Western Punnishment and Ineqaulity

PUNISHMENT AND INEQUALln IN AMERICA

Bruce Western

Russell Sage foundation 0 New York

---

The Russell Sage Foundation, one of the oldest of America's general purpose foundations, was established in 1907 by Mrs. Margaret Olivia Sage for "the improvement of social and living conditions in the United States." The Foundation seeks to fulfill this mandate by fostering the development and dissemination of knowledge about the country's political, social, and economic problems. While the Foundation endeavors to assure the accuracy and objectivity of each book it publishes, the conclusions and interpretations in Russell Sage Foundation publications are those of the authors and not of the Foundation, its Trustees, or its staff. Publication by Russell Sage, therefore, does not imply Foundation endorsement.

BOARD OF TRUSTEES Thomas D. Cook, Chair

Alan S. Blinder Kenneth D. Brody Christine K. Cassel Robert E. Denham Christopher Edley Jr.

John A. Ferejohn

Larry V. Hedges Jennifer L. Hochschild Kathleen Hall Jamieson Melvin J. Konner

Alan B. Krueger Cora B. Marrett Eric Wanner Mary C. Waters

For Lucy, Miriam, and Grace

Library of Congress Cataloging-in-Publication Data

Western, Bruce, 1964-

Punishment and inequality in America / Bruce Western.

p. cm.

Includes bibliographical references and index.

ISBN 978-0-87154-894-8 (cloth) ISBN 978-0-87154-895-5 (paper)

1. Imprisonment-Economic aspects-United States. 2. Imprisonment-Social aspects-United States. 3. Criminal justice, Administration of-Economic aspectsUnited States. 4. Criminal justice, Administration of-Social aspects-United States. I. Tide.

HV9471.W47 2006 365' .973-dc22

2005055259

Copyright © 2006 by Russell Sage Foundation. First papercover edition 2007. All rights reserved. Printed in the United States of America. No parr 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 written permission of the publisher.

Reproduction by the United States Government in whole or in parr is permitted for any purpose.

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences-Permanence of Paper for Printed Library Materials. ANSI Z39.48-1992.

Text design by Genna Patacsil.

RUSSELL SAGE FOUNDATION

112 East 64th Street, New York, New York ID021 109876543

perb comments on the book as a whole and I thank them for their wisdom and generosity. My friend John McCormick indulged many conversations about the main themes and provided helpful comments as the project neared completion. At various stages of the writing process I also received invaluable criticism and assistance from Angus Deaton, David Ellwood, Deborah Garvey, Heather Haveman, Bob Jackman, Christopher Jencks, Jeff Kling, Steve Levitt, Ross Macmillan, Jo McKendry, Abigail Saguy, Rob Sampson, Jeremy Travis, Chris Uggen, and David Weiman. Hillard Pouncy graciously offered his class in Princeton's Woodrow Wilson School as a testing ground for several chapters.

Research for this book was supported by Princeton University, the Russell Sage Foundation, and National Science Foundation grant SES-0004336. Eric Wanner at the Russell Sage Foundation provided sustained support and encouragement, without which much of this research would not have been possible.

Although most of the empirical analysis of the book is new, parts of several chapters were published as "Black-White Wage Inequality, Employment Rates, and Incarceration," American Journal of Sociology 111: 553-78 (co-authored with Becky Pettit); "Incarceration and the Formation and Stability of Marital Unions," Journal of Marriage and the Family 67: 721-34 (coauthored with Leonard M. Lopoo); "Mass Imprisonment and the Life Course: Race and Class Inequality in U.S. Incarceration," American Sociological Review 69: 151-69 (co-authored with Becky Pettit); "The Impact of Incarceration on Wage Mobility and Inequality," American Sociological Review 67: 477-98; and "How Unregulated is the U.S. Labor Market? The Penal System as a Labor Market Institution," American Journal of Sociology 104: 1030-60 (co-authored with Katherine Beckett).

Introduction

In 1831, Alexis de Tocqueville and Gustave de Beaumont were dispatched to America to study the penitentiary, a novel institution generating great discussion among the social reformers of Europe. At that time, two institutions-Auburn State Prison in New York and the Eastern Penitentiary in Philadelphia-offered leading examples of a new approach to the public management of criminals. The institutions were devised for moral correction. Rigorous programs of work and isolation would remedy the moral defects of criminal offenders so they might safely return to society. The penitentiary was billed as a triumph of progressive thinking that provided a humane and rational alternative to the disorderly prisons and houses of correction in Europe. Tocqueville and Beaumont were just two of many official visitors from Europe who toured the prisons in the 1830s, eager to view the leading edge of social reform.

Grand projects in crime control often spring from deep fissures in the social order. Tocqueville and Beaumont saw this clearly, despairing of "a state of disquiet" in French society. Writing in 1833, they traced the need for prison reform to a restless energy in the minds of men "that consumes society for want of other prey."] This moral decline was compounded by the material deprivation of the French working class, "whose corruption, beginning in misery, is completed in prison." Instead of deflecting vice and poverty, the

French prisons made things worse-aggravating immiseration and immorality.² America offered a fresh alternative.

Although the prisons that provided the pretext for Tocqueville's American tour did not figure in his observations on American democracy, democratic aspirations were faintly inscribed on the Auburn and Pennsylvania penitentiaries. The project of rehabilitation assumed an innate moral equality among men that could be restored to criminals through penal discipline. Rehabilitative institutions comprised part of a primitive social democracy that conferred not just the vote and freedom of association but also a minimal equality of life chances. Despite curtailing freedom (and applying corporal punishment), the prison posed no basic threat to democracy because the official ideology of rehabilitation promised to reestablish the social membership of those who had fallen into poverty and crime. In practice, of course, the rehabilitative ideal was regularly compromised and in the South it barely took hold at all. In conception at least, and sometimes in practice, the prison sat comfortably alongside an array of welfare institutions that included not only reformatories and asylums but also public schools, hospitals, and rudimentary schemes for social insurance. Like other welfare institutions, the prison was conceived to rescue the citizenship of the unfortunate, the poor, and the deviant.

The story of this book begins one hundred and forty years later, in the 1970s, when the American penal system embarked on another journey of institutional change. The latest revolution in criminal punishment followed some of the logic of its nineteenth-century predecessor. Shifts in the structure of society and politics forced changes in criminal justice, with large consequences for the quality of American democracy. Through the last decades of the twentieth century, the patchwork system of American criminal justice turned away from the rehabilitative project first attempted in New York and Pennsylvania. By the 1970s, policy experts were skeptical that prisons could prevent crime by reforming their inmates. Incarceration would be used less for rehabilitation than for incapacitation, deterrence, and punishment. Politicians vowed to get tough on crime. State lawmakers abandoned the rehabilitative ideals etched in the law of criminal sentencing and opted for mandatory prison terms, the abolition of parole, and long sentences for felons on their second and third convictions. Tough new sentences were attached to narcotics offenses as the federal government waged first a war on crime, then a war on drugs. Locked facilities proliferated around the country

INTRODUCTIC

to cope with the burgeoning penal population. Prison construction bee an instrument fGt^r regional development as small towns lobbied for COl tional facilities and resisted prison closure.

Prisons themselves changed as a result of the punitive turn in criminal rice. Budgets tightened for education and work programs. But some sc service function remained as the penal system assumed new responsibil for public health, delivering treatment on a large scale for mental illness, berculosis, HIV/ AIDS, and hepatitis C. High-risk inmates were gathem supermax facilities that placed entire prison populations in solitary conf rnenr. In a thousand ways, large and small, the democratic aspirations 01 habilitative corrections were erased and the coercive power of the state pc trated more deeply into the lives of the poor.

Most striking was the increase in the size of the correctional populati Between 1970 and 2003, state and federal prisons grew sevenfold to ho 1.4 million convicted felons serving at least one year behind bars, and 1) cally much longer. Offenders held in county jails, awaiting trial or serv short sentences, added another seven hundred thousand by 2003. In ad tion to the incarcerated populations, another 4.7 million people were urn probation and parole supervision. The entire correctional population of 1 United States totaled nearly seven million in 2003, around 6 percent of 1 adult male population.s

Growth in the penal population signaled more than a change in pub policy. Throughout the twentieth century, African American history l~ been entwined with the history of America's prisons. Blacks have been me likely than whites to go to prison, at least since the 1920s. Southern priso operated quite transparently as instruments of racial domination, usir forced labor to farm cotton and build roads." The prison boom, growiJ quickly in the wake of the civil rights movement, produced a wholly ne scale of penal confinement. The basic brute fact of incarceration in the ne era of mass imprisonment is that African Americans are eight times rno likely to be incarcerated than whites. Incarceration rates climbed to extrao dinary levels among young black men, particularly among those with litt schooling. The Bureau of Iustice Statistics reports that in 2004, over 12 pel cent of black men aged twenty-five to twenty-nine were behind bars, i prison or jaiP Among black men born in the late 1960s who received n. more than a high school education, 30 percent had served time in prison b their mid-thirties; 60 percent of high school dropouts had prison records.

lNlIZODU

By the end of the 1990s, criminal justice supervision was pervasive among young black men. This was a historically novel development in American race relations. We need only go back thirty years, to 1970, to find a time when young black men were not routinely incarcerated. The betrayal of the democratic purpose of rehabilitation had diminished the citizenship of African Americans most of all.

