
tion effects, reduce victimization rates and disproportion-
ately benefit potential black victims.
Comparable analyses of projected effects of alternative pub-
lic policy choices are commonplace (Cook and Ludwig 2001).
Planners explicitly weigh estimates of lost life against cost
in building roads or designing aircraft. No action can be made
perfectly safe and cost inevitably is a consideration. If all
automobiles were design-limited to ten miles an hour and
built with heavy steel bodies, we would have many fewer
traffic fatalities and injuries. Decisions to make cars faster,
lighter, and cheaper produce higher projected rates of injury
and death and higher medical-care costs. The trade-offs are
explicitly discussed. Similar trade-offs should be discussed,
and explicit choices made, in relation to race and sentencing.
Ninth, conduct disparity audits of all criminal justice system
practices to establish whether and to what extent they exacerbate
or ameliorate disparities in outcomes affecting women and mem-
bers of minority groups. This is a variation on the last point
and the argument is much the same. Many facially neutral
practices have racially disparate effects. Here are two exam-
ples.
First, as mentioned earlier, guidelines for decisions about
pretrial release often give weight to social factors such as
the defendant’s educational background, vocational skills,
or family status. These are not nonsensical factors. The key
issue is whether the defendant will appear for trial or run
away, and indicators of social stability or connections are
plausible criteria for predicting the defendant will not disap-
pear. Unfortunately, relatively more blacks and Hispanics
than whites are poor and socially disadvantaged. They will
be less likely to benefit from the social factors and more likely
to be held in jail pending trial. In addition, a large body of
research shows that defendants held in jail before trial, all
else equal, are more likely to receive prison sentences and
to receive longer prison sentences (even after pretrial time
in jail is factored in) than offenders who were released before
trial (Tonry 1995). This means that pretrial detention prac-
tices create another disparate impact for minority defendants.
Happily, these disparate impacts can easily be reduced, once
the problem is recognized. Other factors related solely to
current or past criminality provide as good predictions of
appearance at trial.
The second example is harder. In most countries, defen-
dants from some disadvantaged minority groups are less
likely than majority group members to plead guilty, and
when they do plead guilty to do it later in the process (Tonry
1997). Most systems provide sentence reductions, openly or
otherwise, to defendants who plead guilty. In England, to
give a concrete example, the so-called progressive-loss-of-
mitigation doctrine explicitly provides a one-third reduction

of sentence to defendants who plead guilty at the outset,
with successively lesser reductions for later pleas (hence,
“progressive loss”). Afro-Caribbeans lose on both counts.
They plead guilty less often than whites, and typically later
(Hood 1992). Disparate impact audits would force policy mak-
ers explicitly to decide and declare whether the discount’s
presumed cost-reduction goals outweigh the disparities that
foreseeably are caused. This is a particularly difficult choice
to make, since the noncooperation surely results in part from
minority defendants’ sometimes justified alienation from a
criminal justice system in which they believe they are un-
fairly treated, especially by the police.
Implementing the last two proposals will not be easy, but
they are necessary if American jurisdictions are to reduce
the overrepresentation of minority group members in prisons
and jails. The necessary research and statistical analyses will
not be difficult, but the policy decisions often will be. The
illustration of bail-or-jail criteria given above is easy. Once
the problem is recognized, criteria that do not treat groups of
offenders differently, or that reduce the extent of differential
treatment, can be substituted. More difficult are instances
like the 100-to-1 rule where the goal (try to reduce the direct
and indirect effects of drug trafficking) seems desirable but
the effect (treat black defendants much more harshly than
whites) does not. Among the choices are to accept the un-
wanted racial disparity, to abandon the drug policy goal, or
to compromise. My guess is that compromise is what would
have happened had the choices been starkly presented. Black
crack defendants would still have been sentenced more
harshly than white powder defendants, but the differential,
and the resulting racial disparities, would have been less.
That would have been a small victory, but a victory all the
same.
It could be that forcing policy makers openly to discuss
difficult and sensitive racial issues relating to the criminal
justice system would change the outcomes of policy pro-
cesses. Bad or bigoted motives might be imputed to people
who vigorously promote policies that will treat blacks or

Hispanics or members of other minorities more harshly than
whites or that will worsen disparities. No one wants to be
called a racist or a bigot and that risk might make policy
makers less likely to promote such policies. In lawyers’ lan-
guage, open discussion of racial dimensions of policy might
have a chilling effect. Some good-faith proposals might not
be offered from fear that motives will be misinterpreted or
that epithets will unfairly be brandished. In an ideal world,
of course, that would be a pity. Policies should be based on
evidence and values, openly discussed and critically exam-
ined, and decisions should result from reconciliation of differ-
ent interests and perspectives. In our somewhat less than
ideal world, the trade-offs are not so clear. It would not be
a bad thing if chilling effects associated with open discussion
of racial issues in sentencing policy prevented promotion or
adoption of policies that would worsen racial disparities.
The proposals set out in this chapter would, if taken seri-
ously, move American policies and practices closer to the
mainstream of contemporary Western values and practices.
They will not be easy to adopt or implement, and they might
not work as well as I hope they might. I suspect though,
that if we try, we will stand higher in the esteem of our
descendants, and of our current and future friends in other
lands, than if we continue our current practices or merely
tinker at the edges.

