Bruce Western & Becky Pettit
Dædalus, Summer 2010
In the last few decades,
the institutional contours of American social inequality have been transformed by
the rapid growth in the prison and jail population.1 America’s prisons
and jails have produced a new social group, a group of social outcasts who are joined
by the shared experience of incarceration, crime, poverty, racial minority, and
low education. As an outcast group, the men and women in our penal institutions
have little access to the social mobility available to the mainstream. Social and
economic disadvantage, crystallizing in penal confinement, is sustained over the
life course and transmitted from one generation to the next. This is a profound
institutionalized inequality that has renewed race and class disadvantage. Yet the
scale and empirical details tell a story that is largely unknown.
Though the rate of incarceration is historically high, perhaps the most important
social fact is the inequality in penal confinement. This inequality produces extraordinary
rates of incarceration among young African American men with no more than a high
school education. For these young men, born since the mid-1970s, serving time in
prison has become a normal life event.
The influence of the penal system on social
and economic disadvantage can be seen in the economic and family lives of the formerly
incarcerated. The social inequality produced by mass incarceration is sizable and
enduring for three main reasons: it is invisible, it is cumulative, and it is intergenerational.
The inequality is invisible in the sense that institutionalized populations commonly
lie outside our official accounts of economic well-being. Prisoners, though drawn
from the lowest rungs in society, appear in no measures of poverty or unemployment.
As a result, the full extent of the disadvantage of groups with high incarceration
rates is underestimated. The inequality is cumulative because the social and economic
penalties that flow from incarceration are accrued by those who already have the
weakest economic opportunities. Mass incarceration thus deepens disadvantage and
forecloses mobility for the most marginal in society. Finally, carceral inequalities
are intergenerational, affecting not just those who go to prison and jail but their
families and children, too.
The scale of incarceration is measured by a rate that records the fraction of the
population in prison or jail on an average day. From 1980 to 2008, the U.S. incarceration
rate climbed from 221 to 762 per 100,000. In the previous five decades, from the
1920s through the mid-1970s, the scale of punishment in America had been stable
at around 100 per 100,000. Though the incarceration rate is now nearly eight times
its historic average, the scale of punishment today gains its social force from
its unequal distribution.
Like criminal activity, prisons and jails are overwhelmingly a male affair. Men
account for 90 percent of the prison population and a similar proportion of those
in local jails. The incarceration rate has been growing faster among women in recent
decades, but the social impact of mass incarceration lies in the gross asymmetry
of community and family attachment. Women remain in their communities raising children,
while men confront the possibility of separation through incarceration.2
Age intensifies these effects: incarceration rates are highest for those in their
twenties and early thirties. These are key years in the life course, when most men
are establishing a pathway through adulthood by leaving school, getting a job, and
starting a family. These years of early adulthood are important not just for a man’s
life trajectory, but also for the family and children that he helps support.
Age and sex are the staples of demographic analysis, and the relative youth of the
largely male incarcerated population foreshadows much about the effects of mass
incarceration. Still, it is the profound race and class disparities in incarceration
that produce the new class of social outsiders. African Americans have always been
incarcerated at higher rates than whites, at least since statistics were available
from the late nineteenth century. The extent of racial disparity, however, has varied
greatly over the past century, following a roughly inverse relationship to the slow
incorporation of African Americans as full citizens in American society. In the
late nineteenth century, U.S. Census data show that the incarceration rate among
African Americans was roughly twice that of whites. The demographic erosion of Jim
Crow through the migration of Southern African Americans to the North increased
racial disparity in incarceration through the first half of the twentieth century.
(Racial disparities in incarceration have always been higher in the North than the
South.) By the late 1960s, at the zenith of civil rights activism, the racial disparity
had climbed to its contemporary level, leaving African Americans seven times more
likely to be in prison or jail than whites.
Class inequalities in incarceration are reflected in the very low educational level
of those in prison and jail. The legitimate labor market opportunities for men with
no more than a high school education have deteriorated as the prison population
has grown, and prisoners themselves are drawn overwhelmingly from the least educated.
State prisoners average just a tenth grade education, and about 70 percent have
no high school diploma.3
Disparities of race, class, gender, and age have produced extraordinary rates of
incarceration among young African American men with little schooling. Figure 1 shows
prison and jail incarceration rates for men under age thirty-five in 1980, at the
beginning of the prison boom, and in 2008, after three decades of rising incarceration
rates. The figure reports incarceration separately for whites, Latinos, and African
Americans and separately for three levels of education. Looking at men with a college
education, we see that incarceration rates today have barely increased since 1980.