How can we understand the fabulous growth in the American penal system and its effects on the poor and minority communities from which prison inmates are drawn and ultimately return? This book first details the changing scope of incarceration in America through the 1980s and 1990s, then accounts for the growth in incarceration rates. I then examine the effects of the prison boom on crime, and economic opportunity and the family life of the men who serve time in prison jail.

My main arguments rely on two basic insights of the sociology of politics and crime. First, for political sociology, state power flows along the contours of social inequality. From this perspective; the prison boom was a political project that arose partly because of rising crime but also in response to an upheaval in American race relations in the 1960s and the collapse of urban labor markets for unskilled men in the 1970s. The social activism and disorder of the 1960s fueled the anxieties and resentments of working-class whites. These disaffected whites increasingly turned to the Republican Party through the 1970s and 1980s, drawn by a law and order message that drew veiled connections between civil rights activism and violent crime among blacks in inner cities. For these conservative politics, rehabilitation coddled the criminals who had forfeited their rights to fairness and charity. The young black men of poor urban neighborhoods were the main targets of this analysis. Jobless ghettos, residues of urban deindustrialization, lured many young men into the drug trade and left others unemployed, on the street, and exposed to the scrutiny of police. The punitive sentiment unleashed in the 1970s by rising crime and civil rights activism in the 1960s, institutionalized what had become a chronically idle population of young men with little education. Their life path through adulthood was transformed as a result.

Second, for sociologists of crime, the life path through adulthood normalizes young men, so criminal behavior recedes with age. Adolescents are drawn into the society of adults by passing through a sequence of life course stages-completing school, finding a job, getting married, and starting a family. The integrative power of the life course offers a way out of crime for

adult offenders. Men involved in crime who can find steady work ar ble marriage also become embedded in a web of social supports and tions. These social bonds help criminally active men desist from fur fending. Men coming out of prison, however, have little access to th jobs that usually build work histories and wages. Employers are relu. hire job seekers with prison records, and former inmates are generallj prepared for the routines of steady employment. Prison also disrupi lies. By 2000, over a million black children-9 percent of those un de een-had a father in prison or jail. In around half of all cases, these were living with their children at the time they were incarcerated. The separation of men from their families also takes a toll on conjugal bon women with men in prison, married life is threatened by the strains 0, tion and the temptations of free men who can help support a hou Few couples survive a term of imprisonment. Unmarried men stigmat a prison time can also pay a price. Serving time signals a man's unrel and a prison record can be as repellent to prospective marriage partne is to employers.

A common logic underlies the negative effects of incarceration on mer inmate's job prospects and family life. Although the normal life co integrative, incarceration is disintegrative, diverting young men from t stages that mark a man's gradual inclusion in adult society.

The employment problems and disrupted family life of former if suggests that incarceration may be a self-defeating strategy for crime cc Although incarceration surely prevents those who are locked up from mitring crime in society, inmates are ultimately released with few res: to lead productive lives. Without great hopes for job security or a good riage, crime remains an inviting alternative. Skeptics will counter through the 1990s, when incarceration rates reached their highest J crime rates fell to their lowest levels since the 1960s. Correlation, howe not causation. There were many forces operating at the end of the 195 drive down crime rates. My empirical analysis shows that fully 90 perce the decrease in serious crime from 1993 to 2001 would have happened without the run-up in the incarceration rates. The prison boom contril a little to the decline in crime through the 1990s, but this gain in p safety was purchased at a cost to the economic well-being and family li poor minority communities.

Even more important than the effects of the prison boom on crime a

G PUNISHMENT AND INEQUAlITY IN AMERICA

effects on American inequality. The repudiation of rehabilitation and the embrace of retribution produced a collective experience for young black men that is wholly different from the rest of American society. No other group, as a group, routinely contends with long terms of forced confinement and bears the stigma of official criminality in all subsequent spheres of social life, as citizens, workers, and spouses. This is a profound social exclusion that significantly rolls back the gains to citizenship hard won by the civil rights movement. The new marginality of the mass-imprisonment generation can be seen not only in the diminished rates of employment and marriage of former prisoners. Incarceration also erases prison and jail inmates from our conventional measures of economic status. So marginal have these men become, that the most disadvantaged among them are hidden from statistics on wages and employment. The economic situation of young black men-measured by wage and employment rates-appeared to improve through the economic expansion of the 1990s, but this appearance was wholly an artifact of rising incarceration rates.

To tell this story, I begin by charting the scope of the prison boom. Chapter 1 places the era of mass imprisonment in comparative and historical perspective, underlining the historic novelty of the current period. Chapter 2 explores the causes of the prison boom by relating the growth in incarceration rates to shifts in crime rates. I see little evidence that growth in the penal population is related to either rising crime, or that increased incarceration among young disadvantaged men is associated with increased offending. Chapter 3 continues the search for the causes of rising imprisonment by studying changes in economic and political conditions. Incarceration rates grew most in states that elected Republican governors and adopted punitive regimes of criminal sentencing. Analyzing rates of prison admission for black and white men at different levels of education shows that class inequalities in imprisonment increased as the economic status of less-educated men decreased.

The remaining four chapters study the consequences of the prison boom.

Links between the labor market and the penal system are examined in chapter 4 that measures the hidden inequality in wages and unemployment due to high rates of incarceration. I find that young black men obtained no benefit-either in employment or relative wages-from the record-breaking economic growth in the late 1990s. The invisible inequality that burgeoned through the boom times of the 1990s challenges the claim that robust

INTRODUCTIC

growth by itself, without the supports of social policy, could bring oppc uity to the most disadvantaged. Chapter 5 follows prison and jail inrr from release into society to their experiences in the labor market. Su analysis shows that incarceration significantly reduces the wages, emp rnent, and annual earnings of former inmates, even though their econo opportunities are extremely poor to begin with. The family life of crirn offenders is studied in chapter 6 which analyzes marital disruption and rnestic violence among men coming Out of prison. Here I find that incar. ation undermines marital relations and thus increases a woman's risk of, lence at the hands of her partner. Finally, chapter 7 tests the claim that prison boom drove the fall in crime at the end of the 1990s. I find that large negative effects on crime that are often attributed to imprisonment overstated: The growth in incarceration rates explains only one-tenth of decline in serious crime at the end of the 1990s.

Although the prison boom undermined economic opportunity and SI up families, it cannot explain all the unemployment and female-heao households that underpin much of America's racial inequality. Unernpl, rnent and broken homes are as much a cause of imprisonment as a con; quence. The disadvantaged men who go to prison would still risk unemplc ment and marital instability even if they weren't incarcerated. Instead, t prison boom helps us understand how racial inequality in America was Sl tained, despite great optimism for the social progress of Mrican Americar From this perspective, the prison boom is not the main cause of inequali between blacks and whites in America, but it did foreclose upward rnobili and deflate hopes for racial equality. Perhaps more than adding to inequali between blacks and whites, the prison boom has driven a wedge into d black community, where those without college education are now travelin a path of unique disadvantage that increasingly separates them from colleg( educated blacks.

The prison boom opened a new chapter in American race relations, bu the story of race and class inequalities sustained by political institutions is aJ old one. The punitive turn in criminal justice disappointed the promise 0 the civil rights movement and its burdens fell heavily on disadvantage< Mrican Americans. By cleaving off poor black communities from the main stream, the prison boom left America more divided. Incarceration rates an now so high that the stigma of criminality brands not only individuals, bui an entire generation of young black men with little schooling. Tocqueville

8 PUNISHMENT AND INEQUALITY IN AMERICA

and Beaumont might be surprised that the American prison had failed so completely to realize the promise of its democratic origins. Although the growth in imprisonment was propelled by racial and class division, the penal system has emerged as a novel institution in a uniquely American system of social inequality.

The Scope and Causes of the Prison Boom

CHAPTER 1

Mass imprisonment

If prisons affected no one except the criminals on the inside, they would matter less. But, after thirty years of penal population growth, the impact of America's prisons extends far beyond their walls. By zealously punishing lawbreakers-including a large new class of nonviolent drug offenders-the criminal justice system at the end of the 1990s drew into its orbit families and whole communities. These most fragile families and neighborhoods were the least equipped to counter any shocks or additional deprivations.

We normally relate the prison boom to the problem of crime in America.

Some say that we have more prisoners today because there is more crime. Others say that crime rates have fallen because we've locked up so many dangerous criminals. This book studies the prison boom, but crime is not the main focus. I argue that the prison boom is significant, mostly for its effects on social inequality. Indeed, the penal system has become so large that it is now an important part of a uniquely American system of social stratification.

This is an extravagant claim in many ways. In any given year in the last century, only one in a thousand Americans could be found in prison. Even at the height of the prison boom, in the early 2000s, less than 1 percent of the U.S. population was behind bars. These tiny incarceration rates should not be surprising: Prisons and jails are criminal justice institutions. Their constituents are the small number of criminals who break the law, not the vast

majority of law-abiding citizens. If we are interested in the institutions that affect inequality in America, perhaps we should look at schools, or labor unions, or welfare programs.