References
Albrecht, Hans-Jörg. 2001. “Post-Adjudication Dispositions in

Comparative Perspective.” In Sentencing and Sanctions in
Western Countries, ed. Michael Tonry and Richard S. Frase.
New York: Oxford University Press.
ALI, see American Law Institute
Alschuler, Albert W. 1978. “Sentencing Reform and Prosecuto-
rial Power.” University of Pennsylvania Law Review 126:
550–77.
American Law Institute. 1954. Model Penal Code: Tentative Draft
No. 2. Philadelphia: American Law Institute.
. 1956. Model Penal Code: Tentative Draft No. 5. Philadel-
phia: American Law Institute.
. 1960. Model Penal Code: Tentative Draft No. 12. Phila-
delphia: American Law Institute.
. 1962. Model Penal Code (Proposed Official Draft). Phila-
delphia: American Law Institute.
. 1998. The American Law Institute Seventy-Fifth Anni-
versary 1923–1998. Philadelphia: American Law Institute.
American Law Institute Proceedings. 1954. 31st Annual Meeting.
The American Law Institute. Proceedings. Philadelphia: Amer-
ican Law Institute.
. 1956. 33rd Annual Meeting. The American Law Institute.
Proceedings. Philadelphia: American Law Institute.
. 1960. 37th Annual Meeting. The American Law Institute.
229
. 1961. 38th Annual Meeting. The American Law Institute.
Andenaes, Johannes. 1974. Punishment and Deterrence. Ann Arbor:
University of Michigan Press.
Applebome, Peter. 1996. Dixie Rising: How the South is Shaping
American Values, Politics and Culture. New York: Times Books.
Barclay, Gordon, and Cynthia Tavares. 2002. International Com-
parisons of Criminal Justice Statistics 2000. London: Home
Office, Research, Development, and Statistics Directorate.
Barnes, Jonathan, ed. 1984. The Complete Works of Aristotle:
The Revised Oxford Edition. 2 vols. Princeton, N.J.: Princeton
University Press.
Bauman, Zygmunt. 1991. Modernity and Ambivalence. Cam-
bridge: Polity Press.
Beck, Allen J. 1989. Recidivism of Prisoners Released in 1983.
Washington, D.C.: U.S. Department of Justice, Bureau of Jus-
tice Statistics.
Beck, Ulrich. 1992. Risk Society: Towards a New Modernity.
Trans. Mark Ritter. London: Sage.
. 1996. “Risk Society and the Provident State.” In Risk,
Environment and Modernity: Towards a New Ecology, ed. Scott
Lash, Bronislaw Szerszynski, and Brian Wynne. London: Sage.
, and Elisabeth Beck-Gershheim. 1995. The Normal Chaos
of Love. Cambridge: Polity Press.
Becker, Gary S. 1968. “Crime and Punishment: An Economic
Analysis.” Journal of Political Economy 76: 169–217.
Beckett, Katherine. 1997. Making Crime Pay: Law and Order in
Contemporary American Politics. New York: Oxford Univer-
sity Press.
Begasse, Jen Kiko. 1997. “Oregonians Support Alternatives for
Nonviolent Offenders.” In Sentencing Reform in Overcrowded
Times, ed. Michael Tonry and Kathleen Hatlestad. New York:
Oxford University Press.
Ben-Yehuda, Nachman. 1990. The Politics and Morality of Devi-
ance: Moral Panics, Drug Abuse, Deviant Science, and Reversed
Stigmatization. Albany, N.Y.: State University of New York
Press.
Bennett, William J., John J. DiIulio, and John P. Walters. 1996.
230 REFERENCES
Body Count: Moral Poverty—and How to Win America’s War
against Crime and Drugs. New York: Simon and Schuster.
Bentham, Jeremy. 1789 [1948]. Introduction to the Principles of
Morals and Legislation, ed. Wilfred Harrison. Oxford: Oxford
University Press.
. 1843. The Works of Jeremy Bentham, vol. 4, ed. John
Bowring. London: Simpkin, Marshall.
Bettelheim, Bruno. 1977. The Uses of Enchantment: The Meaning
and Importance of Fairy Tales. London: Thames and Hudson.
Bishop, Donna M. 2000. “Juvenile Offenders in the Adult Crimi-
nal Justice System.” In Crime and Justice: A Review of Re-
search, vol. 27, ed. Michael Tonry. Chicago: University of
Chicago Press.
Blackstone, William. 1766–69. Commentaries on the Laws of En-
gland. 2d ed. Oxford: Clarendon Press.
Blumstein, Alfred. 1993. “Making Rationality Relevant—The
American Society of Criminology 1992 Presidential Address.”
Criminology 31: 1–16.
, and Allen J. Beck. 1999. “Population Growth in U.S.
Prisons, 1980–1996.” In Prisons, ed. Michael Tonry and Joan
Petersilia. Vol. 26 of Crime and Justice: A Review of Research,
ed. Michael Tonry. Chicago: University of Chicago Press.
, and Jacqueline Cohen. 1973. “A Theory of Stability of
Punishment.” Journal of Criminal Law and Criminology 64:
198–201.
, Jacqueline Cohen, Jeffrey Roth, and Christy Visher,
eds. 1986. Criminal Careers and “Career Criminals.” Washing-
ton, D.C.: National Academy Press.
Boerner, David, and Roxanne Lieb. 2001. “Sentencing Reform
in the Other Washington.” In Crime and Justice: A Review
of Research, vol. 28, ed. Michael Tonry. Chicago: University
of Chicago Press.
Boswell, John. 1980. Christianity, Social Tolerance, and Homo-
sexuality. Chicago: University of Chicago Press.
Bottomley, Keith. 1990. “Parole in Transition: A Comparative
Study of Origins, Developments, and Prospects for the
1990s.” In Crime and Justice: A Review of Research, vol. 12,
REFERENCES 231
Bottoms, Anthony E. 1995. “The Philosophy and Politics of Pun-
ishment and Sentencing.” In The Politics of Sentencing Re-
form, ed. C. M. V. Clarkson and R. Morgan. Oxford: Claren-
don.
, and James Dignan. 2004. “Youth Justice in England
and Scotland.” In Youth Crime and Youth Justice: Comparative
and Cross-national Perspectives, ed. Michael Tonry and An-
thony Doob. Chicago: University of Chicago Press.
Boutellier, Hans J. 2000. Crime and Morality: The Significance
of Criminal Justice in Post-Modern Culture. London: Kluwer
Academic.
Bowling, Benjamin. 1999. “The Rise and Fall of New York Mur-
der: Zero Tolerance or Crack’s Decline?” British Journal of
Criminology 39: 531–54.
Braithwaite, John. 1989. Crime, Shame and Reintegration. Cam-
bridge: Cambridge University Press.
. 1999. “Restorative Justice: Assessing Optimistic and
Pessimistic Accounts.” In Crime and Justice: A Review of Re-
search, vol. 25, ed. Michael Tonry. Chicago: University of
Chicago Press.
. 2001. Restorative Justice and Responsive Regulation.
Bureau of Justice Statistics. 1998. State Court Sentencing of Con-
victed Felons, 1994. Washington, D.C.: U.S. Department of
Justice, Bureau of Justice Statistics.
. 2000. Correctional Populations in the United States, 1997.
Washington, D.C.: U.S. Department of Justice, Bureau of Jus-
tice Statistics.
. 2001. Prisoners in 2000. Washington, D.C.: U.S. Depart-
ment of Justice, Bureau of Justice Statistics.
. 2003. Prisoners in 2002. Washington, D.C.: U.S. Bureau
of Justice Statistics.
Burgess, Ernest W. 1928. “Factors Determining Successes or
Failure on Parole.” Part 4 of The Workings of the Indetermi-
nate-Sentence Law and the Parole System in Illinois, by An-
drew W. Bruce, Ernest W. Burgess, and Albert J. Harno.
Springfield: Illinois Board of Parole.
Camus, Albert. 1960 (c. 1959). Reflections on the Guillotine: An
232 REFERENCES
Essay on Capital Punishment. Trans. Richard Howard. Michi-
gan City, Ind.: Fridtjof-Karla Publications.
Caplow, Theodore, and Jonathan Simon. 1999. “Understanding
Prison Policy and Population Trends.” In Prisons, ed. Michael
Tonry and Joan Petersilia. Vol. 26 of Crime and Justice: A
Review of Research, ed. Michael Tonry. Chicago: University
of Chicago Press.
Casper, Jonathan D., David Brereton, and David Neal. 1983. The
Implementation of the California Determinate Sentencing Law.
Washington, D.C.: U.S. Department of Justice.
Christie, Nils. 2000. Crime Control as Industry: Towards Gulags
Western Style. 3d ed. London: Routledge.
Clarke, Stevens H. 1987. Felony Sentencing in North Carolina
1976–1986: Effects of Presumptive Sentencing Legislation.
Chapel Hill: Institute of Government, University of North
Carolina at Chapel Hill.
Clear, Todd R., and Eric Cadora. 2001. “Risk and Correctional
Practice.” In Crime, Risk, and Justice: The Politics of Crime
Control in Liberal Democracies, ed. Kevin Stenson and Robin
R. Sullivan. Derby, U.K.: Willan Publishing.
, Dina R. Rose, and Judith A. Ryder. 2001. “Incarceration
and the Community: The Problem of Removing and Return-
ing Prisoners.” Crime and Delinquency 47: 335–51.
Cohen, Mark A. 1988. “Pain, Suffering, and jury Awards: A
Study of the Cost of Crime to Victims.” Law and Society Re-
view 2(3): 537–55.
Cohen, Stanley. 1972. Folk Devils and Moral Panics. New York:
St. Martin’s Press.
Cook, Philip, and Jens Ludwig. 2001. Gun Violence: The Real
Cost. New York: Oxford University Press.
Cullen, Frank, Bonnie S. Fisher, and Brandon K. Applegate.
2000. “Public Opinion about Punishment and Corrections.”
In Crime and Justice: A Review of Research, vol. 27, ed. Mi-
chael Tonry. Chicago: University of Chicago Press.
Death Penalty Information Center. 2002. Available online at
www.deathpenaltyinfo.org/index.html.
DiIulio, John J. 1990. “Crime and Punishment in Wisconsin.”
Wisconsin Policy Research Institute Report 3(7): 1–56.
REFERENCES 233
Doble, John. 1997. “Survey Shows Alabamians Support Alterna-
tives.” In Sentencing Reform in Overcrowded Times, ed. Mi-
chael Tonry and Kathleen Hatlestad. New York: Oxford Uni-
versity Press.
, and Stephen Immerwahr. 1997. “Delawareans Favor
Prison Alternatives.” In Sentencing Reform in Overcrowded
Times, ed. Michael Tonry and Kathleen Hatlestad. New York:
Doob, Anthony N., and Voula Marinos. 1995. “Reconceptualiz-
ing Punishment: Understanding the Limitations on the Use of
Intermediate Punishments.” University of Chicago Law School
Roundtable 2: 413–33.
Douglas, Mary. 1985. Risk Acceptability According to the Social
Sciences. New York: Russell Sage Foundation.
. 1992. Risk and Blame: Essays in Cultural Theory. Lon-
don: Routledge.
Duff, R. A. 2000. Punishment, Communication, and Community.
Durkheim, Émile. 1933. The Division of Labour in Society. Trans.
George Simpson. (Originally pub. 1893.) New York: Free Press.
Dworkin, Ronald. 1977. Taking Rights Seriously. London: Duck-
worth.
. 1986. Law’s Empire. Cambridge, Mass.: Belknap Press.
Eck, John E., and Edward R. Maguire. 2000. “Have Changes
in Policing Reduced Violent Crime? An Assessment of the
Evidence.” In The Crime Drop in America, ed. Alfred Blumstein
and Joel Wallman. Cambridge: Cambridge University Press.
Edsall, Thomas, and Mary Edsall. 1991. Chain Reaction: The
Impact of Race, Rights, and Taxes on American Politics. New
York: Norton.
Eisner, Manuel. 2001. “Modernization, Self-Control and Lethal
Violence: The Long-Term Dynamics of European Homicide
Rates in Theoretical Perspective.” British Journal of Criminol-
ogy 41: 618–38.
. 2003. “Secular Trends of Violent Crime: Evidence and
Theoretical Interpretations.” In Crime and Justice: A Review
of Chicago Press.
234 REFERENCES
Elias, Norbert. 1978. The History of Manners: The Civilising Pro-
cess, vol. 1. Oxford: Basil Blackwell. (Originally published
1939, Basel: Hans Zum Falken.)
. 1982. State Formation and Civilization: The Civilizing
Process, vol. 2. Oxford: Basil Blackwell. (Originally published
1939, Basel: Hans Zum Falken.)
Ellis, Joseph J. 2002. Founding Brothers: The Revolutionary Gen-
eration. New York: Vintage.
Elson, Alex. 1998. “The Case for an In-depth Study of the Ameri-
can Law Institute.” Law and Social Enquiry 23: 625–40.
Erikson, Kai T. 1966. Wayward Puritans: A Study in the Sociology
of Deviance. New York: John Wiley.
Evans, Malcolm, and Rod Morgan. 1998. Preventing Torture: A
Study of the European Convention for the Prevention of Torture
and Inhuman or Degrading Treatment or Punishment. New
York: Oxford University Press.
Fagan, Jeffrey, Franklin Zimring, and June Kim. 1998. “Declin-
ing Homicide in New York City: A Tale of Two Trends.”
Journal of Criminal Law and Criminology 88: 1277–1323.
Farkas, Steve. 1997. “Pennsylvanians Prefer Alternatives to