Incarceration rates have increased among African Americans and whites who have completed
high school. Among young African American men with high school diplomas, about one
in ten is in prison or jail.
Percentage of Men Aged Twenty to Thirty-Four in Prison or Jail, by Race/Ethnicity
and Education, 1980 and 2008
Source: Becky Pettit, Bryan Sykes, and Bruce Western,
“Technical Report on Revised Population Estimates and NLSY79 Analysis Tables for
the Pew Public Safety and Mobility Project” (Harvard University, 2009).
Most of the growth in incarceration rates is concentrated at the very bottom, among
young men with very low levels of education. In 1980, around 10 percent of young
African American men who dropped out of high school were in prison or jail. By 2008,
this incarceration rate had climbed to 37 percent, an astonishing level of institutionalization
given that the average incarceration rate in the general population was 0.76 of
1 percent. Even among young white dropouts, the incarceration rate had grown remarkably,
with around one in eight behind bars by 2008. The significant growth of incarceration
rates among the least educated reflects increasing class inequality in incarceration
through the period of the prison boom.
These incarceration rates provide only a snapshot at a point in time. We can also
examine the lifetime chance of incarceration – that is, the chance that someone will
go to prison at some point in his or her life. This cumulative risk of incarceration
is important if serving time in prison confers an enduring status that affects life
chances after returning to free society. The lifetime risk of imprisonment describes
how many people are at risk of these diminished life chances.
We calculated the cumulative chance of imprisonment for two birth cohorts, one born
just after World War II, from 1945 to 1949, and another born from 1975 to 1979 (Table
1). For each cohort, we calculated the chances of imprisonment, not jail incarceration.
Prisons are the deep end of the criminal justice system, now incarcerating people
for an average of twenty-eight months for a felony conviction. While there are about
ten million admissions to local jails each year – for those awaiting trial or serving
short sentences – around seven hundred thousand prisoners are now admitted annually
to state and federal facilities.
Cumulative Risk of Imprisonment by Age Thirty to Thirty-Four for Men Born from
1945 to 1949 and 1975 to 1979, by Educational Attainment and Race/Ethnicity
High School Dropouts
High School GED
*Denotes completed high school or equivalency
Source: Becky Pettit, Bryan Sykes, and Bruce
Western, “Technical Report on Revised Population Estimates and NLSY 79 Analysis Tables
for the Pew Public Safety and Mobility Project” (Harvard University, 2009).
These cumulative chances of imprisonment are calculated up to age thirty-four. For
most of the population, this represents the lifetime likelihood of serving prison
time. For the older postwar cohort who reached their mid-thirties at the end of
the 1970s, about one in ten African American men served time in prison. For the
younger cohort born from 1975 to 1979, the lifetime risk of imprisonment for African
American men had increased to one in four. Prison time has become a normal life
event for African American men who have dropped out of high school. Fully 68 percent
of these men born since the mid-1970s have prison records. The high rate of incarceration
has redrawn the pathway through young adulthood. The main sources of upward mobility
for African American men – namely, military service and a college degree – are significantly
less common than a prison record. For the first generations growing up in the post–civil
rights era, the prison now looms as a significant institutional influence on life
The ubiquity of penal confinement in the lives of young African American men
with little schooling is historically novel, emerging only in the last decade. However, this new reality
is only half the story of understanding the significance of mass incarceration in
America. The other half of the story concerns the effects of incarceration on social
and economic inequality. The inequalities produced by contemporary patterns of incarceration
have three characteristics: the inequalities associated with incarceration are invisible
to our usual accounting of the economic well-being of the population; the inequality
is cumulative, deepening the disadvantage of the most marginal men in society; and
finally, the inequality is intergenerational, transmitting the penalties of a prison
record from one generation to the next. Because the characteristic inequalities
produced by the American prison boom are invisible, cumulative, and intergenerational,
they are extremely enduring, sustained over lifetimes and passed through families.
Invisible Inequality. The inequality created by incarceration is often invisible
to the mainstream of society because incarceration is concentrated and segregative.