By 2000, however, the U.S. incarceration rate was unparalleled in the economically developed democracies and unprecedented in U.S. history. Although prison and jail inmates are only a small fraction of the entire population, this chapter will show that the prison boom transformed the institutional landscape traveled by poor black males as they grew out of childhood and became young adults. Imprisonment became commonplace among young black men, more common than military service or college graduation. For black men who dropped out of high school, prison time became a modal event, more common than not. The concentration of imprisonment among young black men, particularly those with little schooling, provides the first piece of evidence for the generalized institutional significance of the American penal system. Empirical evidence for large-scale incarceration justifies the term mass imprisonment-an incarceration so vast as to draw entire demographic groups into the web of the penal system.

INCARCERATION IN HISTORICAL AND COMPARATIVE PERSPECTIVE

Before the prison boom, incarceration was the backstop of the criminal justice system. After school suspension, juvenile hall, warnings from police, arrest, commitment to the adult courts, conviction, and probation, came the county jail and then state prison. The many layers of criminal punishment ensured prison was rarely used, and only then for violent offenders or career criminals who cycled in and out of jail.

The penal system itself is divided among local, state, and federal jurisdictions. County jails account for about a third of the penal population. Jails hold defendants awaiting trial and misdemeanor offenders serving less than a year. John Irwin describes the jail as an instrument for managing "the rabble," mostly disreputable petty offenders who live under the close eye of the police. ¹ State and federal prisons-home to about two-thirds of the penal population-typically hold felony offenders serving a year or more. Most prisoners are serving time for violent, property, or drug crimes. Nine out of ten prison inmates are housed in state facilities. One-third of these, in 1997, had committed homicide, rape, or robbery, and the remainder are mostly property and drug offenders. In the federal system, three out of five prisoners

Source: Maguire and Pastore (1996, table 6.22); Beck and Glaze (2004).

Note: Incarceration rates are shown on the left-hand axis. The prison population is sho the right-hand axis.

by 1997 were drug offenders.² Nearly all prisoners serve at least one yea! most serve much longer sentences. In 1996, state drug offenders aver just over two years in prison, compared to eleven years for murderers. In eral prison the same year, the average time for drug offenders was months."

The great scale of the penal system in the early 2000s is new. On an} for fifty years from 1925 to 1975, about a hundred Americans out of a J dred thousand-just one-tenth of 1 percent of the U.S. population-we prison (figure 1. 1). From 1975, the imprisonment rate began to rapidl crease. By 2003, the share of the population in prison had increased e year for twenty-eight years, standing at nearly half of 1 percent at the be ning of the new century. If we add jail inmates to the count of the inca) ared population, seven-tenths of 1 percent of the u.s. population was lo. up by 2003. This incarceration rate reflects a penal population of 2.1 mil inmates. After more than a quarter of a century of growth, the scale of in ceration exceeded its historic average by a factor of nearly five.

The extent of incarceration in the United States is also unusual by if

national standards. In 1983, the rate of 275 per hundred thousand was about four times higher than in western Europe (figure 1.2). Only Britain's penal population approached American levels, and even in this case the u.s. rate was more than twice as high. By 2001, the imprisonment gap between Europe and the United States had widened. The U.S. incarceration rate had

MASS IMPRlSONME

climbed to 686 per hundred thousand, and European incarceration rai mained close to 1983 levels-around a hundred per hundred thousan less. In 2001, Britain still recorded the highest incarceration rate in WI Europe, but the American imprisonment rate was more than five greater. Indeed, to find close competitors to the American penal syste must look beyond the longstanding democracies of western Europe, to sia (628 per hundred thousand) and South Mrica (400).4

INEQUAUTY UN INCARCERATION: SI1D{u AGE, RACE, AND EDUCATION

By 2000, the U.S. incarceration rate was comparatively and historically but the scale remained small in absolute terms. Even at the height 0 prison boom, less than 1 percent of the population was incarcerated. Ca institutionalization of this size possibly have large effects?

The broad significance of the penal system for American social inequ results from extreme social and economic disparities in incarceration. J, than 90 percent of all prison and jail inmates are men, and throughout book I focus on men's incarceration. Women's incarceration rates have creased more quickly than men's in the twenty years after 1980, but the 11 effect of the prison boom on gender relations is due precisely to the appr mate fact that men go to prison, and women are left in free society to I families and contend with ex-prisoners returning home after release. In ceration is also concentrated among the young. About two-thirds of s prisoners are over eighteen years old but under thirty-five. With this age I tern, only a small number of people are incarcerated at any time, but m more pass through the system at some point. Age and sex disparities in in! ceration magnify the influence of the penal system. In a gendered world! ours, institutions that shape the lives of men or women alone have great 0 sequences for the other sex. The effects of institutions that entangle yOl adults may be sustained over a lifetime.

Gender relations and the life course amplify the effects of a penal syst, that locks up mostly young men, but race and class disparities in incarce tion are significant for inequality in another way. Incarceration is conce trated among the disadvantaged and large race and class disparities in imp: onmenr reinforce lines of social disadvantage. High incarceration ra among less educated, less skilled, financially disadvantaged, and minor; men are unmistakable. The 1997 survey of state and federal prisoners sho-

16 PUNISHMENT AND INEQUALITY IN AMERICA

that state inmates average fewer than eleven years of schooling. A third were not working at the time of their incarceration, and the average wage of the remainder is much lower than that of other men with the same level of education. African Americans and Hispanics also have higher incarceration rates than whites. Blacks and Hispanics together account for about two-thirds of the state prison population. The black-white disparity in imprisonment is especially large. Black men are six to eight times more likely to be in prison than whites.

The demographic contours of imprisonment produced large differences in incarceration rates across the population (table 1.1). Through the last two decades of the twentieth century the national incarceration rate of the United States grew from about one-fifth of 1 percent of the population to seven-tenths of 1 percent. Because nearly all prison and jail inmates are men of working age, the incarceration rate in this group is nearly three times the national average. Incarceration rates for minority men are much higher. By 2000, more than 3 percent of Hispanic men and almost 8 percent of African American men of working age were in prison or jail.

The black-white difference in incarceration rates is especially striking.

Black men are eight times more likely to be incarcerated than whites and large racial disparities can be seen for all age groups and at different levels of education. The large black-white disparity in incarceration is unmatched by most other social indicators. Racial disparities in unemployment (2 to 1), nonmarital childbearing (3 to 1), infant mortality (2 to 1), and wealth (1 to 5) are all significantly lower than the 8 to 1 black-white ratio in incarceration rates." If white men were incarcerated at the same rate as blacks there would be more than six million people in prison and jail, and the incarceration rate would include more than 5 percent of the male working-age population.

Age, race, and educational disparities concentrate imprisonment among the disadvantaged. White men aged twenty to forty saw their incarceration rates rise from .6 to 1.6 percent between 1980 and 2000. The incarceration rate for young Hispanic men in 2000 was three times higher. Large blackwhite differences in the incarceration rate can also be seen for men under age forty. Three out of every two hundred young white men were incarcerated in 2000, compared to one in nine young black men. Incarceration of the poor is deepened by the severe educational disadvantage of prison and jail inmates. Among young men who had never been to college, 5.5 percent of Hispanic and 17 percent of black men under age forty-one were in prison or more likely to be incarcerated than their college-educated counterparts. Twenty years later, the difference was more than eight times. In 2000, one in three black dropouts were locked up, compared to just one in twenty-five of their college-educated counterparts.

In sum, disparities in incarceration produced astonishing rates of penal confinement among less-educated and minority men. Among the most socially marginal men-Mrican Americans in their twenties and thirties who had dropped out of high school-incarceration rates were nearly fifty times the national average.

INCARCERATION AS INSTITUTIONAL MEMBERSHIP

Studies of social inequality usually cast working-age men in the roles of worker or job seeker. For workers and job seekers, the key institutional influences on economic well-being are labor unions and the welfare state. By raising the wages of union members and paying out unemployment benefits, labor organization and social policy significantly modify and augment the distribution of rewards in the labor market. Prisons and jails, on the other hand, aren't usually treated as labor market institutions. Instead, they lie on the horizon of social life, separating the deviant few from the mainstream. Because prisons and jails regulate deviance, incarceration is usually thought to mark a criminological, not an economic, status. In any case, formal controls on deviance have not been extensive enough to broadly influence economic opportunity. The scale of imprisonment in the 1990s challenges us to view penal institutions as significant economic influences on the men who serve time in prison or jail. Is the reach of the penal system sufficiently extensive to justify its comparison to other institutions that we normally see as shaping young men's labor market experiences?