Prison.” In Sentencing Reform in Overcrowded Times, ed. Mi-
versity Press.
Farrington, David P. 1986. “Age and Crime.” In Crime and Jus-
tice: An Annual Review of Research, vol. 7, ed. Michael Tonry
and Norval Morris. Chicago: University of Chicago Press.
Federal Courts Study Committee. 1990. Report. Washington,
D.C.: Administrative Office of the U.S. Courts.
Feeley, Malcolm M., and Jonathan Simon. 1992. “The New Pe-
nology: Notes on the Emerging Strategy of Corrections and
Its Implications.” Criminology 30: 449–74.
Feinberg, Joel. 1970. Doing & Deserving: Essays in the The-
ory of Responsibility. Princeton, N.J.: Princeton University
Press.
Feld, Barry C. 1999. Bad Kids: Race and the Transformation of
the Juvenile Court. New York: Oxford University Press.
Florida Legislative Office of Program Policy Analysis and Gov-
ernment Accountability. 1995. Policy Review of Reincarcera-
REFERENCES 235
tion in Florida’s Prisons Administered by the Department of
Corrections. Tallahassee: Florida Legislature.
Fogelson, Robert M. 1977. Big-City Police. Cambridge, Mass.:
Harvard University Press.
Foucault, Michel. 1977. Discipline and Punish: The Birth of the
Prison. Trans. Robert Hurley. New York: Pantheon.
Frankel, Marvin E. 1972. Criminal Sentences: Law without Order.
New York: Hill and Wang.
Frase, Richard, and Debra Dailey. 2004. “Sentencing Policy in
Minnesota, 1975–2000.” In Crime and Justice: A Review of
Research, vol. 3, ed. Michael Tonry. Chicago: University of
Chicago Press.
Freiberg, Arie. 2001. “Three Strikes and You’re Out—It’s Not
Cricket: Colonization and Resistance in Australian Sentenc-
ing.” In Sentencing and Sanctions in Western Countries, ed.
Michael Tonry and Richard S. Frase. New York: Oxford Uni-
versity Press.
Friedman, Lawrence. 1993. Crime and Punishment in American
History. New York: Basic Books.
Gannon, Marie. 2001. Crime Comparisons between Canada and
the United States. Ottawa: Statistics Canada.
Garland, David. 1990. Punishment and Modern Society: A Study
in Social Theory. Chicago: University of Chicago Press.
. 1991. “Sociological Perspectives on Punishment.” In
Crime and Justice: A Review of Research, vol. 14, ed. Michael
Tonry. Chicago: University of Chicago Press.
. 1996. “The Limits of the Sovereign State: Strategies of
Crime Control in Contemporary Society.” British Journal of
Criminology 36(4): 445–71.
. 2001a. The Culture of Control: Crime and Social Order in
Contemporary Society. Chicago: University of Chicago Press.
, ed. 2001b. Mass Imprisonment in the United States: So-
cial Causes and Consequences. London: Sage.
Gest, Ted. 2001. Crime & Politics: Big Government’s Erratic Cam-
paign for Law and Order. New York: Oxford University Press.
Giddens, Anthony. 1990. The Consequences of Modernity. Cam-
bridge: Polity Press.
236 REFERENCES
. 1991. Modernity and Self-Identity. Cambridge: Polity
Press.
. 1998. “Risk Society: The Context of British Politics.”
In The Politics of Risk Society, ed. Jane Franklin. Malden,
Mass.: Polity Press.
Gilmore, Grant. 1974. The Death of Contract. Columbus: Ohio
State University Press.
Goode, Erich, and Nachman Ben-Yehuda. 1994. Moral Panics:
The Social Construction of Deviance. Cambridge, Mass.: Black-
well.
Goodrich, Herbert F., and Paul A. Wolkin. 1961. The Story of
the American Law Institute 1923–1961. St. Paul, Minn.: Ameri-
can Law Institute.
Gottfredson, Don M., Leslie T. Wilkins, and Peter B. Hoffman.
1978. Guidelines for Parole and Sentencing. Lexington, Mass.:
Lexington Books.
Grayling, Anthony C. 2001. “The Last Word on Anger.” The
Guardian, December 8, 2001.
Greenberg, David F. 1988. The Construction of Homosexuality.
Chicago: University of Chicago Press.
Greene, Judith A. 1999. “Zero Tolerance: A Case Study of Police
Policies and Practices in New York City.” Crime and Delin-
quency 45: 171–87.
Greenwood, Peter W., and Allan Abrahamse. 1982. Selective
Incapacitation. Prepared for the National Institute of Justice.
Santa Monica, Calif.: Rand Corp.
Gurr, Ted Robert. 1981. “Historical Trends in Violent Crimes:
A Critical Review of the Evidence.” In Crime and Justice: An
Annual Review of Research, vol. 3, ed. Michael Tonry and
Norval Morris. Chicago: University of Chicago Press.
. 1989. “Historical Trends of Violent Crime: England,
Western Europe, and the United States.” In Violence in
America: The History of Crime, vol. 1, ed. Ted Robert Gurr.
Newbury Park, Calif.: Sage.
Gusfield, Joseph R. 1963. Symbolic Crusade: Status Politics and
the American Temperance Movement. Urbana: University of
Illinois Press.
REFERENCES 237
Hagan, John, and Ronit Dinovitzer. 1999. “Collateral Conse-
quences of Imprisonment for Children, Communities, and
Prisoners.” In Prisons, ed. Michael Tonry and Joan Petersilia.
Vol. 26 of Crime and Justice: A Review of Research, ed. Mi-
Hall, Stuart, Chas Critcher, Tony Jefferson, John Clarke, and
Brian Roberts. 1978. Policing the Crisis: Mugging, the State,
and Law and Order. London: Macmillan.
Hallsworth, Simon. 2000. “Rethinking the Punitive Turn: Econ-
omies of Excess and the Criminology of the Other.” Punish-
ment and Society 2(2): 145–60.
Hampton, Jean. 1984. “The Moral Education Theory of Punish-
ment.” Philosophy and Public Affairs 13(3): 208–38.
Harcourt, Bernard E. 2001. Illusion of Order: The False Promise
of Broken Windows Policing. Cambridge, Mass.: Harvard Uni-
versity Press.
Hart, H. L. A. 1968. Punishment and Responsibility. Oxford: Ox-
ford University Press.
Heaney, Gerald W. 1991. “The Reality of Guidelines Sentencing:
No End to Disparity.” American Criminal Law Review 28:
161–233.
Hood, Roger. 1992. Race and Sentencing. Oxford: Clarendon.
. 2002. “Criminology and Penal Policy: The Vital Role of
Empirical Research.” In Ideology, Crime and Criminal Justice:
A Symposium in Honour of Sir Leon Radzinowicz, ed. Anthony
Bottoms and Michael Tonry. Cullompton, Devon: Willan.
Hope, Tim, and Richard Sparks, eds. 2000. Crime, Risk and
Insecurity: Law and Order in Everyday Life and Political Dis-
course. London: Routledge.
Hough, Mike, and Julian V. Roberts. 1997. Attitudes to Punish-
ment: Findings from the British Crime Survey. Home Office
Research Study no. 179. London: Home Office.
Howell, James C., and J. David Hawkins. 1998. “Prevention of
Youth Violence.” In Youth Violence, ed. Michael Tonry and
Mark H. Moore. Vol. 24 of Crime and Justice: A Review of
Research, ed. Michael Tonry. Chicago: University of Chicago
Press.
Hull, Natalie E. H. 1990. “Restatement and Reform: A New Per-
238 REFERENCES
spective on the Origins of the American Law Institute.” Law
and History Review 8: 55–96.
Janson, Carl-Gunnar. 2004. “Youth Justice in Sweden.” In Youth
Crime and Youth Justice: Comparative and Cross-national Per-
spectives, ed. Michael Tonry and Anthony Doob. Chicago:
University of Chicago Press.
Jareborg, Nils. 2001. “Sentencing Law, Patterns, and Policies in
Sweden.” In Penal Reform in Overcrowded Times, ed. Michael
Tonry. New York: Oxford University Press.
Jefferson, Tony, and Wendy Hollway. 2000. “The Role of Anxi-
ety in Fear of Crime.” In Crime, Risk and Insecurity: Law and
Order in Everyday Life and Political Discourse, ed. Tim Hope
and Richard Sparks. London: Routledge.
Jenkins, Philip. 1998. Moral Panic: Changing Concepts of the
Child Molester in Modern America. New Haven, Conn.: Yale
University Press.
Johnson, Paul. 1976. A History of Christianity. New York: Athe-
neum.
Kagan, David. 1989. “How America Lost Its First Drug War.”
Insight 8–17 (Nov. 20, 1989).
Kahan, Dan M. 1997. “Some Realism about Retroactive Criminal
Lawmaking.” Roger Williams University Law Review 3(1):
95–117.
. 1998a. “The Anatomy of Disgust in Criminal Law.”
Michigan Law Review 96: 1621–57.
. 1998b. “Punishment Incommensurability.” Buffalo
Criminal Law Review 1: 691–708.
, and Tracey L. Meares. 1998. “The Wages of Antiquated
Criminal Procedure.” University of Chicago Law Forum 1998:
197–214.
. 1998. “The Coming Crisis of Criminal Procedure.”
Georgetown Law Journal 86: 1153–84.
Kennedy, Joseph E. 2000. “Monstrous Offenders and the Search
for Solidarity through Modern Punishment.” Hastings Law
Journal 51: 829–908.
Kennedy, Randall. 1997. Race, Crime, and the Law. New York:
Pantheon Books.
Kensey, Annie, and Pierre Tournier. 2001. “French Prison Num-
REFERENCES 239
bers Stable since 1988, but Populations Changing.” In Penal
Reform in Overcrowded Times, ed. Michael Tonry. New York:
Kesteren, John van, Pat Mayhew, and Paul Nieuwbeerta. 2000.
Criminal Victimization in Seventeen Industrialised Countries:
Key Findings from the 2000 International Crime Victims Survey.
The Hague: WODC.
King, Ryan S., and Marc Mauer. 2001. Aging behind Bars: Three
Strikes Seven Years Later. Washington, D.C.: The Sentencing
Project.
Kleiman, Mark A. R., and David Cavanagh. 1990. “A Cost-Bene-
fit Analysis of Prison Cell Construction and Alternative Sanc-
tions.” Unpublished manuscript. Cambridge, Mass.: Kennedy
School of Government, Guggenheim Program in Criminal Jus-
tice Policy and Management.
Knapp, Kay. 1984. The Impact of the Minnesota Sentencing Guide-
lines: Three-Year Evaluation. St. Paul, Minn.: Minnesota Sen-
tencing Guidelines Commission.
Kommer, Max. 1994. “Punitiveness in Europe.” European Jour-
nal of Criminal Policy and Research 2(1): 29–43.
Kuhn, André. 1998. “Sanctions and Their Severity.” In Crime
and Criminal Justice Systems in Europe and North America
1990–1994, ed. K. Kangasunta, M. Joutsen, and N. Ollus.
Helsinki: European Institute for Crime Prevention and Con-
trol (HEUNI).
. 2001. “Incarceration Rates across the World.” In Penal
Reform in Overcrowded Times, ed. Michael Tonry. New York:
Kuhn, Thomas. 1996. The Structure of Scientific Revolutions. 3d
ed. Chicago: University of Chicago Press.
Kurki, Leena. 2001. “International Standards for Sentencing and
Punishment.” In Sentencing and Sanctions in Western Coun-
tries, ed. Michael Tonry and Richard S. Frase. New York:
Kyvsgaard, Britta. 2004. “Youth Justice in Denmark.” In Youth
240 REFERENCES
Lane, Roger. 1980. “Urban Police and Crime in Nineteenth-Cen-
tury America.” In Crime and Justice: An Annual Review of
Research, vol. 2, ed. Norval Morris and Michael Tonry. Chi-
cago: University of Chicago Press.
. 1992. “Urban Police and Crime in Nineteenth-Century
America.” In Modern Policing, ed. Michael Tonry and Norval
Morris. Vol. 15 of Crime and Justice: A Review of Research,
. 1999. “Murder in America: A Historian’s Perspective.”
In Crime and Justice: A Review of Research, vol. 25, ed. Mi-
Lappi-Seppälä, Tapio. 2001. “Sentencing and Punishment in
Finland: The Decline of the Repressive Ideal.” In Sentencing
and Sanctions in Western Countries, edited by Michael Tonry
and Richard S. Frase. New York: Oxford University Press.
Lieb, Roxanne, Vernon Quinsey, and Lucy Berliner. 1998. “Sex-
ual Predators and Social Policy.” In Crime and Justice: A
Review of Research, vol. 23, ed. Michael Tonry. Chicago: Uni-
versity of Chicago Press.
Lindblom, Charles E., and D. K. Cohen. 1979. Usable Knowledge.
London: Yale University Press.
Lipson, Albert J., and Mark A. Peterson. 1980. California Justice
under Determinate Sentencing: A Review and Agenda for Re-
search. Santa Monica, Calif.: Rand Corp.
Lupton, Deborah. 1999. Risk. London: Routledge.
Males, M., and D. Macallair. 1999. “Striking Out: The Failure of
California’s ‘Three-Strikes-and-You’re-Out Law.’” Stanford
Law and Policy Review 11: 65–74.
Mauer, Marc. 1999. The Race to Incarcerate. New York: New
Press.
, and Meda Chesney-Lind, eds. 2002. Invisible Punish-
ment: The Collateral Consequences of Mass Imprisonment. New
York: New Press.
Mayhew, Pat, and Jan van Dijk. 1997. Criminal Victimisation