We have seen that steep racial and class disparities in incarceration have produced
a generation of social outliers whose collective experience is wholly different
from the rest of American society. The extreme concentration of incarceration rates
is compounded by the obviously segregative function of the penal system, which often
relocates people to far-flung facilities distant from their communities and families.
As a result, people in prison and jail are disconnected from the basic institutions – households
and the labor market – that dominate our common understanding and measurement of the
population. The segregation and social concentration of incarceration thus help
conceal its effects. This fact is particularly important for public policy because
in assessing the social and economic well-being of the population, the incarcerated
fraction is frequently overlooked, and inequality is underestimated as a result.
The idea of invisible inequality is illustrated by considering employment rates
as they are conventionally measured by the Current Population Survey, the large
monthly labor force survey conducted by the Census Bureau. For groups that are weakly
attached to the labor market, like young men with little education, economic status
is often measured by the employment-to-population ratio. This figure, more expansive
than the unemployment rate, counts as jobless those who have dropped out of the
labor market altogether. The Current Population Survey is drawn on a sample of households,
so those who are institutionalized are not included in the survey-based description
of the population.
Figure 2 shows the employment-to-population ratio for African American men under
age thirty-five who have not completed high school. Conventional estimates of the
employment rate show that by 2008, around 40 percent of African American male dropouts
were employed. These estimates, based on the household survey, fail to count that
part of the population in prison or jail. Once prison and jail inmates are included
in the population count (and among the jobless), we see that employment among young
African American men with little schooling fell to around 25 percent by 2008. Indeed,
by 2008 these men were more likely to be locked up than employed.
Employment to Population Ratio, African American Men Aged Twenty to Thirty-Four
with Less than Twelve Years of Schooling, 1980 to 2008
Source: Becky Pettit, Bryan
Sykes, and Bruce Western, “Technical Report on Revised Population Estimates and
NLSY79 Analysis Tables for the Pew Public Safety and Mobility Project” (Harvard
Cumulative Inequality. Serving time in prison or jail diminishes social and economic
opportunities. As we have seen, these diminished opportunities are found among those
already most socioeconomically disadvantaged. A burgeoning research literature examining
the economic effects of incarceration finds that incarceration is associated with
reduced earnings and employment.4
We analyzed panel data from the National Longitudinal Survey of Youth (NLSY), one
of the few surveys that follows respondents over a long period of time and that
interviews incarcerated respondents in prison. The NLSY began in 1979, when its
panel of respondents was aged fourteen to twenty-one; it completed its latest round
of interviews in 2006. Matching our population estimates of incarceration, one in
five African American male respondents in the NLSY has been interviewed at some point
between 1979 and 2006 while incarcerated, compared to 5 percent of whites and 12
percent of Latino respondents. Analysis of the NLSY showed that serving time in
prison was associated with a 40 percent reduction in earnings and with reduced job
tenure, reduced hourly wages, and higher unemployment.
The negative effects of incarceration, even among men with very poor economic opportunities
to begin with, are related to the strong negative perceptions employers have of
job seekers with criminal records. Devah Pager’s experimental research has studied
these employer perceptions by sending pairs of
fake job seekers to apply for real jobs.5In each
pair, one of the job applicants was randomly assigned a résumé indicating a criminal
record (a parole officer is listed as a reference), and the “criminal” applicant
was instructed to check the box on the job application indicating he had a criminal
record. A criminal record was found to reduce callbacks from prospective employers
by around 50 percent, an effect that was larger for African Americans than for whites.
Incarceration may reduce economic opportunities in several ways. The conditions
of imprisonment may promote habits and behaviors that are poorly suited to the routines
of regular work. Time in prison means time out of the labor force, depleting the
work experience of the incarcerated compared to their nonincarcerated counterparts.
The stigma of a criminal conviction may also repel employers who prefer job applicants
with clean records. Pager’s audit study offers clear evidence for the negative effects
of criminal stigma. Employers, fearing legal liability or even just unreliability,
are extremely reluctant to hire workers with criminal convictions.
A simple picture of the poor economic opportunities of the formerly incarcerated
is given by the earnings mobility of men going to prison compared to other disadvantaged
groups. The NLSY data can be used to study earnings mobility over several decades.