I address this question by comparing involvement in the penal system to labor union membership and enrollment in government social programs, and later study the economic effects of imprisonment more directly. To begin, however, if men's involvement in the penal system rivals their involvement in unions and social programs, we are challenged to count incarceration as a major institutional presence in the economic lives of young men.

union membership and participation in government welfare and other soc programs. Although unions and social programs are important institutio for young whites, their reach is at least equaled by the penal system for you blacks. Among whites, union membership (9.7 percent) and social progra participation (6.7 percent) are far more common than incarceration (1.6 pc cent). Hispanic men show a similar pattern, though they are less likely participate in government programs. Black men under age forty have an iJ carceration rate of 11.5 percent, and are just as likely to be in prison or jail in a labor union, and about twice as likely to be incarcerated, as to receigovernment benefits. High school dropouts participate at high rates in we fare and other government programs, but white and Hispanic dropouts aJ at least as likely to be incarcerated as to be in a union or receive some assi: tance from a welfare program. Their black counterparts also participate healily in government programs, but a third are in prison or jail. The incarcera tion rate for these blacks exceeds any support they receive for health care training, and income maintenance through the welfare state.

20 PUNISHMENT AND INEQUALITY IN AMERICA

These figures indicate that by the end of the 1990s, penal confinement had become common for African American and less educated men, compared to involvement in labor unions and social programs. The economic institutions that we normally associate with the welfare of young men was being eclipsed in the era of the prison boom, especially among the most disadvantaged-African American youth who had dropped out of high school.

INCARCERATION OVER THE LIFE CYCLE

Incarceration rates offer a snapshot of the extent of penal confinement. Time series of incarceration rates tell us how the extent of penal confinement has shifted historically. We can also study not the level of incarceration at a particular time, but how the risk of incarceration accumulates over an individual's life. This kind of life course analysis asks what is the likelihood an individual will go to prison by the time he is twenty-five, thirty, or thirty-five. Instead of providing a snapshot of the risk of incarceration, the life course analysis tries to characterize a typical biography.

The life course perspective provides more than just a way of thinking about the risks of incarceration; it also provides a comprehensive social analysis. For students of the life course, the passage to adulthood is a sequence of well-ordered stages that affect life trajectories long after the early transitions are completed. Today, arriving at adult status involves moving from school to work, then to marriage, to establishing a home and becoming a parent. Completing this sequence without delay promotes stable employment, marriage, and other positive life outcomes. The process of becoming an adult thus influences success in fulfilling adult roles and responsibilities.

As an account of social integration, life course analysis has attracted the interest of students of crime and deviance.v Criminologists point to the normalizing effects of life course transitions. Steady jobs and good marriages build social bonds that keep would-be offenders in a daily routine. They enmesh men who are tempted by crime in a web of supportive social relationships. Strong family bonds and steady work restrict men's opportunities for antisocial behavior and offer them a stake in normal life. For persistent lawbreakers, the adult roles of spouse and worker offer a pathway out of crime'? Those who fail to secure the markers of adulthood are more likely to persist in criminal behavior. This idea of a normalizing, integrative, life path offers a powerful alternative to claims that criminality is a stable trait present in

MASS IMPRlSONMEl'-

some, but absent in others. Above all else, the life-course account of ern dynamic, describing how people change as their social context evolves age.

Imprisonment significantly alters the life course. In most cases, me! tering prison will not be following typical life trajectories. Time in juv incarceration and jail and weak connections to work and family divert r prison inmates from the usual path followed by young adults. Spells ol prisonment-thirty to forty months on average-further delay entry the conventional adult roles of worker, spouse, and parent. Diversions j the normal life course are not always negative. Military service, for exan has been identified as a key event that redirects life trajectories. Glen I describes military service as a "legitimate timeout" that offered disad taged servicemen in World War II an escape from family hardship." Simi] imprisonment can provide a chance to reevaluate life's direction." Typic however, imprisonment has negative effects. In contrast to the legirir time out of military service, imprisonment is an illegitimate timeout confers an enduring stigma. Employers of less-skilled workers are reluc to hire men with criminal records. The stigma of a prison record also ere legal barriers to skilled and licensed occupations, rights to welfare bene and voting rights.t? Later chapters will show that ex-prisoners earn lc wages and suffer more unemployment than similar men who have not l incarcerated. Former prisoners are also less likely to get married or live, the mothers of their children. By eroding opportunities for employment marriage, incarceration may also lead former inmates back to a life of cri The volatility of adolescence may last well into midlife for men sen prison time. In short, imprisonment is a turning point to fewer opportl ties and attenuated citizenship. The life course significance of incarcerai motivates analysis of the evolving probability of prison incarceratior young men age through their twenties and thirties.

THE HISTORICAL SIGNIfiCANCE Of THE PRISON BOOM FOR THE UIFE COURSE

Biographies unfold in particular historical contexts. To choose a famous ample, boys growing up through the Great Depression started their work lives young, in adolescence, to help support their families. Having seen depredations of mass unemployment, they valued economic security, of at the expense or more lucrative employment in later life. They also dela

22 PUNISHMENT AND INEQUALITY IN AMERlCA

marriage and fatherhood as they struggled to establish themselves economically before starting a household. The imprint of history on this birth cohort of depression era boys makes a generation-a cohort of children whose collective coming of age is shaped decisively by historic forces of social change. I I

These youth were to become the "Greatest Generation," as familiar in the public imagination as they were to professional demographers. World War II drew nearly all of the able-bodied among them into the military. For those from poor families or with histories of delinquency, military service was a turning point. As servicemen, the children of the Great Depression often received additional schooling, and those who survived the war with mind and body intact could also take advantage of the GI Bill. The GI Bill massively subsidized the collective mobility of the American working class, through its support for college education and home ownership. After the war, even the most needy and troublesome youth who attended school under the GI Bill would come to enjoy good jobs and higher incomes as they moved into midlife. The Greatest Generation, forged as much by the GI Bill as wartime, escaped the constraints of family background and personal history to share in the great social and economic benefits of the first decades of the postwar period.t-

Throughout the twentieth century, history has left its mark on generations through great programs in social improvement. The GI Bill is the leading example, but the hundred-year emergence of mass public education also transformed the passage to adulthood. For successive cohorts since the 1900s, the expansion of public education contributed to an increasingly orderly and compressed transition to adulrhood.P We might also think of a Civil Rights Generation, African Americans growing up after school desegregation, and under the umbrella of antidiscrimination protections. These black men and women, growing up through the 1960s and 1970s, enjoyed great gains in schooling and employment, significantly closing the gap with whites.f These examples show how individual lives, confronted with the transformative force of military service and education, have been redirected to produce significant episodes of collective mobility.

The prison boom, too, can be viewed as a major social change that has reordered the biographies of those growing up through the 1980s and the 1990s. In the historic context of the prison boom, incarceration has reshaped adulthood for entire birth cohorts. In this way, the growth of America's prisons is similar to other social transformations that precipitated major sh life trajectories.

Of course, prison time is not chosen in the same way as school arten or military service. Men must commit crime to enter prison. Despit qualification for imprisonment, the penal system has no necessary mon over young men involved in crime. A variety of institutions compete f risdiction over the life course. 15 Criteria for entry into prison, the milir, school are historically variable. During World War II, the scale of the war effort ensured that all able-bodied young men were potential se men, and most were drafted. As the number of college places expander ing the 1960s and 1970s, young men became potential college stu qualifYing less on the basis of social background, and more througb demic achievement. The prison emerged through the 1980s and 1990, major institutional competitor to the military and the educational syste least for young black men with little schooling. Much more than for cohorts, the official criminality of men born in the late 1960s was c mined by race and class.

In the past, going to prison was a marker of extreme deviance, reserve violent and incorrigible offenders. Just as the threshold for military se was lowered during World War II, the threshold for incarceration was ered by the forces driving prison boom. We will see how incarcerarior came more common for convicted felons, and how the criminalization 0 drug trade swept up large numbers of small-time offenders. These tr, suggest the novel normality of criminal justice sanction in the lives of re cohorts of disadvantaged minority men. Richard Freeman, for exarr writes that "participation in crime and involvement in the criminal ju: system has reached such levels as to become part of normal economic lift many young men." 16 John Irwin and James Austin echo this observat "For many young males, especially African Americans and Hispanics, threat of going to prison or jail is no threat at all but rather an expectec accepted part of life."] ? David Garland similarly observes that for "yo black males in large urban centers ... imprisonment ... has come to I regular, predictable part of experience. "18 All these claims of pervasive prisonment suggest a wholly new experience of adult life for recent cob of disadvantaged, minority men.

The widely claimed significance of mass imprisonment in the lives young African American men suggests two hypotheses. First, imprisonm

L4 I^JUNl.) HM.t.l'l! AI'IU 1l'ltI...(.Uf\Ll1 r lr~ l\1VICru\.../\

by the 1990s became a modal life event for young black men with little education. Second, the prevalence of imprisonment among African American men in the 1990s rivals in frequency more familiar life events such as military service and college graduation.

LIfETIME RISKS OF IMPRISONMENT

To place the risks of imprisonment in the context of the life course, I calculated the likelihood that a man would go to prison by age thirty-five. Imprisonment by that age provides a good estimate of the lifetime risk, because very few are incarcerated for the first time after their mid-thirties. Although a number of different incarceration statistics are reported in tables 1.1 and 1.2, it is important to remember that the following figures describe the deep end of the penal system for which there are lengthy terms of confinement for a felony conviction. By focusing on prison and bracketing jail, these figures understate the full reach of the penal system. 19

Two birth cohorts of black and white men are contrasted to judge the effects of the prison boom. The older cohort is born just after World War II, from 1945 to 1949, and reaches their early thirties in the late 1970s. The older cohort passes through their twenties and early thirties before the most rapid increase in imprisonment rates. The younger cohort is born during the Vietnam War, from 1965 to 1969, and reaches their early thirties at the height of the prison boom. How much have the risks of imprisonment changed from one birth cohort to the other?