in Eleven Industrialized Countries: Key Findings from the 1996
International Crime Victims Survey. The Hague: Dutch Minis-
try of Justice.
McIvor, Gill. 1995. “CSOs Succeed in Scotland.” In Intermediate
REFERENCES 241
Sanctions in Overcrowded Times, ed. Michael Tonry and Kate
Hamilton. Boston: Northeastern University Press.
Meares, Tracey L., and Dan M. Kahan. 1998. “Law and (Norms
of ) Order in the Inner City.” Law and Society Review 32:
805–38.
Messinger, Sheldon, and Phillip Johnson. 1978. “California’s
Determinate Sentence Laws.” In Determinate Sentencing: Re-
form or Regression. Washington, D.C.: U.S. Government Print-
ing Office.
Minnesota Legislative Auditor. 1997. Recidivism of Adult Felons.
St. Paul: State of Minnesota, Office of the Legislative Auditor.
Moffitt, Terrie E. 1993. “Adolescence-Limited and Life-Course-
Persistent Antisocial Behavior: A Developmental Taxon-
omy.” Psychological Review 100: 674–701.
Moore, Mark H. 1995. “Learning by Doing: Linking Knowledge
to Policy in the Development of Community Policing and
Violence Prevention in the United States.” In Integrating
Crime Prevention Strategies: Propensity and Opportunity, ed.
P. O. Wikström, R. V. Clarke, and J. McCord. Stockholm:
National Council for Crime Prevention.
Morgan, Rod. 2001. “International Controls on Sentencing and
Punishment.” In Sentencing and Sanctions in Western Coun-
tries, ed. Michael Tonry and Richard S. Frase. New York:
Morris, Allison. 2004. “Youth Justice in New Zealand.” In Youth
Morris, Norval. 1974. The Future of Imprisonment. Chicago: Uni-
versity of Chicago Press.
Murphy, Jeffrie G. 1985. “Retributivism, Moral Education, and
the Liberal State.” Criminal Justice Ethics 4: 3–11.
Musto, David. 1973. The American Disease: Origins of Narcotic
Control. New Haven, Conn.: Yale University Press.
. 1987a. The American Disease: Origins of Narcotic Con-
trol. Expanded edition. (Originally published 1973.) New
. 1987b. “Changing Attitudes toward Drugs and Alcohol
242 REFERENCES
in the United States.” In Drugs and Crime: Workshop Proceed-
ings, ed. Jeffrey A. Roth, Michael Tonry, and Norval Morris.
A Report of the 1986 National Academy of Sciences Confer-
ence on Drugs and Crime Research. Washington, D.C.: Na-
tional Academy of Sciences.
Nagel, Ilene H., and Stephen J. Schulhofer. 1992. “A Tale of
Three Cities: An Empirical Study of Charging and Bargaining
Practices under the Federal Sentencing Guidelines.” Southern
California Law Review 66: 501–66.
Ohlin, Lloyd E. 1951. Selection for Parole: A Manual of Parole
Prediction. New York: Russell Sage Foundation.
Parent, Dale. 1988. Structuring Sentencing Discretion: The Evolu-
tion of Minnesota’s Sentencing Guidelines. Stoneham, Mass.:
Butterworth.
Pease, Ken. 1985. “Community Service Orders.” In Crime and Jus-
tice: An Annual Review of Research, vol. 6, ed. Michael Tonry
Pelikan, Jaroslav. 1985. Jesus through the Centuries: His Place
in the History of Culture. New Haven, Conn.: Yale University
Press.
Peter D. Hart Research Associates, Inc. 2002. The New Politics
of Criminal Justice: Summary of Findings. Washington, D.C.:
Hart Associates.
Posner, Richard. 1977. Economic Analyses of Law, 2d ed. Boston:
Little, Brown.
. 1980. “Optimal Sentences for White-Collar Offenders.”
American Criminal Law Review 17: 409–18.
Roberts, Julian V. 1992. “Public Opinion, Crime, and Criminal
Justice.” In Crime and Justice: A Review of Research, vol. 16,
, and Loretta J. Stalans. 1997. Public Opinion, Crime, and
Criminal Justice. Boulder, Colo.: Westview Press.
, Loretta J. Stalans, David Indermaur, and Mike Hough.
2002. Penal Populism and Public Opinion. New York: Oxford
University Press.
Rose, Dina, and Todd R. Clear. 1998. “Incarceration, Social Capi-
tal, and Crime: Implications for Social Disorganization The-
ory.” Criminology 36(3): 441–80.
REFERENCES 243
Rothman, David J. 1980. Conscience and Convenience. Boston:
Little, Brown.
Schulhofer, Stephen J., and Ilene H. Nagel. 1997. “Plea Negotia-
tions under the Federal Sentencing Guidelines.” Northwest-
ern University Law Review 91: 1284–1316.
Schünemann, Bernd. 1998. “Zum Stellenwert derpositiven Gen-
erolprävention in einer dualistischen Straftheorie.” In Posi-
tive Generalpravention, ed. Andrew von Hirsch, Bernd Schüne-
mann, and Nils Jareborg. Heidelberg: C. F. Mueller Verlag.
Simon, Jonathan. 2001. “Fear and Loathing in Late Modernity:
Reflections on the Cultural Sources of Mass Imprisonment in
the United States.” Punishment and Society 3(1): 21–34.
, and Malcolm M. Feeley. 1995. “True Crime: The New
Penology and the Public Discourse on Crime.” In Punishment
and Social Control, ed. Thomas G. Blomberg and Stanley Co-
hen. New York: Aldine de Gruyter.
Spierenburg, Pieter, ed. 1984. The Emergence of Carceral Institu-
tions: Prisons, Galleys, and Lunatic Asylums, 1550–1900. Rot-
terdam: Erasmus Universiteit.
Stanko, Elizabeth A. 2000. “Victims R US: The Life History of
‘Fear of Crime’ and the Politicization of Violence.” In Crime,
Risk and Insecurity: Law and Order in Everyday Life and Politi-
cal Discourse, ed. Tim Hope and Richard Sparks. London:
Routledge.
Stenson, Kevin. 2000. “Some Day Our Prince Will Come: Zero
Tolerance Policing and Liberal Government.” In Crime, Risk
and Insecurity: Law and Order in Everyday Life and Political
Discourse, ed. Tim Hope and Richard Sparks. London: Rout-
ledge.
, and Robert R. Sullivan, eds. 2001. Crime, Risk and
Justice: The Politics of Crime Control in Liberal Democracies.
Cullompton, Devon: Willan.
Sutherland, Edwin H. 1950. “The Diffusion of Sexual Psycho-
path Laws.” American Journal of Sociology 56: 142–48.
Tak, Peter J. P. 2001. “Sentencing and Punishment in The Neth-
erlands.” In Sentencing and Sanctions in Western Countries,
ed. Michael Tonry and Richard S. Frase. New York: Oxford
University Press.
244 REFERENCES
Tonry, Michael. 1994. “Proportionality, Parsimony, and Inter-
changeability of Punishments.” In Penal Theory and Practice,
ed. R. A. Duff, S. E. Marshall, R. E. Dobash, and R. P. Dobash.
Manchester, U.K.: Manchester University Press.
. 1995. Malign Neglect: Race, Crime, and Punishment in
America. New York: Oxford University Press.
. 1996. Sentencing Matters. New York: Oxford University
Press.
. 1997. “Ethnicity, Crime, and Immigration.” In Ethnic-
ity, Crime and Immigration: Comparative and Cross-national
Perspectives, ed. Michael Tonry. Chicago: University of Chi-
cago Press.
, ed. 2001. Penal Reform in Overcrowded Times. New
, and Richard S. Frase, eds. 2001. Sentencing and Sanctions
in Western Countries. New York: Oxford University Press.
, and David A. Green. 2003. “Criminology and Public
Policy in the U.S. and U.K.” In The Criminological Foundations
of Penal Policy, a volume prepared in honour of Professor
Roger Hood, ed. Lucia Zedner and Andrew Ashworth. Ox-
ford: Oxford University Press.
, and Kate Hamilton, eds. 1995. Intermediate Sanctions
in Overcrowded Times. Boston: Northeastern University Press.
, and Kathleen Hatlestad, eds. 1997. Sentencing Reform
in Overcrowded Times: A Comparative Perspective. New York:
, Lloyd E. Ohlin, and David P. Farrington. 1991. Human
Development and Criminal Behavior. New York: Springer-
Verlag.
Törnudd, Patrik. 1993. Fifteen Years of Declining Prisoner Rates.
Helsinki: National Research Institute of Legal Policy.
. 1997. “Sentencing and Punishment in Finland.” In Sen-
tencing Reform in Overcrowded Times: A Comparative Perspec-
tive, ed. Michael Tonry and Kathleen Hatlestad. New York:
Tremblay, Richard, and Wendy M. Craig. 1995. “Developmental
Crime Prevention.” In Building a Safer Society: Strategic Ap-
proaches to Crime Prevention, ed. Michael Tonry and David
REFERENCES 245
P. Farrington. Vol. 19 of Crime and Justice: A Review of Re-
search, ed. Michael Tonry. Chicago: University of Chicago
Press.
Turner, S., P. W. Greenwood, E. Chen, and J. Fain. 1999. “The
Impacts of Truth-in-Sentencing and Three-Strikes Laws:
Prison Populations, State Budgets, and Crime Rates.” Stanford
Law and Policy Review 11: 75–91.
Tyler, Tom. 1990. Why People Obey the Law. New Haven, Conn.:
Yale University Press.
. 2003. “Procedural Justice, Legitimacy, and the Effec-
tive Rule of Law.” In Crime and Justice: A Review of Research,
vol. 30, ed. Michael Tonry. Chicago: University of Chicago
Press.
, and Robert Boeckmann. 1997. “Three Strikes and
You’re Out, But Why?” Law and Society Review 31: 237–65.
Virginia Criminal Sentencing Commission. 1997. Annual Report.
Richmond: Virginia Criminal Sentencing Commission.
von Hirsch, Andrew. 1993. Censure and Sanctions. Oxford: Claren-
don.
, and Andrew Ashworth, eds. 1998. Principled Sentenc-
ing. 2d ed. Oxford: Hart Publishing.
Wacquant, Loic. 2001. “Deadly Symbiosis: When Ghetto and
Prison Meet and Merge.” Punishment & Society 3: 95–134.
Walker, Nigel. 1991. Why Punish? Oxford: Oxford University
Press.
. 1999. Aggravation, Mitigation, and Mercy in English
Criminal Law. London: Blackstone.
Walker, Samuel. 1977. A Critical History of Police Reform: The
Emergence of Professionalism. Lexington, Mass.: Lexington
Books.
. 1998. Popular Justice: A History of American Criminal
Justice. 2d ed. New York: Oxford University Press.
Weigend, Thomas. 1997. “Germany Reduces Use of Prison Sen-
tences.” In Sentencing Reform in Overcrowded Times, ed. Mi-
versity Press.
. 2001. “Sentencing and Punishment in Germany.” In
Sentencing and Sanctions in Western Countries, ed. Michael
246 REFERENCES
Tonry and Richard S. Frase. New York: Oxford University
Press.
Weiss, Carole H. 1983. “Ideology, Interests, and Information:
The Basis of Policy Decisions.” In Ethics, the Social Sciences,
and Policy Analysis, edited by D. Callahan and B. Jennings.
London: Plenum Press.
. 1986. “Research and Policy-Making: A Limited Partner-
ship.” In The Use and Abuse of Social Science, ed. F. Heller.
London: Sage.
Whitman, James Q. 2003. Harsh Justice: Criminal Punishment
and the Widening Divide between American and Europe. New
Wilson, James Q. 1993. The Moral Sense. New York: Free Press.
. 1995. “Crime and Public Policy.” In Crime, ed. James
Q. Wilson and Joan Petersilia. San Francisco, Calif.: ICS Press,
Institute for Contemporary Studies.
, and Richard Herrnstein. 1985. Crime and Human Na-
ture. New York: Simon and Schuster.
Wright, Ronald F. 2002. “Counting the Cost of Sentencing in
North Carolina, 1980–2000.” In Crime and Justice—A Review
of Chicago Press.
Yankelovich, Dan. 1991. Coming to Public Judgment: Making
Democracy Work in a Complex World. Syracuse, N.Y.: Syra-
cuse University Press.
Young, Warren, and Mark Brown. 1993. “Cross-national Com-
parisons of Imprisonment.” In Crime and Justice: A Review
of Chicago Press.
Zedlewski, Edwin W. 1987. Making Confinement Decisions.
Washington, D.C.: U.S. Department of Justice, National Insti-
tute of Justice.
Zimring, Franklin E. 1982. The Changing Legal World of Adoles-
cence. New York: Free Press.
. 1989. “Toward a Jurisprudence of Family Violence.” In
Family Violence, ed. Lloyd Ohlin and Michael Tonry. Vol. 11
of Crime and Justice: A Review of Research, ed. Michael Tonry
REFERENCES 247
, and Gordon Hawkins. 1988. “The New Mathematics
of Imprisonment.” Crime and Delinquency 34: 425–36.
. 1991. The Scale of Imprisonment. Chicago: University
of Chicago Press.
. 1995. Incapacitation: Penal Confinement and the Re-
straint of Crime. New York: Oxford University Press.
. 1997. Crime is Not the Problem: Lethal Violence in
America. New York: Oxford University Press.
, Gordon Hawkins, and Sam Kamin. 2001. Punishment
and Democracy: Three Strikes and You’re Out in California.
248 REFERENCES
Index
abortion appropriations laws, proposals for, 19–20