We calculated the chances that a poor man in the lowest fifth of the earnings distribution
in 1986 would move up and out of the lowest fifth by 2006. Among low-income men who
are not incarcerated, nearly two-thirds are upwardly mobile by 2006 (Figure 3). Another
group in the NLSY has very low levels of cognitive ability, scoring in the bottom
quintile of the Armed Forces Qualifying Test, the standardized test used for military
service. Among low-income men with low scores on the test, only 41 percent are upwardly
mobile. Upward mobility is even less common among low-income high school dropouts.
Still, we observe the least mobility of all among men who were incarcerated at some
point between 1986 and 2006. For these men, only one in four rises out of the bottom
quintile of the earnings distribution.
Twenty-Year Earnings Mobility among Men in the Bottom Quintile of Earnings
Distribution in 1986, National Longitudinal Survey of Youth (NLSY) Men
AFQT stands for Armed Forces Qualifying Test.
Source: Becky Pettit, Bryan Sykes, and Bruce Western,
"Technical Report on Revised Population Estimates and NLSY79 Analysis Tables for
the Pew Public Safety and Mobility Project" (Harvard University, 2009).
Intergenerational Inequality. Finally, the effects of the prison boom extend also
to the families of those who are incarcerated. Through the prism of research on
poverty, scholars find that the family life of the disadvantaged has become dramatically
more complex and unstable over the last few decades. Divorce and nonmarital births
have contributed significantly to rising rates of single parenthood, and these changes
in American family structure are concentrated among low-income mothers. As a consequence,
poor children regularly grow up, at least for a time, with a single mother and,
at different times, with a variety of adult males in their households.
High rates of parental incarceration likely add to the instability of family life
among poor children. Over half of all prisoners have children under the age of eighteen,
and about 45 percent of those parents were living with their children at the time
they were sent to prison. About two-thirds of prisoners stay in regular contact
with their children either by phone, mail, or visitation.6 Ethnographer
Megan Comfort paints a vivid picture of the effects of men’s incarceration on the
women and families in their lives. She quotes a prisoner at San Quentin State Prison
Nine times out of ten it’s the woman [maintaining contact with prisoners]. Why?
Because your homeboys, or your friends, if you’re in that lifestyle, most the time
they’re gonna be sittin’ right next to your ass in prison. . . . The males, they
don’t really participate like a lot of females in the lives of the incarcerated.
. . . They don’t deal with it, like first of all they don’t like to bring to reality
that you’re in prison; they don’t wanna think about that . . . Or some of ’em just
don’t care. So the male’s kinda like wiped out of there, so that puts all the burden
on the woman.7
Partly because of the burdens of incarceration on women who are left to raise families
in free society, incarceration is strongly associated with divorce and separation.
In addition to the forced separation of incarceration, the post-release effects on
economic opportunities leave formerly incarcerated parents less equipped to provide
financially for their children. New research also shows that the children of incarcerated
parents, particularly the boys, are at greater risk of developmental delays and
Against this evidence for the negative effects of incarceration, we should weigh
the gains to public safety obtained by separating violent or otherwise antisocial
men from their children and partners. Domestic violence is much more common among
the formerly incarcerated compared to other disadvantaged men. Survey data indicate
that formerly incarcerated men are about four times more likely to assault their
domestic partners than men who have never been incarcerated. Though the relative
very high, around 90 percent of the partners of formerly incarcerated report no
domestic violence at all.
The scale of the effects of parental incarceration on
children can be revealed simply by statistics showing the number of children with
a parent in prison or jail. Among white children in 1980, only 0.4 of 1 percent
had an incarcerated parent; by 2008 this figure had increased to 1.75 percent. Rates
of parental incarceration are roughly double among Latino children, with 3.5 percent
of children having a parent locked up by 2008. Among African American children,
1.2 million, or about 11 percent, had a parent incarcerated by 2008 (Figure 4).
Number of Children under Eighteen with a Parent in Prison or Jail, 1980
Source: Becky Pettit, Bryan Sykes, and Bruce Western, "Technical Report
on Revised Population Estimates and NLSY79 Analysis Tables for the Pew Public Safety
and Mobility Project" (Harvard University, 2009).
The spectacular growth in the American penal system over the last three decades
was concentrated in a small segment of the population, among young minority men
with very low levels of education. By the early 2000s, prison time was a common
life event for this group, and today more than two-thirds of African American male
dropouts are expected to serve time in state or federal prison. These demographic
contours of mass imprisonment have created a new class of social outsiders whose
relationship to the state and society is wholly different from the rest of the population.