I answered this question using life-table methods that calculate the probability that a man with a clean record will go to prison for the first time at age twenty, then at twenty-one, twenty-two, and so on. Adding up these probabilities at each age, and adjusting for mortality, gives an estimate of the probability that a man has ever been to prison by age thirty-five. To make these calculations, I assembled a variety of data sources, including prison inmate surveys, administrative data on state and federal prison populations, household survey data on the noninstitutional population, and vital statistics on mortality. More information about the analysis and some checks on the quality of the estimates are reported later this chapter.

Life table estimates show how the risk of ever being imprisoned grows as men get older (figure 1.3). For white men born in the late 1940s, the probability of going to prison for the first time by age twenty was relatively small, less than half of 1 percent. By the time they were in their late twenties, this probability had climbed to over 1 percent, that is, just over one in a hund Growth in the risk of imprisonment slows a little by the time these men" in their mid-thirties to 1.4 percent. For the younger birth cohort, born in late 1960s and growing up through the prison boom, the risks of impris ment until age twenty-five are similar to those of the older cohort. The ef of the prison boom on the path through young adulthood can be seen fr age twenty-five to thirty-five. From their mid-twenties onwards, white n who have never been to prison became much more likely through the 19 to serve time. Indeed, by their early thirties, the cumulative risk of impris ment in the younger cohort of white men was more than double that those born twenty years earlier.

The changing risk of imprisonment over the life course is strikingly pa lei among black men. The imprisonment risk is so much higher than for earlier generation, in fact, that we might of think of the prison boom transforming young adulthood for black men. For black men born in the I

1940s, around one in ten served time in prison by their early thirties, in 1979. Those born twenty years later were more than twice as likely to go to prison. Changes in imprisonment over the life course help us understand how the adulthood of black men has been transformed by the prison boom. Most obviously, prison has become commonplace for African American men born since the late 1960s. More than twenty percent have spent at least a year-and typically two-locked up for a felony conviction. As for whites, the chances of incarceration have also increased for those in their early thirties. This may be because repeat offenders, on their second and third convictions, became more likely to be sentenced to prison. The growing likelihood of imprisonment for those entering midlife suggests that incarceration has become more disruptive. As they reach their late twenties most men are beginning to establish and support a household. For black men born in the late 1960s, much more than for their counterparts born twenty years before, this pathway to manhood is increasingly blocked as the penal system reaches deeper into the life course.

Calculating lifetime risks of imprisonment for high school dropouts, graduates, and the college-educated shows how the lives of the disadvantaged have been changed by rising incarceration rates. Imprisonment has become relatively common for white high school dropouts (figure 1.4). By 1999, one in nine would go to prison by their early thirties. The lifetime risks of imprisonment decline as we go up the educational ladder. White high school graduates were only 3.6 percent likely to go to prison by their early thirties in 1999, less than half the risk faced by white dropouts. College-educated whites were largely spared from the prison boom, their lifetime risk of imprisonment growing from just .5 to .7 of 1 percent from 1979 to 1999.

The cumulative risks of imprisonment for black men at the end of the 1990s are extremely high. Incredibly, a black male dropout, born in the late 1960s had nearly a 60-percent chance of serving time in prison by the end of the 1990s. At the close of the decade, prison time had indeed become modal for young black men who failed to graduate from high school. The cumulative risks of imprisonment also increased to a high level among black high school graduates. Nearly one out of five black men with just twelve years of schooling went to prison by their early thirties. Among all noncollege blacks-the bottom half of the education distribution-nearly a third had gone to prison in the younger cohort compared to just one in eight, two decades earlier.' ? As for whites, virtually all the increase in the risk of imprisonment among blacks fell on those with just a high school education. In fact, my estimates indicate that the lifetime risk of imprisonment actually declined slightly for college-educated blacks through the last decades of the twentieth century.

IMPRISONMENT AND OTHER LIFE-COURSE MILESTONES

Just as we can compare incarceration rates to labor union membership and participation in government programs, the cumulative risk of imprisonment can be compared to other life experiences that mark the transition to adulthood. College graduation, military service, and marriage are all important markers of progress through adult life. Each of these milestones moves young men forward in life to establishing a household and a steady job. Comparing imprisonment to these other life events indicates how the pathway through adulthood has been changed by the prison boom.

Table 1.3 shows the chances, for men born 1965 to 1969, of experiencing different life events by their early thirties in 1999. The risks of each life event are different for blacks and whites, but racial differences in imprisonment greatly overshadows any other inequality. Whites by their early thirties are more than twice as likely as blacks to hold a bachelor's degree. Blacks are about 50 percent more likely to have served in the military. However, black men in their early thirties are about seven times more likely than whites to have a prison record. Indeed, recent birth cohorts of black men are more likely to have prison records (22.4 percent) than military records 07.4 percent) or bachelor's degrees (12.5 percent). The share of the population with prison records is particularly striking among men with less than a college education. Whereas few such whites have prison records, nearly a third of black men with less than a college education have been to prison. Black men in their early thirties in 1999 without college were more than twice as likely to have served time in prison than to have served in the military. By 1999, imprisonment had become a common life event for black men that sharply distinguished them from white men.

MASS IMPRISONMENT

David Garland coined the term "mass imprisonment" to refer to the high rate of incarceration in the contemporary United States. In Garland's definition, mass imprisonment has two characteristics. First, he writes, "mass imrisonment implies a rate of imprisonment ... that is markedly abo' historical and comparative norm for societies of this type." 21 Indeec chapter has shown that the rate of incarceration in America by the late was far higher than in western Europe and without precedent in U.S. h Second, Garland argues, the demographic concentration of imprison produces not the incarceration of individual offenders, but the "syste imprisonment of whole groups of the population." 22 The empirical ill; of mass imprisonment are more slippery in this case. When will the inc arion rate be high enough to imprison, not the individual, but the grou

The picture painted by the statistics here help us answer this que, Not only did incarceration become common among young black men; end of the 1990s, its prevalence exceeded that of the other life events We ally associate with passage through the life course. More than college gr; tion or military service, for example, incarceration typified the biograph African American men born since the late 1960s.

Because of the nature of imprisonment, as an official sign of crirnin the collective experience of incarceration is as much a relational as a bi:

phical fact. The mass imprisonment generation-black men without college education born since 1965-is set apart from the mainstream by official criminality. Through its extent, concentration, and designation of deviance, mass imprisonment converts young black men with little schooling from a demographic category into a social group. As such, they share the same life chances and are ascribed the same social status by state officials, employers, and others in power. In the era of mass imprisonment, to be young, black, and male, even if never having gone to prison, is to arouse suspicion and fear. To go to prison, even if not young, black, and male, is to acquire something of that identity. The idea of mass imprisonment as constitutive of a social group, shares something with Loic Wacquant's idea of Jim Crow and the ghetto as "race-making" institutions.P Mass imprisonment, however, makes not the whole race. Instead, it divides the race, as the real experience of pervasive incarceration is confined just to those without college education. Imprisonment of the group, mass imprisonment, results not just from a high level of incarceration but from a high level of incarceration, unequally distributed.

CONCLUSION

The empirical evidence in this chapter supports three claims. First, the last two decades of the twentieth century produced a penal system that is without precedent in American history, and unlike any other in the advanced democracies. The growth in imprisonment has been sustained over the three decades from 1975 and incarceration rates in the early 2000s were five times higher than the incarceration rates that prevailed for most of the twentieth century. Although the u.s. incarceration rate had long been higher than in most western European countries, the imprisonment gap between Europe and the United States widened significantly in the period of the prison boom.

Second, race and class disparities in imprisonment are large, and class disparities have grown dramatically. From 1980 to 2000, black men were about six to eight times more likely to be in prison or jail than whites and Hispanics, about three times more likely. Some researchers claim that racial disparity has grown as incarceration increased, but I found no strong evidence for this trend. Class inequality increased, however, as a large gap in the prevalence of imprisonment opened between college-educated and non college men in the 1980s and the 1990s. Indeed, the lifetime risks of imprisonment roughl} doubled for all men from 1979 to 1999, but nearly all of this increased risk was experienced by those with just a high school education.

Third, imprisonment has become a common life event for recent birtl: cohorts of black men without college education. In 1999, about 30 percent of these had gone to prison by their early thirties. Among black male high school dropouts, the risk of imprisonment had increased to 60 percent, establishing incarceration as a normal stopping point on the route to midlife. Underscoring the historic novelty of the prison boom, these risks of imprisonment are about three times higher in 1999 than twenty years earlier.