partisan politics of, 18, 45–46, 77 arson, as Index Crime, 114
sensibilities regarding, 175 assassinations, legislation as result of, 93, 95
abstinence, from alcohol, 110–111 assault. See aggravated assault
abuse Atkins v. Virginia, 75
contemporary sensibilities of, 144–145 Atlanta, comparative homicide rates in,
punishments for, 67–68 121–122
acceptances of offer, in civil law disputes, attorneys
175 fees for, 192
acquittals, real offense sentencing for, power with real offense sentencing,
146–150 149–150
administration, as civil service focus, 207 trial (see defense attorneys; prosecutors)
administration agencies, delegating policy U.S. appointed, 208
making to, 210–213, 219, 221 Australia
aesthetic preferences, changing cycles of, crime rates in, 128
64–65, 78–79, 83–84, 101 punishment policies in, 97, 148, 151, 162
affirmative action, partisan politics of, 39–
40, 46 beheadings, as punishment, 10, 66, 72–73
age, as criminality factor, 146, 152–155 social status and, 58
aggravated assault Belgium, long-term crime trends in, 102–103
direct costs of, 192 Bell v. Wolfish, 173
as Index Crime, 114 benefits, of crime and punishment. See cost-
just punishments for, 203 benefit analysis
reporting patterns for, 115, 118 Bentham, Jeremy, 141–142, 171, 191, 193
short-term trends of, 116–118 rational actor proposals of, 189–190
alcohol use bias. See also disparity analyses
cycles of public attitudes on, 80–81 in criminal justice policies, 215, 223
nineteenth-century prohibition of, 80, 82, Bias, Len, 5, 139, 204–205
109–111 bigotry, in criminal justice system, 226–227
recovery vs. relapse of, 199 black Americans
alienation, 163. See also disintegrative disparity analysis for, 20, 96, 214–215,
shaming 221–225
Alvarez, The People v., 17 disparity audits on, 225–227
American Law Institute (ALI) drug use/drug policy associations, 109,
on Model Penal Code, 176–177, 180–186 112, 133–134, 221–222
organizational objectives of, 177–179 penal policy as damaging to, 3, 9, 11, 63,
Andrade, The People v., 17, 137 214–215
anger, penal policies based on, 83, 96, 200– as target of partisan politics, 39–40, 76,
202, 204 158, 198
strategies for changing, 204, 208 Black Caucus, Congressional, on crack co-
anthropology, risk society analysis in, caine, 204
46–52 Blair, Tony, 13
antiterrorism programs, politics of, 44–46, boot camp, 217
218 Bourne, Peter, 81
anxiety broken windows policing, 120, 122–123
moral (see moral panics) Buckley, William F., 133
public (see public fear/anxiety) bureaucratization, long-term crime trends
apostates, persecution of, 65, 78 and, 107, 138
appellate courts burglaries, statistics on, 28–29
accounting of real offense sentencing, short-term trends, 113–114, 116–118
147–148 three-strikes law impact on, 123
in preventive detention debate, 169, Bush, George, 39
173–174 Bush, George W., 4, 219
249
California collateral effects, of punishment, 171–172,
harsh legislation in (see three-strikes law) 189–191, 193
political distortion of policies in, 212 commensurability, central premise of, 170
recent crime rates in, 123–124 communicative theories, of punishment,
sentencing reform in, 16, 217–218 159–160
uniform sentencing law in, 125–127 community
Canada in cost-benefit analysis, 171–172, 189–
crime rates in, 12, 14 191, 193, 214
international comparisons, 127–129 as local legal culture, 209–210
imprisonment rates in, American rate vs., socialization, crime prevention through,
128–129 107–108, 161–162
capital crimes, preventive detention for, community justice programs, 137–138
169, 172 community service, as punishment, 158,
capital punishment. See also specific tech- 163–166
nique consciousness, impact on crime control poli-
in America cies, 10, 160
changing cycles of, 66–67, 75, 175 consequentialists, beliefs of, 202–203
de facto moratorium on, 97 conservatives and conservatism
endorsement of, 6, 77, 135 American Law Institute as, 178–179
of mentally retarded offenders, 75 impact on imprisonment rates, 23, 25, 32,
politics of, 45–46, 77 46–51
social status impact on, 58, 222–223 partisan politics of, 38–41, 136,
European examples of, 58, 66–67, 72–73 196–197
career ladders, civil service, in prisons, 22 on preventive detention, 170, 172–173
Chicago, comparative homicide rates in, Constitution, U.S.
121–122 Eighth Amendment of, punishment appli-
child abuse cations of, 67, 75
moral panic and, 91–95 ex post facto principle of, 144, 157
punishments for, 67–68 contracts
child labor, sensibilities support of, 6 American Law Institute on, 178
child molestation. See sexual psychopaths/ in civil law disputes, 175
predators corporal punishment
children, as criminals. See juvenile of- constitutionality of, 97
fenders examples of, 66–67, 72–73
chronocentrism, 55 cost-benefit analysis
church officials, impact on crime control pol- of crime and punishment
icies, 6–7, 10 accounting of offenders, 171–172,
citizen reporting 189–194, 214
as crime rate factor, 114–118 Bentham’s rational rules for, 189–190
on personal drug use, 131–133 fiscal vs. opportunity, 41, 191
civil law disputes, acceptances of offers in, measurement methods, 170–171
175 of crime prevention, 191–192, 224–225
civil liberties/rights crack cocaine, 112, 115, 204, 218–219
crime control policy trade-off between, 8, credibility, of government, 43–44, 46
96, 175, 224–225 crime(s)
as sentencing reform focus, 214, 218 cost-benefit analysis of
as timeless idea, 77–78, 136–137 accounting of offenders, 171–172,
Civil Rights Movement, 39–40, 214 189–194
civil service/servants Bentham’s rational rules for, 189–190
career, politics and influence of, 22, fiscal vs. opportunity, 41, 191
206–210 measurement methods, 170–171
politicians vs., 9–10 expectancy tables for prediction of, 49
prison career ladders in, 22 legislative definition of, 145
Civil War, impact on crime policy, 63, 78, as most pressing problem, 16, 36–37
109 origin debate on, 12–13, 17
civility. See respect political reconfiguration of, 41–46
civilization curves, 72 severity of (see offense severity)
“civilizing process,” sensibilities as, 71–72, crime control policies. See criminal justice
75, 107 policies
Clacton riots, as moral panics, 86–88 crime prevention
classicism, recurring cycles of, 6, 64, 83, 101 costs vs. benefits of, 191–192, 224–225
Cleveland, comparative homicide rates in, situational, rehabilitation vs., 48
121–122 through community socialization, 107–
Clinton, Bill, 8, 38, 136, 196, 212, 219 108, 161–162
cocaine trafficking, 100-to-1 sentencing law crime rates
for, 4–5, 15, 81 crime control policies impact on, 97–139
remediation of, 218–219, 221, 226 American perspectives, 11–12, 14, 40,
repeal of, 133–134, 204 118–119, 134–135
cocaine use city comparisons, 121–122
cycles of public attitudes on, 6–9, 15–16, drug policy cycles and, 100, 113,
68–69, 80–81 130–139
punishments for, 68, 112 drug use cycles and, 100, 109–112
recent self-reported data on, 131–132 international comparisons, 118–119,
trends of, 112, 115 127–130
cognitive dissonance, 77 intolerance cycles and, 112–113,
Coker v. Georgia, 67, 75 119–127
250 INDEX
long-term influences, 100–109 expressive content of, 42–44, 52, 55, 197
oscillation of human values and, as formerly unthinkable, 141–146
100–101 for children, 146, 150–156
short-term influences, 113–127 disintegrative shaming as, 144, 146,
state comparisons, 124–125 156–167
as decreasing real offense sentencing, 146–150
factors influencing, 12, 14, 96, 118–127 impact on crime rates, 97–139
in short-term, 113–118 American perspectives, 11–12, 14, 40,
impact on imprisonment rates, 23, 27–34 118–119, 134–135
impact on penal policies, 5, 14, 96 city comparisons, 121–122
international comparisons, 12, 14, 27–28 drug policy cycles and, 100, 113,
crime control policies and, 118–119, 130–139
127–130 drug use cycles and, 100, 109–112
democracy and social status in, 57–59, international comparisons, 118–119,
81 127–130
standardized surveys for, 117–118, 127 intolerance cycles and, 112–113,
long-term, 100–109 119–127
in European countries, 101–105 long-term influences, 100–109
as unrelated to crime policies, 100–101 oscillation of human values and,
in Western countries, 104–109 100–101
postmodernist perspectives of, 52–53 short-term influences, 113–127
short-term state comparisons, 124–125
as decreasing, 113–118 in last third of twentieth century, 3–5, 8,
factors influencing, 12, 14, 96, 118–127 22–23
criminal(s) postmodern perspectives of, 51–56,
costs of crime and punishment for, 171– 196–200
172, 189–194, 203, 214 political courage to repeal, 15, 18–19
immigrants as, 109–110, 112, 195–196 preventive, 48, 107–108, 161–162, 191,
repeat offending by (see recidivism) 196, 224
rights of, 136–137 professional model for, 208–210
stereotypes of, 49–50, 214 public safety trade-off between, 8, 143,
young 170, 177, 224–225
rehabilitation of, 105, 151 repressive, 196
unthinkable punishments for, 146, retroactivity of, 144–145
150–156 sensibilities impact on, 70–78
criminal justice policies cycles of, 7–8, 25, 59–61, 85
anecdotes illustrating abhorrence of, criminal record
195–196 in juvenile punishment policies, 152–154
career civil service influence on, 206–210 in real offense sentencing, 149
collective tasks for changing, 199–200 recidivism rates and, 188
ethical introspection of, 199–204 criminologists, on risk society analysis,
institutional change for, 199–200, 49–50
204–213 “criminology of the other,” in postmodern
proposals for, 19–20 angst, 25–26, 52, 54, 56
remediation of, 199, 213–227 cruelty, 17, 21, 26, 69
cost of (see cost-benefit analysis) evolving standards of, 26, 74, 84, 143
crime rates impact on, 5, 14, 96 culture
delegation of authority for, 210–213, 219, impact on punishment policies, 24–26, 57
221 drug use and, 109–112
drug policy cycles impact on, 130–139 social status differentiation and, 57–59,
Musto’s hypotheses on, 130–131 81, 90
self-reported drug use and, 131–133 through career civil servants, 209–210
in Europe timeless ideas and, 77–78
as egalitarian, 24–25, 27 risk society analysis based on, 46–50
for juveniles, 150–151 postmodernist angst and, 51–56
shaming as, 161–162, 165–166 self-defense behaviors of, 90
status differentiation with, 57–59 cycles, social
vs. Western countries, 97, 195–196, of drug law attitudes, 6–9, 15–16
206 of human life, 100–101
evolution of harsh, 23–61 of human rights, 6–8
crime rates and, 23, 27–34 impact on crime policies, 7–8, 25
democracy and degradation impact on, proportionality perspectives, 141–146,
24–25, 57–59 202
elimination of strict construction of, 158 sensibilities perspectives, 59–61, 79–85
partisan politics impact on, 23, 38–41, of political movements, 59–61
136, 196–197, 206–210 of tolerance/intolerance
political reconfiguration and, 23, 41– sensibilities impact on, 6–8, 60, 85
46, 108 truths and trivia about, 64–69
postmodernist angst and, 23–24, 51–56 cynicism, political, impact on penal policies,
public opinion and, 23, 34–38, 133 5, 11, 41, 60, 197–198, 216
“risk society” and, 23, 46–51
sensibilities impact on, 25, 59–61, 70– Dahmer, Jeffrey, 138
78, 210 Davis, Richard Allen, 205
social status impact on, 57–59, 81 day fines, 165
social values impact on, 73–75, 77–78, death, as cost of crime, 171, 192
209–210 death penalty. See capital punishment
INDEX 251
debasement, deliberate, 144, 146, 156–167 double jeopardy doctrine, 174
decency, evolving standards of, 74–75, 84 draft evasion, 67
decision making, as Model Penal Code drug laws
theme, 177 drug use cycles impact on, 100, 109–112
defense attorneys as ethnically disparate, 109, 133–134,
real offense sentencing and, 150 221–222
in right to appeal sentences, 173–174 impact on criminal justice policies, 100,
degradation, of prisoners 113, 130–139
deliberate (see disintegrative shaming) Musto’s hypotheses on, 130–131
as punishment policy component, 24–25 self-reported drug use and, 131–133
social status impact on, 57–59, 81 as prohibitionistic, 111–112, 204
delegation, of policy-making authority, proportionality perspectives of, 143–144
210–213, 219, 221 public opinion on, 80–82, 203
delinquency, as juvenile offense factor, 154 referenda for reform of, 16–17, 20, 76,
democracy, impact on punishment policies, 218–219
24–25 as too harsh and wasteful, 10–11, 15,
social hierarchy factor, 57–59 133, 204
Democratic politics, impact on imprison- drug trafficking
ment rates, 32, 39, 196–197 100-to-1 sentencing law for, 4–5, 15, 81
“deserved punishment,” Model Penal Code remediation of, 218–219, 221, 226
and, 170, 176–179, 185 repeal of, 133–134, 204
detention, pretrial. See preventive detention cycles of punishment policy for, 68, 80,
determinate sentencing, 106, 125 143, 204, 226
deterrence drug treatment programs, effectiveness de-
cost vs. benefit of, 189–191 bate about, 16, 20
current state of knowledge on, 203 drug use
Detroit, comparative homicide rates in, cycles of public attitudes on, 6–9, 15–16,
121–122 68–69, 80–83
Deukmeijian, George, 17 drug policy relationship with, 100,
DiFrancesco, United States v., 174 109–112
dignity. See respect recent self-reported data on, 131–132
disapprobation, in reactions to crime, stereotyping, 111–112
162–163 due process
discretion, as Model Penal Code theme, 177 for criminals, 136
discrimination preventive detention and, 170, 172–173
political movements impact on, 42–44 Durkheim, Émile, 99, 108, 157, 161
unconscious, in crime policies, 215, 223
disgust, with offenders, 162. See also expres- early release laws, proposals for, 19
sive punishment economics
disintegrative shaming, 156–167 changing sensibilities attributed to, 73,
based on communicative theories, 107, 145
159–160 of crime and punishment, 170–171, 189–
based on moral-educative theories, 157, 194, 214
160–162, 166 political movements impact on, 42–43
expressive perspectives of, 157, 159–163 educational programs, government role in
Kahan’s arguments on, 146, 156–159, failure of, 43
163–166 egalitarian values, impact on punishment
object of, 144, 146 policy, 24–25, 27, 57–59
public knowledge and opinion in, 158, Eighth Amendment, of U.S. Constitution,
163–167 punishment applications of, 67, 75
overbreadth perspectives, 165 elections, as political force on crime, 41,
parochialism and, 165–166 44–46, 207
as under-informed, 164–165 elemental fairness, of punishments, 13
rehabilitation vs., 158, 161, 164 emotion, public. See also moral panics
reintegrative shaming vs., 162–163 collective tasks for changing, 199–200
dismissed charges, real offense sentencing ethical introspection, 200–204
for, 146–150 institutional change, 199–200, 204–213
disparity analyses, proposals for, 19–20, 96, remediation, 199, 213–227
214–215 penal policies based on, 5, 9, 11, 18, 96
as sentencing legislation requirement, policy makers avoidance of acting on, 19
221–225 short-term, in career civil service,
disparity audits, as remediation, 225–227 206–207
disproportionate punishments empathy, impact on crime policy, 200–201
for minority defendants, 214–215, employment, in cost-benefit analysis, 171,
221–225 192, 214–215
Model Penal Code prevention of, 170, equality, as timeless idea, 77
176, 180 equity, as Model Penal Code theme, 182
public imposition of, 141–143 ethics
strategies for reducing, 211, 214, 216–227 introspection for policy changes, 199–204
dispute resolution as timeless idea, 77–78, 144–145
in civil law, 175 ethnic minorities. See minority Americans;
as criminal justice focus, 138 specific ethnicity
diversion programs, for drug offenders, 218 ethnocentrism, 55, 76
domestic violence ethos, social definition of, 67, 70–71
contemporary sensibilities of, 145, 166 Europe
punishments for, 67–68 capital punishment policy in, 97
252 INDEX
civil servant selection in, 206–207 Gacy, Richard, 138
crime rates in, 12, 14, 26 gang wars, 115, 121, 215
imprisonment trends and, 27–34 Gantt, Harvey, 39
international comparisons, 127–128 gay rights
long-term, 12–14, 101–105 cycles of sensibilities on, 6, 64–66
short-term, 116–118 politics of, 45
expansion of human rights, 21–22 gender-neutrality, for punishments, 154
imprisonment rates in, 21–22, 144 generational criminal activities, of immi-
crime rate trends and, 32–34 grants, 195–196
juvenile crime policies in, 150–151 Georgia, Coker v., 67, 75
punishment policies of Germany
as egalitarian, 24–25, 27 crime rates in
shaming as, 161–162, 165–166 impact on imprisonment, 28–33
status differentiation with, 57–59, long-term, 12–14, 27–28, 102, 104
225–226 social status impact on, 57–59
vs. Western countries, 97, 195–196, 206 Durkheimian punishment policies of, 161
ex post facto principle, 144, 157 Giuliani, Rudolph, 108, 120
executions. See capital punishment Goldwater, Barry, 38
existential angst, historical example of, good-faith proposals, on racial policies, 227
90–91 good time credits, in Model Penal Code, 182
expressive punishment governing-through-crime, 55
in criminal justice policies, 42–44, 52, 55, government, American
197 collapse of confidence in, 43–44, 46
in disintegrative shaming, 157, 159–163 determination of punishment, 14, 19,
146, 217
Fair Sentencing Act (1981), of North Caro- partisan politics of, 9–10, 17, 19, 136,
lina, 126 196–197 (see also politicians)
fairness, of punishments, 13, 77, 202 in career civil service, 206–210
with imposition of disproportionate, sovereignty in response to crime, 41–46
141–143 delegation to specialized agencies,
individualized vs., 58–59, 70–71 210–213, 219, 221
in real offense sentencing, 147–148 with drugs, 133–134
false consciousness, 10 by juveniles, 146, 150–156
family, of offenders, in cost-benefit analy- pessimism about effectiveness of, 134
sis, 171–172, 189–194, 214 postmodernist perspectives of, 52–53
fear. See public fear/anxiety real offense sentencing and, 148, 150
federal government Great Britain, crime rates in
sentencing guidelines of, 146, 217 long-term, 12–14, 85, 101–102
sovereign response to crime, 41–46 social status impact on, 57–59, 225–226
delegation to specialized agencies, 211, guilty pleas, disparate social factors of,
213, 219 225–226
Federal Reserve Board, as delegation exam- gun control
ple, 212 legislative origins of, 93–95
fees politics of, 18, 45–46, 77
for attorneys, 192 sensibilities regarding, 175
for security firms, 191 Gun Control Act (1968), 93
feminist movement, impact on penal poli- gun violence
cies, 166–167, 221 costs of, 193
fines, as punishment, 164–166 statistics on, 28–29
Finland, crime rates in, 12–14 zero-tolerance policing impact on, 121
impact on imprisonment, 28–30, 32–33 hangings, social status and, 58
first-offender penalties Harrison Act (1914), 80, 112
for non-violent drug crimes, 134 Hatch, Orrin, 15, 134
recidivism rates and, 186–189 health, personal responsibility for, 132–133
fiscal costs, of imprisonment policies, 41, health care system, government role in fail-
170–171, 191 ure of, 43
Flood, Gerald F., 184 Helms, Jessie, 39
Florida Hendricks, Kansas v., 173
prison building programs in, 221 heretics
prisoner population caps in, 220 as moral panics example, 65, 88–90
recent crime rates in, 124–125 persecution of, 6, 10, 78, 84
sentencing commissions in, 211 heroin trafficking, punishments for, 68
forgiveness, as crime policy basis, 195, heroin use, recent self-reported data on,
198–199 131–132
formalism, 83 Hispanic Americans
France drug use/drug policy associations, 109,
capital punishment examples from, 66– 112
67, 72–73 penal policy as damaging to, 3, 11
crime rates in, 27, 128 disparity analyses of, 214–215, 225,
impact on imprisonment, 30–31 227
social status impact on, 57–59 homicide
Frankel, Marvin, 210 just punishments for, 203, 224
freedom, as timeless idea, 77 statistics on, 28–32
Freud, Sigmund, 198 city comparisons, 121–122
Friedman, Milton, 133 international comparisons, 128–129
frustration, impact on crime policy, 200 short-term trends, 113–114
INDEX 253
[homicide] (continued) Japanese Americans, sensibilities on intern-
state comparisons, 124–125 ment of, 6, 76
zero-tolerance policing impact on, Jefferson, Thomas, 78–79, 141–142
121–123 Jesus, changing views of public beliefs in,
homosexuals, cycles of sensibilities on, 6, 65, 79
45, 64–66 Jim Crow laws, 78
Horton, Willie, 39, 60 Johnson, Garry, 133
Houston, comparative homicide rates in, judges
121–122 authority to circumvent three-strikes law,
human behavior, impact of oscillations of, 10, 17, 137
100–101 as remediation, 216–217
human rights partisan politics of, 9–10, 19, 207–210
cycle of sensibilities impact on, 6–8 as politically influenced civil servants,
European expansion of, 21–22 206–210
ex post facto principle and, 144 professionalization of, 208–210
moral panics and, 96 public opinion influence on, 184–185
radical movements for, 166–167, 221 role in real offense sentencing, 146–149
as timeless idea, 77–78, 137 training for, 207–208
humiliation, deliberate, 144, 146, 156–167 jury awards, in cost-benefit analysis, 192
hypocrisy, in criminal justice system, just deserts
222–223 central premise of, 170, 176, 179
hysteria. See moral panics Model Penal Code and, 176–179, 185
retributivists vs. consequentialists views
Illinois, recent crime rates in, 124–125 on, 202–203
immigrants and immigration justice
drug use/drug policy associations, 109– as civil service focus, 207–208
110, 112 as timeless idea, 77
generational criminal activities of, juvenile court, long-term crime trends and,
195–196 105
partisan politics of, 46, 60, 76 juvenile offenders
imprisonment formerly unthinkable sentencing for, 146,
cost vs. benefit of, 191–194 150–156
determination of (see prison sentences) rehabilitation of, 105, 151
effectiveness of, current state of knowl-
edge on, 203, 218 Kahan, Dan, 146, 156–159, 163–166
rehabilitation vs., 34, 36, 48–49, 105, Kanka, Megan, 60, 92, 139
151, 176–177 Kansas, prisoner population caps in, 221
selective, for juvenile offenders, 153–156 Kansas v. Hendricks, 173
state-based population caps for, 220 Kant, Immanuel, 141, 176, 186, 193
imprisonment rates, American Kemp, McClesky v., 222–223
vs. European countries, 21–22, 26, 128 Kennedy, Robert, 93, 95
crime rates trends and, 27–34, 118–119 King, Martin Luther, 93, 95
vs. other Western countries, 3, 21, 28, Klaas, Marc, 205
97–98, 119, 128–129 Klaas, Polly, 5, 60, 92, 95, 139, 205
incapacitation. See imprisonment Koon v. United States, 137
indeterminate sentencing, 146, 148, 173, 179 Kross, Anna M., 184
long-term crime trends and, 105–106
Index Crimes, classification of, 114 law-and-order movements, 60
indigenous peoples, North American, sensi- legislators and legislation
bilities on near-extermination of, delegation of authority to specialized
5–6 agencies, 210–213
individualization determinate sentencing and, 126–127
as aspect of contemporary life, 47–48, 82 motivation for penal policies, 3–5, 15, 18,
as Model Penal Code theme, 173, 177 145
punishment policies consideration of, 58– partisan politics of, 9–10, 17, 136,
59, 70–71 196–197
as timeless idea, 77 as stuck in current policies, 11–12, 15,
inner voices, 198 18, 219
innocence, presumption of, 172–173 “windows of opportunity” for, 94–96
insecurity(ies) lex talionis principle, of proportional punish-
postmodernist perspectives of, 51–52, 54 ment, 141–142
public, 83, 142 liberal reformers
institutional change, for policy changes, on preventive detention, 170, 172–173
204–213 view on policy making, 95, 136, 179, 193
career civil service and, 206–210 liberty. See freedom
as collective task, 199–200 live-and-let-live attitudes, 80
delegated policy making and, 210–213, Los Angeles, comparative homicide rates in,
219, 221 121–122
moral panics impact on, 204–205 Lynaugh, Penry v., 75
institutional memory, 208, 212
interest groups, impact on penal policies, mandatory minimum sentencing, 135
18, 44–46, 77 for drug offenses, 68, 81, 134
International Crime Victimization Survey Model Penal Code basis for, 180, 183, 185
(ICVS), 28, 116–118 partisan politics of, 40–41, 206
trend comparisons with, 127–128 repeal of, 19, 216–217
intolerance. See tolerance/intolerance manslaughter, just punishments for, 203
254 INDEX
Manson, Charles, 138–139 moral entrepreneurs, 98, 199
Mari moral panics
Marijuana Tax Act (1937), 80, 112 Clacton riots as, 86–88
marijuana use collective tasks for controlling effects of,
legalizing of, 81 199–200
for medical use, 8, 16, 134, 137 ethical introspection, 200–204
punishments for, 68, 112 institutional change, 199–200, 204–213
public attitude cycles on, 6–9, 15–16, remediation, 199, 213–227
68–69, 80–81 connotative meanings of, 86, 92–93, 96
recent self-reported data on, 131–132 examples of, 5, 65, 85–86, 204
marital violence. See domestic violence impact on crime trends, 98–100, 198
maximum sentencing laws impact on penal policies, 5, 9, 19, 51
Model Penal Code basis for, 180, 183 in contemporary U.S., 137–139
real offense sentencing and, 149 impact on proportionality, 143–144
McCaffrey, Barry, 219 Salem witch trials as, 86, 88–91
McCarthyism, 78 sexual psychopaths and, 86, 91–96
McClesky v. Kemp, 222–223 short-term vs. long-term sensibility cy-
McVeigh, Timothy, 97 cles, 98–100
media sensationalism theoretical basis and applications of, 85–
impact on public opinion, 34–35, 37, 60 96, 98
as moral panics factor, 5, 85–86, 88 and “windows of opportunity,” 94–96
in political agendas, 44–45 morality/immorality, views of
Megan’s Laws, 92–94, 135 cycles of, 64–69
memory, institutional, 208, 212 in partisan politics, 40–41, 46, 198–199
mentally ill, sensibilities on sterilization of, 6 post-Enlightenment debates on, 176
mentally retarded offenders, culpability limi- prohibitionistic, 109–112
tations of, 75 as timeless idea, 77, 142, 145
mercy, in punishment policies, 59 motives, in crime policies
Miami, comparative homicide rates in, as bad, 222–223, 226–227
121–122 of politicians, 3–5, 15, 18, 145
middle class, postmodernist perspectives of, motor vehicle theft
51–56 short-term trends of, 113–114, 116–118
minimum sentencing laws. See mandatory three-strikes law impact on, 123
minimum sentencing Mower, The People v., 16, 137
Minnesota, sentencing commission in, 211, multiculturalism, 24–26, 57–59, 90
219 timeless ideas based on, 77–78
minority Americans murder. See homicide
disparity analysis for, 20, 96, 214–215, Musto, David, 111–113, 130, 134
221–225
disparity audits on, 225–227 narcotics laws, 80, 112
drug use/drug policy associations, 109– National Crime Victimization Survey
110, 112 (NCVS), U.S., 116–118
penal policy as damaging to, 3, 9, 11 National Rifle Association (NRA), 18, 46
postmodernist perspectives of, 51–56, 76 nationality groups, sentencing legislation im-
misdemeanor policing, 120 pact analyses on, 221–225
mitigation, of sentencing, Model Penal Code negativism, of moral panics, connotative
provisions for, 181–182 meanings of, 86, 96
mob psychology, 198 Netherlands, long-term crime trends in,
Model Penal Code 102–103
American Law Institute proceedings on, New York, recent crime rates in, 124–125
176–186 New York, Williams v., 149
applications of, 38, 164 New York City
authorized prison sentences, 180 comparative homicide rates in, 121–122
authorized probation sentences, 180–181 zero-tolerance policing in, 108, 119–123
good time credits, 182 Nixon, Richard, 39, 80
just deserts and, 170, 176, 179 normative theories
key drafters of, 178–179, 186 in civil liberty issues, 175
main themes of, 170, 177 in crime trends, 108, 111–112, 129–130
mitigation of sentences, 181–182 determinate sentencing and, 126–127
parole release, 183 of punishment, 141, 157–160, 162, 166
public sentiments on, 179, 183–186 North Carolina
purposes of sentencing, 179–180 prisoner population caps in, 220–221
reconsideration of sentences, 181 sentencing laws in, 126, 211, 220
retribution theme of, 170, 176, 185, 193 not-guilty verdicts, real offense sentencing
truth-in-sentencing laws consistency for, 146–150
with, 180, 182, 185
moderation, in punishment policies, 59 offenders. See criminal(s)
modernity offense severity
erosion of in contemporary policies, in expressive punishment, 162
25–26 in juvenile crimes, 151–152, 154–155
postmodernist angst perspectives of, in proportional punishments, 141–143,
51–56 183, 202
Mods and Rockers, moral panics about, for remediation, 217–218
86–88 in real offense sentencing, 149
moral-educative theories, of punishment, tools for identification of high-rate, 156
157, 160–162, 166 Oklahoma, prison building programs in, 221
INDEX 255
100-to-1 sentencing law, for cocaine traf- of partisan politics, 39–41, 196–197
ficking, 4–5, 15, 81 research influence on policy making,
remediation of, 218–219, 221 94–96
repeal of, 133–134, 204 politicians
opium use, U.S. intolerance of, 112 civil servants vs., 9–10
opportunity costs, of imprisonment policies, collapse of confidence in, 41–44, 46
41, 191 courage to repeal harsh penal laws, 15,
order maintenance policing, 120 18–19
Oregon, sentencing reform in, 16, 211, 22 impact on imprisonment rates
0ostracism, 163. See also disintegrative Democratic, 32, 39
shaming election agendas and, 44–46
out-of-pocket costs, of crime and punish- partisan perspectives of, 38–41, 76,
ment, 192 136, 196–197
Republican, 23, 25, 39, 46–51
pain and suffering as leading vs. following public opinion,
for consequentialists beliefs on, 202–203 34–38
as crime and punishment cost, 171, motivation to pass and maintain penal
191–192 laws, 3–5, 15, 18, 145
pardon(s), legitimacy of, 212–213 vulnerability to emotional overreaction,
pardon boards, 213 9, 17–18, 88
parole politics. See also partisan politics
factors influencing policies on, 10, 59, of career civil service, 206–210
105, 212 polls, public opinion, in crime research, 16,
Model Penal Code provisions for, 177, 28, 35–36, 43
179, 183 populist punitiveness, in postmodernist
parole commissions, 212 angst, 52, 56, 197
parsimony, as proportional punishment prin- post-Enlightenment debates, on morality, 176
ciple, 142 postmodernism
partisan politics crime rate perspectives, 52–53
of affirmative action, 39–40, 46 impact on crime policies, 23–24, 51–56,
of American government, 9–10, 17–19 77, 196–200
of career civil service, 206–210 Powell, Lewis, 223
ethnic Americans as target of, 39–40, 76, preventive detention
158, 198 central premise of, 169, 172
impact on crime policy, 23, 38–41, 76, disparate social factors of, 225
136, 196–197 due process debates regarding, 170,
morality debates in, 39–41, 45–46, 60, 172–173
77, 198–199 prosecutorial appeals and, 173–174
Pataki, George, 15, 134 preventive policies, 173, 196
penal culture prison building programs, state-based, 221
sense and sensibility in, 3–20 prison sentences. See also imprisonment
cycles and, 63–84 American measurement of, 6, 9, 22
strict construction principle of, 174 crime policy changes impact on, 134–135
penal policies. See criminal justice policies as harsh and wasteful, 10, 18–19, 196,
penitentiaries. See also imprisonment; 209
prison sentences impact on crime trends
state-based building programs for, 221 long-term, 105–106
Penry v. Lynaugh, 75 short-term, 118–119
The People v. Alvarez, 17 international comparisons, 21–22, 144,
The People v. Andrade, 17, 137 225–226
The People v. Mower, 16, 137 Model Penal Code provisions for, 170,
The People v. Romero, 17 177, 179–182
perfectionism, as human frame of reference, 101 public opinion on, 34, 36–37
personal agendas. See self-interest changing cycles of, 67–69, 209–210
personal origins, of crime, 12–13, 17 remediation strategies for, 199, 213–227
personal responsibility as rite of passage, 70–71
for health and well-being, 132–133 statutes for
partisan politics of, 40–41 maximum, 180, 183
Philadelphia, comparative homicide rates in, minimum, 19, 40–41, 68, 81, 134–135,
121–122 180, 183, 185
plea bargaining state-based, 52, 125–127, 133, 148,
with drug offenses, 81 150, 210
real offense sentencing and, 150 prisoners’ rights movement, 136–137
point system, in juvenile punishment poli- probation
cies, 152–154 long-term crime trends and, 67, 105
Poland, crime rates in, 128 Model Penal Code authorization of,
police 180–181
in cost-benefit analysis, 191 problem solving, as criminal justice focus,
impact on crime trends 138
long-term, 106–107 procedural symmetry, in sentencing, 174
short-term, 120 productivity, in cost-benefit analysis, 171,
political movements, impact on crime pol- 192
icy, 23, 41–46 professional model, of criminal law,
cycles of, 59–61 208–210
political science progressive-loss-of-mitigation doctrine,
of career civil service, 206–210 225–226