The criminal justice system has become so pervasive that we should count prisons and jails among the key institutions that shape the life course of recent birth cohorts of African American men. By the end of the 1990s, black men with little schooling were more likely to be in prison or jail than to be in a labor union or enrolled in a government welfare or training program. Black men born in the late 1960s were more likely, by 1999, to have served time in state or federal prison than to have obtained a four-year degree or served in the military. For noncollege black men, a prison record had become twice as common as military service.

Although the great institutional interventions in the life course of the twentieth century had progressive effects, mass imprisonment threatens the reverse. The growth of military service during World War II and the expansion of higher education exemplify projects of "administered mobility" -that detached the fate of disadvantaged groups from their social backgrounds. Inequalities in imprisonment indicate the opposite effect, in which the life path of poor minorities was cleaved from the well-educated majority and disadvantage was deepened, rather than diminished. More strikingly than patterns of military enlistment, marriage, or college graduation, prison time differentiates the young adulthood of black men from that of most others. Convict status inheres now, not in individual offenders, but in entire demographic categories.

Why did incarceration rates rise so greatly, particularly among black men with little schooling? Two main explanations have been offered. One suggests that crime has increased. Although aggregate crime rates did not rise steadily through the 1980s and 1990s, some have claimed that urban street crime proliferated as joblessness increased in inner-city communities. Against this

argument, others say that the growth in incarceration rates was due largely to the changes in politics and policy. In this scenario, a crackdown on crime, beginning in the 1970s, intensified criminal punishment even though criminal offending did not increase. In the next two chapters, I weigh evidence for both these explanations of mass imprisonment.

APPENDIX: CALCULATING RATES AND RISKS OF INCARCERATION

Detailed incarceration rates for age-race education groups are estimated using data from the Survey of Inmates of State and Federal Correctional Facilities (1974, 1979, 1986, 1991, 1997),24 and the Survey of Inmates of Local Jails (1978, 1983, 1989, 1996).25 These figures are combined with counts of the noninstitutional population from the Current Population Survey and counts of military personnel to determine the size of the population.

The life table calculations reported in figures 1.3 and 1.4 are described in detail by Becky Pettit and Bruce Westetn.2⁶ The cumulative risks of imprisonment use the surveys of inmates of state and federal correctional facilities (1974 to 1997), to calculate age-specific risks of prison admissions for fiveyear birth cohorts from 1945 to 1949 to 1965 to 1969.

To help assess the accuracy of these estimates, I compared the cumulative risks to two other statistics. First, the Bureau of]ustice Statistics has reported lifetime risks of imprisonment using data from 1991 survey of prison inmates." These imprisonment risks are not defined for any particular birth cohort, nor are they calculated at different levels of education. Still, they do offer a rough guide to the prevalence of imprisonment in recent birth cohorts. The second source of data is a panel survey, the National Longitudinal Survey of Youth (NLSy) , that interviewed a national sample of young men every year until 1994, and then every other year after that (Center for Human Resource Research 2004). The survey recorded men's levels of education and whether they were interviewed in prison. The NLSY only provides data for one birth cohort, born 1957 to 1964, and a small sample of incarcerated men. Still, like the B]S estimates, it provides a check that my estimates agree with other data sources and methodologies.

CHAPTER 2

Inequality, Crime, and the Prison Boom

Extraordinary incarceration rates among young, less-educated black men at the end of the 1990s have a seemingly obvious explanation: black youth with little schooling commit a great deal of crime. Indeed, criminologists report high rates of serious violence among young black men, with strong indications that violence is concentrated among the poorest.' Even more suggestively, the emergence of mass imprisonment coincided with a twenty-year rise in economic inequality that stalled the economic progress of less-educated black men during the 1980s and 1990s. Unemployment and stagnant wages may have driven these men to crime. The story is more complicated than these patterns suggest, however, because trends in incarceration haven't tracked trends in the crime rate. The incarceration rate has steadily grown since the 1970s, through waves of violence in the late 1980s, and major gains in public safety through the late 1990s. Looking at race or class differences in crime at a point in time suggests that crime and incarceration are closely linked. Looking at changes in overall rates of serious offending suggests that trends in imprisonment and crime are unrelated.

Although the prison boom was not obviously driven by mounting crime rates, crime may have become more serious in poor communities while declining among the middle class. Under these conditions, the incarceration rate may not follow trends in aggregate crime rates, but increased crime

INEQUALITY, CRIME, AND THE PRlSON BOOM

among the poor could raise rates of arrest and prison admission. The key e pineal question, which few researchers have examined directly, asks if pc young men were more involved in crime at the height of the prison boom 2000 than twenty years earlier.

Here I examine the connections between social inequality, crime, and iJ prisonment. I begin by asking why socially marginal groups, like the pc and racial minorities, might be more involved in crime. Although ma studies describe links between race, class, and crime, I know of no empiric test that simply records whether crime increased among low-status you. men in the period of the prison boom. I provide several such tests and fiJ that young, poor men-black and white-were much less involved in crir in 2000 than in 1980, despite a large increase in their chances of incarcer tion. How can crime go down while incarceration goes up? To answer tl question, I go on to look at the stages of criminal processing from offendiJ to arrest, conviction, and imprisonment. This analysis shows that the crirr nal justice system became more punitive in the two decades from 1980, iJ creasing the risk of imprisonment for those who are arrested and increasir the time served of those locked up.

CRIME AND INEQUALITY

If the prison boom reflected trends in crime, we would expect that poorly ec ucated and Mrican American men were breaking the law more at the end ( the 1990s than twenty years earlier. Certainly, sociologists and economis have often argued that crime flourishes amid poverty and racial divisior Robert Merton famously claimed that frustration at blocked opportuniri, drives the poor to crime so they might obtain the material success enjoye legally by the middle class.? Economists make similar claims from a cost-ben efir perspective. Gary Becker argued that temptations of crime will b strongest when its benefits-the income from robbery or drug dealing, sayare high and its costs are low. Severe punishment can raise the costs of crime but so can the legitimate job opportunities that provide alternatives to illega activity." In addition to the influence of economic rewards, steady work sub jects daily life to supervision and routine. Continuously employed men hav. fewer opportunities to get involved in crime. Young unemployed men cal spend more time with their idle friends and may be more weakly committee to the roles of worker and provider.

These explanations emphasize the motives that push the poor into crime

but the home life and neighborhoods of the middle class can also erect barriers to criminal behavior. A stable marriage, like a steady job, creates everyday routines for husbands who might otherwise be on the street, getting into trouble. The social bonds of orderly, closely knit, neighborhoods also inhibit delinquency and crime. Communities lacking these social connections-in which families are weakly tied to employers, voluntary organizations, and friends-risk high rates of violence and other crime."

Unemployment, family instability, and neighborhood disorder combine to produce especially high rates of violence among young black men. Although black men made large economic strides between 1940 and 1970, their unemployment rate has been double that of whites since the 1970s. High rates of black unemployment accompany large numbers of femaleheaded households and neighborhoods of concentrated poverty. Because of low marriage rates among African Americans, and because black adolescents are more likely to grow up in female-headed households than whites, black youth are more loosely tied by the family bonds that prevent criminal offending. Poor black neighborhoods, in which poverty and its demographic correlates are highly concentrated, also lack the web of social networks that can supervise children after school, watch the street, and quickly seek help if it's needed.

Several statistical studies have found close links between violent crime and economic and racial inequality. Usually examining cities or states, these studies uncover high rates of homicide and other violence in areas of severe poverty, with large numbers of female-headed households. One of the most thorough studies of this kind, by Kenneth Land and his colleagues, finds that between 1960 and 1980, in central cities, broader metropolitan areas, and states, the highest rates of homicide are found in localities with the highest rates of poverty, unemployment, and divorce.> We also see the effects of family structure and the neighborhood environment when the focus is on urban violence. The large number of female-headed black families in metropolitan areas has been found to explain a significant share of the homicide rate among black adolescents.f Poor segregated neighborhoods with weak community ties and concentrated disadvantage have also been found to have more homicides, robberies, and burglaries." Although the quantitative studies find strong evidence that criminal violence is stratified by race and class, they provide little sense of the role of crime in the everyday life of the poor.

Ethnographies of poor urban areas offer a more vivid picture of the perva-

sive presence of crime and its close connection to incarceration. The inn: city drug trade occupies a special place in this research, providing econon opportunity for young men in neighborhoods with high rates of unemplc ment, In his ethnography of Hispanic drug gangs in New York, Philip Bourgois argues that "the insult of working for entry-level wages amidst e traordinary opulence is especially painful" for Spanish Harlem youths." TI inequality drives young Puerto Rican men "deeper into the confines of th. segregated neighborhood and the underground economy. "9 Sudl Venkatesh and Steven Levitt analyze the economic significance of drug de; ing for Chicago's "outlaw capitalism." Drug trafficking thrived in the va uum of legitimate employment in Chicago's Southside neighborhooc Chicago youth spoke to Venkatesh and Levitt of their "gang affiliation ar their drive to earn income in ways that resonated with representations work in the mainstream corporate firm. Many approached [gang] involv ment as an institutionalized path of socioeconomic mobility for down-ani out youth." IO In Elijah Anderson's account, violence follows the drug trade crime becomes a voracious force in the poor neighborhoods of Philadelphi.