256 INDEX
prohibition, 80, 82 public opinion on
nineteenth-century alcohol, 109–111 as complicated, 16–17, 36
twentieth-century drug, 111–112 in imprisonment research, 34–38
proportionality principle, of punishment shaming, 158, 163–167
in Model Penal Code, 170, 176, 180 xenophobic perspectives of, 76
public opinion cycles on, 141–146, 202 restitution as, 164–166
for repeat offenders, 143–146 shaming as
prosecutors disintegrative, 144, 146, 156–167
partisan politics of, 9–10, 19, 136, reintegrative, 162–163
207–210 social values impact on, 209–210
as politically influenced civil servants, cycles of, 5, 13, 63–69
206–210 evolutionary nature of, 73–75, 77–78
professionalization of, 208–210 sensibilities as harsh, 6, 8–9, 12, 16,
real offense sentencing and, 149–150 96, 133
sentence appeals by, 169, 173–174 punishment incommensurability. See disinte-
Protestant Reformation, impact on crime pol- grative shaming
icy, 65, 78
psychology, as crime policy basis, 198 Quakers, as moral panics example, 65,
public fear/anxiety. See also moral panics 88–90
impact on penal policies, 5, 9, 11, 37,
198, 204 race and racism
postmodernist perspectives of, 51–52, 54, in partisan politics, 39–40, 198
83, 197 in penal policies, 3, 9, 11, 60, 222
public judgment sentencing legislation consideration of
on drug use, 132–133 disparity audits on outcomes, 221–225
public opinion vs., 35–36, 40, 145 disparity impact analyses for, 19–20,
public opinion 96, 214–215
on disintegrative shaming, 158, 163–167 radical movements
overbreadth perspectives, 165 for human rights, 166–167, 221
parochialism and, 165–166 for institutional changes
as under-informed, 164–165 in career civil service, 206–210
evidence of changing, 4, 15–16, 85 in policy making authority, 210–213, 221
impact on civil servants, 207, 209–210 rape
impact on imprisonment rates, 23, 34–38 direct costs of, 192
impact on penal policies, 3–5, 8–10, 76 punishments for, 67–68, 75–76
impact on proportionality, 141–143, 202 as just, 203, 224
on Model Penal Code, 179, 183–186 reporting patterns for, 115, 118
public judgment vs., 35–36, 40, 145 statistics on, 29
public safety short-term trends, 113–114, 116–118
crime control policy trade-off between, 8, three-strikes law impact on, 123
143, 170, 224–225 rationality, in punishment, 198
as Model Penal Code theme, 170, 177 Bentham’s rules for, 189–190
punishment(s) Reagan, Ronald, 17, 39
Bentham’s rational rules for, 189–190 real offense sentencing
communicative theories of, 159–160 attorneys’ power with, 149–150
community service as, 158, 163–166 constitutionality of, 149
cost-benefit analysis of federal guidelines for, 148–149
accounting of offenders, 171–172, issues with, 147–148
189–194 objectives of, 146–147
Bentham’s rational rules for, 189–190 rejection of, 148, 150
fiscal vs. opportunity, 41, 191 recidivism
measurement methods, 170–171 formerly unthinkable policies on,
preventive strategies, 191–192, 144–146
224–225 imprisonment influence debate on,
crime patterns/rates impact on, 5, 14 186–187
fairness of, 13, 202 lack of interest in, 170–171
with imposition of disproportionate, measurement methods for, 170, 186
141–143 prediction of, 49, 154
individualized vs., 58–59, 70–71 proportional punishment for, 143
in real offense sentencing, 147–148 research perspectives for, 186–189
fines as, 164–166 reconsideration, of sentencing, Model Penal
gender-neutral, 154 Code provisions for, 181
impact on crime rates, 11–12, 14, 40 Red Scares (1920s), 78
moral-educative theories of, 157, 160– reform initiatives
162, 166 on drug laws, 16–17, 20, 76, 218–219
nonincarcerative, 158, 163–166 (see also on policy making, 95, 136, 179, 193
shaming) on preventive detention, 170, 172–173
normative theories of, 141, 157–160, 162, 166 on sentencing, 95, 136, 213–227
postmodernist perspectives of, 51–56, state-based, 211, 217–218, 220
196–200 rehabilitation
professional blueprint for (see Model Pe- effectiveness of, current state of knowl-
nal Code) edge on, 203, 218
proportionality principle of, 141–142, of juvenile offenders, 105, 151
170, 176, 180 long-term crime trends and, 105
proposals for ameliorating excesses of, Model Penal Code consideration of, 177,
19–20 179
INDEX 257
[rehabilitation] (continued) San Diego, comparative homicide rates in,
public opinion on, 34, 36 121–122
risk society analysis for, 48–49 Scandinavia
shaming vs., 158, 161, 164 Durkheimian punishment policies of, 161
situational crime prevention vs., 48 long-term crime trends in, 12–14,
reintegrative shaming, 162–163 102–104
relevant conduct, in real offense sentencing, scapegoats, criminals as, 7, 24
149 Schmoke, Kurt, 133
religious pluralism, recurring cycles of school prayer, politics of, 45–46
prohibitionistic, 110 security firms, fees for, 191
tolerance vs. intolerance, 6–7, 10, 78–79, selective incapacitation, of juvenile offend-
100 ers, 153–156
truths and trivia about, 64–66 self-awareness, criminal justice system’s
remediation, for policy changes, 213–227 need for more, 199–204
amendment of mandatory laws, 217 self-defense behaviors, of culture, 90
disparity audits, 225–227 self-interest
goals of, 199, 213–214 of American politicians, 3, 9–10, 17, 19
impact analyses for women and ethnic in career civil service, 207–208, 210
groups, 221–225 impact on crime policy, 23, 38–41, 76,
impact projection requirements, 219–220 136, 196–198, 202
for minority groups, 214–216 morality debates on, 40–41, 46
repeal of mandatory laws, 216–217 sensibilities
resource and appropriation provision re- as “civilizing process,” 71–72, 107
quirements, 220–221 collective tasks for changing, 199–200
“safety valve” laws, 215–217 ethical introspection, 200–204
sentencing guideline revisions, 217–218 institutional change, 199–200, 204–213
sunset clauses, 218–219 remediation, 199, 213–227
Reno, Janet, 219 current policies supported by, 7–11,
repeat offending. See recidivism 209–210
reporting, by citizens as formerly unthinkable, 141–146
as crime rate factor, 114–118 as harsh and punitive punishment, 6,
on personal drug use, 131–133 8–9, 12, 16, 96, 133, 196
repressive policies, 196 historical policies supported by, 5–6,
Republican politics, impact on imprison- 63–64
ment rates, 23, 25, 39, 46–51, impact on crime policies, 25, 51, 70–78
196–197 based on understanding and forgive-
resources, sentencing legislation consider- ness, 195, 198–200
ation of in contemporary U.S., 137–139,
authorization provisions, 220–221 195–200
impact projections, 219–220 cycles of, 59–61, 79–84, 134–136
respect, for prisoners drug use perspectives, 130–139
as punishment policy component, 24–25, long-term crime trends and, 100–101,
57, 159 105–109
social status impact on, 57–59 social definition of, 5, 70–71, 85
restitution, as punishment, 164–166 truths and trivia about, 64–69
retribution unthought thought examples of, 169–172
consequentialist view vs., 202–203 sentence appeals, prosecutorial
as Model Penal Code theme, 170, 176, central premise of, 169
185, 193, 202 as commonplace today, 174
retroactivity, of criminal policies, 144–145 in preventive detention debate, 173–174
reverse J-shaped curve, of violence rates, sentencing
104–106, 130 capital (see capital punishment)
Right-to-Life movement, 18 as fragmenting in U.S., 138
“risk society” “going rates” for, 209
historical example of, 90–91 imprisonment as (see prison sentences)
impact on crime policies, 23, 46–51, 218 Model Penal Code provisions for, 170, 177
postmodernist perspectives of, 51–52, 54 authorization of, 180
rite of passage, prison sentences as, 70–71 mitigation of, 181–182
robberies purposes of, 179–180
direct costs of personal, 192 reconsideration of, 181
statistics on, 28–29 real offense, federal guidelines for,
international comparisons, 128–129 146–150
short-term trends, 113–114, 116–118 reform initiatives for, 95, 136, 213–227
three-strikes law impact on, 123–125, sentencing commissions, delegating policy
203 making to, 210–213, 219, 221
Rockefeller Drug Laws (1972), 15, 81, 134 severity, of crimes. See offense severity
romanticism, recurring cycles of, 6, 64, 79– sexual favors, criminal policies on, 144–145
80, 83, 101 sexual psychopaths/predators
Romero, The People v., 17 American periods of concern about,
91–93
“safety-valve” legislation moral panics and, 86, 91–96, 138–139,
for drug offenses, 134 143, 205
as remediation, 215–217 partisan politics of, 40–41
Salem witch trials, as moral panics, 86, registration laws for, 91–93
88–91 lack of support for, 40
258 INDEX
sexual tolerance, cycles of sensibilities on, real offense sentencing and, 148, 150
6, 45, 64–66 Speck, Richard, 138
shaming stare decisis, 174
disintegrative, 156–167 state government, sovereign response to
expressive perspectives of, 157, crime, 41–46
159–163 delegation to specialized agencies, 210–
Kahan’s arguments on, 146, 156–159, 212, 219, 221
163–166 for drug offenses, 133–134
object of, 144, 146 for juvenile crimes, 146, 150–156
public knowledge and opinion in, 158, pessimism about effectiveness of, 134
163–167 postmodernist perspectives of, 52–53
rehabilitation vs., 158, 161, 164 real offense sentencing and, 148, 150
reintegrative, 162–163 sentencing statutes, 125–127
single issue movements, in platform poli- Stephen, James Fitzjames, 176, 186
tics, 18, 44–46, 77 stereotype(s)
situational crime prevention, rehabilitation of criminals, 49–50
vs., 48 of drug abusers, 111–112
slavery, sensibilities support of, 5, 76, 78 of minorities, 214–215
social costs, of crime and punishment, 171– penal policies based on, 96
172, 189–194, 214 stigmatization, 162, 165. See also disintegra-
social hierarchies tive shaming
disadvantaged American strict construction principle, of penal cul-
penal policy as damaging to, 3, 9, 11, ture, 174
214 substantive symmetry, in sentencing, 174
social determinism of, 13, 215, 225 sunset clauses, as remediation, 218–219
punishment policy based on, 24–25, 57, super-ego, on misbehavior, 198
81 supportive reactions, to crime, 162–163
drug use and, 109–112 surveys
in European vs. Anglo-American cul- in public opinion research, 16, 28, 35–36,
tures, 57–58, 225–226 43
for juvenile offenders, 152–155 for victimization rates, 28, 116–118
postmodern perspectives of, 51–54 symmetry principle, for sentencing, 174,
social programs, politics of, 43, 46, 196 180
social value(s) sympathy, impact on crime policy, 201
about right and wrong
drug use related to, 111–112 Taylor, Linda, 39
impact on crime policy, 99–100, 108, technology, people’s beliefs about, 101
198 teenagers, as criminals. See juvenile of-
cycles of changing (see cycles) fenders
impact on penal policies, 209–210 temperance movement, 80, 109
in contemporary U.S., 137–139, Texas
202–204 prison building programs in, 221
evolutionary nature of, 73–75, 77–78 prisoner population caps in, 220
expressive, 159–162, 166, 197 recent crime rates in, 124–125
as harsh and punitive, 6, 8–9, 12, 16, theft
96, 133 as Index Crime, 114
oscillations of, 5, 13, 63–69, 100–101 statistics on, 28–29
long-term crime trends and, 99–100, 108, short-term trends, 113–114, 116–117
111–112 three-strikes-and-you’re-out law, 135
in partisan debates, 40–41, 46 three-strikes law
on personal responsibility American endorsement of, 5–6, 8, 109,
for health and well-being, 132–133 135
partisan politics of, 40–41 as disproportionate punishment, 143,
postmodernist perspectives of, 51–56, 205
196–199 effectiveness of, 11–12, 120, 125
social welfare agencies, role in juvenile impact on crime trends, 109, 119,
crime policies, 151 123–127
socialization judges authority to sidestep, 10, 17, 137
as crime prevention strategy, 107–108, remediative amendment for, 217
161–162 partisan politics of, 40–41, 196–197
Durkheimian framework for, 99, 108 repeal of
sociology arguments for, 19
influence on policy making, 94–96, 189, political reluctance to, 15, 17–18
193 as remediation, 216–217
on origins of crime, 12–13, 17 timeless ideas
risk society analysis in, 46–52 impact on punishment policies, 77–78
postmodernist angst and, 51–56 social clarity on, 78, 142
“soft on crime,” 3, 8, 18, 136 tolerance/intolerance
sovereignty, in response to crime, 41–46 of criminals, 112–130
delegation to specialized agencies, 210– cycles of
213, 219, 221 American examples of, 79–84, 112
with drugs, 133–134 impact on proportionality, 143–144,
by juveniles, 146, 150–156 202
pessimism about effectiveness of, 134 sensibilities impact on, 6–8, 60, 85,
postmodernist perspectives of, 52–53 204
INDEX 259
[tolerance/intolerance] (continued) vigilantism, for sexual offenders, 93
short-term vs. long-term, 98–101 violent crimes
truths and trivia about, 64–69 just punishment for, 203
as timeless idea, 77–78 statistics on, 28–32
“tough on crime” reverse J-shaped vs. U-shaped curves
impact on imprisonment rates, 41, 50 of, 104–106, 130
strategies for being, 4, 13, 15, 18 stereotyping of, 214–215
truth-in-sentencing laws, 8, 106, 119, 135 Virginia
Model Penal Code consistency with, 180, juvenile crime policies in, 146, 150–156
182, 185 sentencing commission in, 211
Virginia, Atkins v., 75
U-shaped curve, of violence rates, 104–106, virtue, as timeless idea, 77
130 voting rights, sensibilities on denial to
understanding, as crime policy basis, 195, women, 6
198–200
Uniform Commercial Code, 178 war(s), impact on crime policy, 55–56, 63,
Uniform Crime Reports (UCR) 67, 78, 109
Index Crimes of, 114 Washington, sentencing reform in, 16, 211,
short-term trends in, 113–114, 116, 118 217, 220
Uniform Determinate Sentencing Law Washington, George, 78–79
(1976), of California, 125–127 Watts, United States v., 149
United States Wechsler, Herbert, 178, 181–183, 185–186
civil servant selection in, 206–207 “wedge issues,” in partisan politics, 23,
crime rate trends in, 12, 14, 27–28, 96 39–41
imprisonment rate and, 27–34 welfare programs
international comparisons, 118–119, government role in failure of, 43
127–130 partisan politics of, 39–40, 46, 196
long-term, 104–109 Welfare Queen, 39
with policy changes, 134–135 well-being, personal responsibility for,
short-term, 113–118 132–133
governance issues (see government) Western countries
imprisonment rates in crime trends in
crime rate trends and, 27–34 international comparisons, 118–119,
vs. European countries, 21–22, 26 127–130
vs. other Western countries, 3, 21, 28, long-term, 104–109
97–98, 128–129, 205 short-term, 119
punishment policies in imprisonment rates in, vs. America, 3,
as egalitarian, 24–25, 27, 57–59 21, 28, 97–98, 119, 128–129
evolution of harsh, 23–61, 73–75, 77– punishment policies of
78, 108, 133, 196 only U.S. as extreme, 205–210
as extreme, 205–210 vs. European, 195–196, 206, 225
as harsh and wasteful, 10–11, 15, 18– Williams, Jerry Dewayne, 203
19, 209 Williams v. New York, 149
moral panics recent impact on, Wilson, Pete, 108, 120, 196
137–139 “windows of opportunity,” moral panics
postmodern perspectives of, 23–24, and, 94–96
51–56, 77, 196–200 witch trials, as moral panics, 6, 84
social status differentiation with, 57– of Salem, 86, 88–91
59, 81 Witte v. United States, 149
tolerance cycles of, 79–84 Wolfish, Bell v., 173
sentencing measurements in, 6, 9, 22 women offenders, disparity analysis for, 20,
changing cycles of, 67–69 221–225
as fragmenting, 138 women’s rights
public opinion on, 34, 36–37 politics of, 45
sexual predator concerns in, 91–93 sensibilities on
three-strikes law in, 5–19, 40–41, 109, radical movements for, 166–167, 221
135, 143, 205 as supporting denial of, 6, 76
United States, Koon v., 137 sentencing legislation consideration of
United States, Witte v., 149 disparity audits on outcomes, 225–227
United States Constitution disparity impact analyses for, 221–225
Eighth Amendment of, punishment appli- wrongful conduct, criminal definition of,
cations of, 67, 75 145
ex post facto principle of, 144, 157
United States v. DiFrancesco, 174 xenophobia, 76
United States v. Watts, 149
utilitarianism, 142 youth. See juvenile offenders
youth gangs, 115, 121
values. See social value(s)
victimization rate zeitgeist, social definition of, 70
short-term trends of, 116–118 zero-tolerance policing
survey tools for, 28, 116 effectiveness of, 8, 11–12
victims’ movement, impact on penal poli-
cies, 167
260 INDEX

Thinking About Crime

Uploaded by

Thinking About Crime

Uploaded by

Thinking About Crime:

Sense and Sensibility in

OXFORD UNIVERSITY PRESS

The American Street Gang:

Isaac Newton, not generally known as a modest man, once,

From capital punishment to three-strikes-and-you’re-out to

1 Sense and Sensibility in American Penal Culture 3

2 Why So Many Americans Are in Prison 21

3 Cycles and Sensibilities 63

4 Moral Panics and “Windows of Opportunity” 85

5 Crime Trends and the Effects of Crime Control

6 Formerly Unthinkable Policies 141

7 Unthought Thoughts 169

8 Better Understanding, People, and Policies 195

Sense and Sensibility

In searching for ways to address rising crime rates in the

4 THINKING ABOUT CRIME

6 THINKING ABOUT CRIME

SENSE AND SENSIBILITY IN AMERICAN PENAL CULTURE 7

8 THINKING ABOUT CRIME

SENSE AND SENSIBILITY IN AMERICAN PENAL CULTURE 9

10 THINKING ABOUT CRIME

12 THINKING ABOUT CRIME

SENSE AND SENSIBILITY IN AMERICAN PENAL CULTURE 13

14 THINKING ABOUT CRIME

SENSE AND SENSIBILITY IN AMERICAN PENAL CULTURE 15

16 THINKING ABOUT CRIME

SENSE AND SENSIBILITY IN AMERICAN PENAL CULTURE 17

18 THINKING ABOUT CRIME

SENSE AND SENSIBILITY IN AMERICAN PENAL CULTURE 19

20 THINKING ABOUT CRIME

Why So Many Americans Are in Prison

American imprisonment rates, more than 700 per 100,000

22 THINKING ABOUT CRIME

WHY SO MANY AMERICANS ARE IN PRISON 23

24 THINKING ABOUT CRIME

Crime Rates and Trends

The first explanation for why so many Americans are in

28 THINKING ABOUT CRIME

WHY SO MANY AMERICANS ARE IN PRISON 29

The imprisonment rate, however, fell through the mid-1970s

30 THINKING ABOUT CRIME

with periods of steep increase alternating with steep declines.

WHY SO MANY AMERICANS ARE IN PRISON 31

German parliament enacted laws discouraging use of short

The second explanation for why American punishment poli-

36 THINKING ABOUT CRIME

Crime and punishment have been high on American political

38 THINKING ABOUT CRIME

WHY SO MANY AMERICANS ARE IN PRISON 39

So what is needed is an explanation for why crime and pun-

WHY SO MANY AMERICANS ARE IN PRISON 41

42 THINKING ABOUT CRIME

WHY SO MANY AMERICANS ARE IN PRISON 43

WHY SO MANY AMERICANS ARE IN PRISON 45

One explanation builds on literatures in sociology and an-

46 THINKING ABOUT CRIME

[T]he contemporary obsession with risk has its roots in

WHY SO MANY AMERICANS ARE IN PRISON 47

So far, so good. Stripped of jargon, there is nothing here

WHY SO MANY AMERICANS ARE IN PRISON 51

52 THINKING ABOUT CRIME

[B]y the 1970s society’s vulnerability to high rates of crime

Garland describes how political campaigns and the mass me-

WHY SO MANY AMERICANS ARE IN PRISON 53

[T]his new element of precariousness and insecurity is

The vulnerability of the middle class is doubly important.

54 THINKING ABOUT CRIME

WHY SO MANY AMERICANS ARE IN PRISON 55

The challenge historian James Whitman set for himself was

58 THINKING ABOUT CRIME

Cycles and Sensibilities

Each of the preceding seven stories helps explain the origins

WHY SO MANY AMERICANS ARE IN PRISON 59

WHY SO MANY AMERICANS ARE IN PRISON 61

Cycles and Sensibilities