Surrounded by violence and by indifference to the innocent victims of drug dealers and users alike, the decent people are finding it harder and harder to maintain a sense of community. Thus violence comes to regulate life in the drug-infested neighborhoods and the putative neighborhood leaders are increasingly the people who control the violence. I I

The picture drawn by the ethnographic research is of poor neighborhood: chronically short of legitimate work and embedded in a violent and illeg; market for drugs.

The perils of the drug trade and other street crime include not just th threat of violence, but also the risk of incarceration. The seemingly mechani cal connection between crime and incarceration is captured by Sullivan's ac count of how knifepoint robberies by Hispanic youth led progressively fron arrest, to jail time, to imprisonment. The inevitability of incarceration, Sulli van reflects, illustrates the "limits of confrontational street crime as a sourc of income. One, two, even several crimes may be perpetrated with impunity but continued involvement in such visible and violent crime does lead to se rious sanctions." 12 The great prevalence of incarceration in high-crime neigh borhoods is probably most extreme in Washington, D.C. Donald Bramar

38 PUNISHMENT AND INEQUALITY IN AMERICA

observes the experiences of Londa, a twenty-year-old mother of three living in the heart of the District. In the two-block radius of Londa's residence, Braman counted sixty-four arrests for drug possession and distribution over the course of a year. During that period, 120 men living within that two-block radius were admitted to the D.C. correctional system. Talking about the children in the neighborhood, Londa says, "I look around here and none of these kids have fathers. It's a mess what's happened."l3 Qualitative observations, like these, match in their details my statistical finding of pervasive incarceration among young men, particularly black men, with little schooling.

CRIMINAL ACTIVITY AMONG DISADVANTAGED MEN

It is often observed that trends in incarceration are, at best, only loosely related to trends in crime.l? Figure 2.1 compares the imprisonment rate to the index crime rate from 1970 to 2000. The index crime rate is calculated from the FBI's Uniform Crime Reports-serious crimes reported to the police, including murders, rapes, robberies, aggravated assaults, burglaries, larcenies, motor vehicle thefts, and arsons. The imprisonment rate grew steadily through periods of increasing crime in the 1970s, and declining crime in the 1990s. The correlation between incarceration and crime is a statistically insignificant, -.2. Statistics like these have led some commentators to discount any relationship between crime and punishment. Nils Christie points the point bluntly: "The explosion in the number of prisoners in the USA cannot be explained as 'caused by crime.' It has to do with penal policy." 15

This disposes of the issue too quickly. For recent ethnographies of urban poverty, incarceration is an occupational hazard of street crime, and crimeparticularly drug dealing-has been an important part of the ghetto economy at least since the 1980s. In this context, crime and incarceration appear closely related. But if crime among the ghetto poor were the main driver of the rise in incarceration rates, we would expect to see that young black men with little schooling were more involved in drug dealing and other crime in 2000 than in 1980. To test this hypothesis, I now turn to an empirical examination of crime among disadvantaged male youth.

There is surprisingly little empirical work examining trends in crime among poor young men. I provide a simple analysis using data on self-reported offending and criminal victimization. The 1979 and 1997 cohorts of the National Longitudinal Surveys of Youth (NLSy) are unusual in pro\ ing data on self-reported criminal activity at different points in time. 1 NLSY asked two national samples of young men and women in 1980 ;0 2000 about their criminal activity over the past year. I analyzed crime amc male NLSY respondents, aged fifteen to eighteen. Because the NLSY resp( dents are younger than adult prisoners, the survey data do not provide dir evidence on those at risk of imprisonment. Still, juvenile and adult cri rates move roughly together and virtually all adult felons have a history of venile offending. If increased criminality among lower-class men increa: the prison population, we would also expect to see increased criminal among lower-class youth.

Data on juveniles also have important advantages for studying trends crime. Some writers attribute a large decline in adult crime between 19 and 2000 to the rise in imprisonment. 16 Juvenile incarceration, however, 1 not increased nearly as much as adult incarceration. In 1979, the juvenile j carcerated population numbered 71,922 compared to 108,931 in 1999Y the juvenile incarcerated population increased by about 50 percent, the adt prison population increased by 430 percent. IS If we observe large declines in juvenile crime, as we have for adults, it is much less likely that these are explained by rising incarceration.

Table 2.1 describes criminal activity among male youth in the NLSY. Respondents were asked to report whether they had attacked someone, vandalized property, stolen something, or sold drugs in the past year. I calculated crime rates for all male youth, and for those whose household incomes fell below the poverty line.

Consistent with broader trends, older teenagers in the NLSY were much less involved in crime in 2000 than in 1980. Rates of crime declined significantly in all offense categories. Property crime fell the most as the number of youth who reported vandalism or larceny (stealing) dropped by between 65 and 75 percent. The decline in violent assaults is smaller but still substantial. In 1980, 15 percent of male youth said they had attacked someone with the intention of seriously hurting them, or killing them, compared to fewer than 11 percent in 2000.

If mushrooming criminality among the poor drove the prison boom, we would expect to see rising crime among poor youth. In fact, like the rest of the population, youth from poor families were less involved in delinquency in 2000 than in 1980. Property crime was halved among poor whites, and fell by even more among the poor minority youth, whom many writers saw as drawn to crime by the decline of inner-city economies. Although research claims that drug dealing replaced legitimate economic opportunity in ghetto neighborhoods, poor black and white youth in the NLSY survey were selling drugs far less in 2000 than in 1980. For example, 16 percent of black teenagers from poor families said they sold drugs in 1980, compared to just 5 percent in 2000. Only violent crime did not fall substantially. Poor juveniles were getting into serious fights about as often in 2000 as they were twenty years earlier. In short, data on self-reported crime among young men from poor families look very similar to aggregate crime rates. Levels of crime at the end of the 1990s were much lower than in the early 1980s. There is no evidence here that disadvantaged youth have become more involved in crime. Indeed, they are much less involved in crime at the peak of the prison boom in 2000, that at the beginning in 1980. What's more, because juvenile incarceration rates have not greatly increased, the decline in crime shown in the survey data is unlikely to be an artifact of rising incarceration.

The NLSY data provide no evidence of rising crime among needy youth, but self-report data are subject to errors in which respondents may under port very serious crime. Another approach looks at victimization data. V timization is typically recorded by surveys, in which individuals are asked they have been assaulted, had property stolen, and so on. Because a great d, of crime happens among acquaintances and neighbors, victimization data f poor men will also be informative about their levels of criminal activity. T

National Crime Victimization Survey annually asks a national sample, twelve years and older, about their experiences of household, property, and violent crime. I calculated property and violent victimization rates for young men, aged twenty-two to thirty, and for young male high school, dropouts, the group who have experienced the largest increases in incarceration. Because sample sizes are quite small for dropouts, I pooled data for 1980 to 1983 and for 1997 to 2000.

Table 2.2 reports victimization rates for the two periods for all young men and for those who dropped out of high school. Between the early 1980s and the late 1990s there was a large decline in rates of criminal victimization. In the mid-1980s, one of two men in their twenties was a victim of property crime or violence. By the late 1990s, fewer than one in five was, a decline of about 70 percent. Men with little education also shared in the gains in public safety. Victimization rates for young dropouts declined by between 60 and 75 percent, in line with national trends.

INEQUALITY, CRIME, AND THE PRISON BOOM

Unlike the earlier analysis of juvenile crime, victimization figures amc young men may reflect the effects of growing imprisonment. Men may safer by the end of the 1990s because prisons had taken most of the crimir off the streets. I adjusted for this effect by taking account of the grow punitiveness of the criminal justice system. This can be measured by , number of prison admissions for every arrest. In the early 1980s, admissic accounted for about 1. 8 percent of all arrests. By the late 1990s, that fig! was about 3.6 percent. By this measure, punitiveness between 1980 a 2000 has doubled. Assuming each person imprisoned would otherwise be sponsible for ten crimes against men aged twenty-two to thirty, I can adjr victimization rates to reflect the growth in criminal punishment.

The second panel of table 2.2 reports adjusted victimization rates rl: take account of the rise in imprisonment. Although the declines in rates 2 smaller once growing imprisonment is accounted for, they are still substa tial. Among young white dropouts, victimization has fallen by nearly 50 pc cent and fallen even more among young and less-educated minorities.

These data leave us with a puzzle. Between 1980 and 2000, self-repor« crime fell significantly among disadvantaged youth. Teenagers from poe households in 2000 were less involved in violence and drug dealing rha twenty years earlier. Declines in vandalism and theft were especially large an were mirrored by large declines in criminal victimization among young les. educated men. High school dropouts in their twenties, whose incarceratio rates are now extraordinarily high, are 60 to 70 percent less likely to be vic rirnized by crime in 2000 than in 1980. How have large reductions in crim among disadvantaged young men become associated with large increases i incarceration?

LINKING CRBMIE TO PUNISHMENT

We can understand how reduced crime is associated with increased impris onrnenr by following each stage of criminal processing. Table 2.3 compare, the number of crimes to rates of arrest and prison admission. About one mil, lion violent crimes are reported to the police each year. The number of vio. lent crimes increased from over nine hundred thousand to 1.36 million from 1980 to 1990. But from 1990 to 2001 the level of violence fell. Just under half of the complaints to police resulted in an arrest. However, the chances that an arrest would result in prison roughly doubled, from 13 to 28 percent.

Time served in prison by violent offenders also increased significantly, from thirty-three months in 1980 to fifty-three months on average by 2001. Because time served and the rate of prison admission both increased, the incarceration rate for violent crime rose from seventy-six to 208 per hundred thousand, despite the decline in the level of violence.

Property offenders, mostly burglars and car thieves, show a similar pattern. About eight times more property crime is reported to police than violent crime. Like violent crime, property crime increased from 1980 to 1990, but then fell over the next ten years. Fewer than one in five property crimes result in an arrest, much lower than the rate for violence. From 1980 to 2001, the property offender's chances of imprisonment nearly doubled, from 6 to 11 percent. Time served for property crime also increased by 75 percent. The increasing chances of imprisonment given an arrest and increasing time in prison more than doubled the incarceration rate.

Finally, figures on drug crimes show a similar pattern of intensified law enforcement. Unlike violent and property offenses, there are no crime statistics on levels of drug use and trafficking, neither in victimization surveys nor in the police reports of the Uniform Crime Survey. The numbers of drug arrests, however, are recorded. Unlike arrest trends for other crime types, drug arrests increased by 170 percent in the two decades from 1980. The prison admissions for each arrest increased sixfold, from two to 12 percent. Many drug offenders who were admitted to prison in the 1990s were parolees readmitted not for new offenses, but for violating the conditions of their parole. Failing a drug test, for example, is a common parole violation. In the 1990s, parole revocation doubled for drug offenders, increasing the number of people reentering prison and not convicted of new crimes. Timeserved also increased sharply so that by 2001, released drug offenders had served about two years in state prison. These factors-the large increase in drug arrests, the growing likelihood of imprisonment given arrest, the increased risk of parole revocation, and increasing time served-produced more than a tenfold increase in the drug crime incarceration rate from 1980 to 2001.

These figures on arrest, prison admission, parole revocation, and time served explain why trends in crime bear only a slight relationship to the scale of imprisonment. At every stage of criminal processing, from policing, to the court hearing, to parole, criminal justice officials decide on the disposition of offenders and these effects on the scale of imprisonment far overshadow fluctuations in the level of crime.

DRUG USE AND DRUG ENfORCEMENT

How do we interpret the large increase in the number of drug arrests f[i 1980 to 2001? Propelled by policy initiatives of first the Nixon and then 1 Reagan administration, drug enforcement escalated dramatically through 1 1970s and 1980s. I'll have more to say about America's war on drugs in 1following chapter, but for now we can see its quantitative extent reflected in the fourfold growth in drug arrest rates from the late 1960s to 2001 (see figure 2.2). Drug arrests had always shown a large racial disparity. In the early 1970s, blacks were about twice as likely as whites to be arrested for a drug offense. The great growth in drug arrest rates through the 1980s had a large effect on African Americans. At the height of the drug war in 1989, arrest rates for blacks had climbed to 1,460 per hundred thousand compared to 365 for whites. Throughout the 1990s, drug arrest rates remained at these historically high levels. These trends may be related to trends in drug use or to trends in drug enforcement. Because there are no crime statistics on drug use, we cannot systematically compare offending rates to arrest and imprisonment rates. However, we can look further at the link between drug use and drug arrests trends in drug use by examining social surveys and hospital reports of drug-related emergency room visits.

Figure 2.3 shows trends in drug Use with data from the Monitoring the Future Survey and the Drug Abuse Warning Network. The survey asked a national sample of high school seniors whether they had ever used drugs in the past year. In 1980, just Over 50 percent said they had, compared to around 40 percent in 2000. White high school students consistently reported more drug use than black students. National samples of adults, studied by the National Survey on Drug Abuse (NSDA), similarly show that drug use among adults declined from 20 to 11 percent from 1979 to 2000. Like the high school survey, the NSDA shows that levels of drug use do not differ much between blacks and whites.

The survey responses may not provide a good indication of trends in serious drug use. A shorter time series is available from the Drug Abuse Warning Network that records the number of drug-related emergency room visits from hospitals in twenty-one cities. Whites had roughly twice to three times the number of drug-related emergency room visits than blacks. Whites increased their share of drug-related emergency care through the 1990s. Al-

48 PUNISHMENT AND INEQUALITY IN AMERICA

though data on drug use are patchy, there is little evidence that mounting drug use or relatively high rates of drug use among blacks fueled the increase in drug arrests during the 1990s.

CRIME AS CONTEXT RATHER THAN CAUSE

We have seen that there is no consistent, positive relationship between crime and incarceration rates through the prison boom period. Still, several writers claim that trends in crime are not the cause but the context for the rise in incarceration rates.t? In this account, the growth in crime in the 1960s before the prison boom exposed middle-class whites to serious risks of victimization for the first time. The growth in crime contributed to new feelings of vulnerability among the affluent and created a political opening for a change in crime policy that ultimately increased the incarceration rate.

How much did crime really increase? Crime rates were not measured very accurately until the Census Bureau began its victimization survey in 1972. Still, before 1972, murder rates were measured more accurately than other crime statistics." From 1965 to 1980, the annual number of murders in the United States increased from about ten thousand each year to more than twenty thousand, an increase from 5.1 to 9.6 deaths per hundred thousand residents. Although measured less accurately, the overall violent crime ratewhich includes rapes, robberies, and assaults as well as homicides-increased threefold between 1965 and 1980, from 200 to 597 crimes per hundred thousand." The large growth in crime rates predated the explosion of the pe-

nal population.

Although the incarceration trends have not tracked trends in crime, the

largest increases in the 1960s and 1970s are in states with the largest increases in incarceration rates twenty years later (figure 2.4). Southern states such as Louisiana, Mississippi, Oklahoma, and Texas all experienced big increases in their homicide rates. By the end of the 1990s, these four had the highest rates of incarceration. At the other end of the scale, small Midwestern and New England states saw only modest increases in serious violence through the 1970s, and incarceration rates increased only a little in subsequent decades.

This empirical pattern provides a clue. Crime rates themselves may not

have driven the prison boom, but long-standing fears about crime and other social anxieties may form the backdrop for the growth in imprisonment. While crime was rising in the late 1960s, urban riots, racial tensions, and economic recession closed a chapter in American postwar social history. period of rapid social change ushered in a new economy, characterized l: ban deindustrialization, and a new politics, characterized by law-and-, appeals to white suburban voters. We turn to these trends, and their ef in the next chapter.

CONCUJSION

The research and data I have presented so far help us untangle the link tween social inequality, crime, and the growth of imprisonment amon~ advantaged men. In trying to understand why so many less educated anc nority men are going to prison, we should carefully distinguish the two to this story.

First, there is good evidence that disadvantaged men are, at any ~

time, highly involved in crime and that this is closely associated with the high rate of their imprisonment. Quantitative studies and field observation in poor neighborhoods show that poor young men are greatly involved in violence and other crime. Black homicide offending offers the clearest example. Several studies have shown that blacks are roughly seven times more likely to be imprisoned for murder than white men, but are also seven times more likely to be arrested for murder and to be murdered than whites. High rates of homicide among black men fully explain the parallel high rates of imprisonment for murder. However, for less serious offenses, race differences in incarceration are not well explained by high crime rates. Black men are much more likely than whites to be arrested for a drug offense, and go to prison if arrested, even though they are no more likely to use drugs than whites. Criminologists estimate that about 80 percent of black-white difference in imprisonment rates is due simply to the high involvement of black men in crime.P This number has likely declined with growth in the share of drug offenders in prison. We also have seen research showing that economically disadvantaged men are also more involved in crime than middle-class men. It seems reasonable that a large fraction of the class inequality in incarceration is also attributable to class differences in offending, but there is little direct evidence on this point.

Second, although high crime rates among the disadvantaged largely explain their incarceration at a given time, trends in crime and imprisonment are only weakly related over time. Poor and minority men were much less involved in crime in 2000 than twenty years earlier, matching declines in crime in the population as a whole. Although disadvantaged men became much more law-abiding, their chances of going to prison rose to historically high levels. Statistics on criminal processing showed that there were three main causes of the growth in imprisonment, each of which is unrelated to trends in crime. First was a significant increase in the use of imprisonment for those who are convicted of a crime. Second was that those who go to prison are now serving longer sentences. Third was a dramatic increase in the prosecution and incarceration of drug offenders. Indeed, 45 percent of the increase in the state prison population is explained by the rising incarceration among drug offenders.²³ The increased risk of imprisonment given arrest, and increased time served, show that courts are treating drug and other offenders more harshly than before. What's more, crime statistics do not measure the level of drug offending so most crime statistics shed little light on a I of state prison population growth.

Crime is not consistently related to imprisonment trends in the and 1990s, but states that experienced the largest increases in seric lence through the 1960s and 1970s, also experienced the largest gain: prisonment decades later. These trends lead us to look at the politics ( policy and the economic context in which criminal justice is adminis

DMU Timestamp: March 08, 2013 19:51