Social marginality is deepened by the inequalities produced by incarceration. Workers
with prison records experience significant declines in earnings and employment. Parents
in prison are likely to divorce or separate, and through the contagious effects
of the institution, their children are in some degree “prisonized,” exposed to the
routines of prison life through visitation and the parole supervision of their parents.
Yet much of this reality remains hidden from view. In social life, for all but those
whose incarceration rates are highest, prisons are exotic institutions unknown to
the social mainstream. Our national data systems, and the social facts they produce,
are structured around normative domestic and economic life, systematically excluding
Thus we define carceral inequalities as invisible, cumulative, and intergenerational.
Because they are so deeply concentrated in a small disadvantaged fraction of the
population, the social and economic effects of incarceration create a discrete social
group whose collective experience is so distinctive yet unknown that their disadvantage
remains largely beyond the apprehension of public policy or public conversation.
The redrawing of American social inequality by mass incarceration amounts to a contraction
of citizenship – a contraction of that population that enjoys, in T. H. Marshall’s
words, “full membership in society.”9 Inequality of this kind threatens
to be self-sustaining. Socioeconomic disadvantage, crime, and incarceration in the
current generation undermine the stability of family life and material support for
children. As adults, these children will be at greater risk of diminished life chances
and criminal involvement, and at greater risk of incarceration as a result.
Skeptics will respond that these are false issues of social justice: the prison
boom substantially reduced crime, and criminals should forfeit their societal membership
in any case. The crime-reducing effects of incarceration are hotly debated, however.
Empirical estimates of the effects of incarceration on crime vary widely, and often
they turn on assumptions that are difficult to test directly. Researchers have focused
on the sharp decline in U.S. crime rates through the 1990s, studying the influence
of rising prison populations. Conservative estimates attribute about one-tenth of
the 1990s crime decline to the growth in imprisonment rates.10 Though
the precise impact of incarceration on crime is uncertain, there is broad agreement
that additional imprisonment at high rates of incarceration does little to reduce
crime. The possibility of improved public safety through increased incarceration
is by now exhausted.
Studies of the effects of incarceration on crime also focus only on the short term.
Indeed, because of the negative effects of incarceration on economic opportunities
and family life, incarceration contributes to crime in the long run by adding to
idleness and family breakdown among released prisoners. Scale matters, too. If the
negative effects of incarceration were scattered among a small number of serious
criminal offenders, these effects may well be overwhelmed by reduction in crime
Today, however, clear majorities of the young men in poor communities are going
to prison and returning home less employable and more detached from their families.
In this situation, the institutions charged with public safety have become vitally
implicated in the unemployment and the fragile family structure characteristic of
high-crime communities. For poorly educated young men in high-incarceration communities,
a prison record now carries little stigma; incentives to commit to the labor market
and family life have been seriously weakened.
To say that prison reduces crime (perhaps only in the short run) is a spectacularly
modest claim for a system that now costs $70 billion annually. Claims for the crime-reducing
effects of prison, by themselves, provide little guidance for policy because other
approaches may be cheaper. Measures to reduce school dropout, increase human capital,
and generally increase employment among young men seem especially promising alternatives.
Results for programs for very young children are particularly striking. Evaluations
of early childhood educational programs show some of their largest benefits decades
later in reduced delinquency and crime.11 For adult men now coming out
of prison, new evaluations show that jobs programs reduce recidivism and increase
employment and earnings.12 The demographic concentration of incarceration
accompanies spatial concentration. If some portion of that $70 billion in correctional
expenditures were spent on improving skills and reducing unemployment in poor neighborhoods,
a sustainable and socially integrative public safety may be produced.
Much of the political debate about crime policy ignores the contemporary scale of
criminal punishment, its unequal distribution, and its negative social and economic
effects. Our analysis of the penal system as an institution of social stratification,
rather than crime control, highlights all these neglected outcomes and leaves us
pessimistic that widespread incarceration can sustainably reduce crime. The current
system is expensive, and it exacerbates the social problems it is charged with controlling.
Our perspective, focused on the social and economic inequalities of American life,
suggests that social policy improving opportunity and employment, for young men
in particular, holds special promise as an instrument for public safety.
Our perspective on inequality points to a broader view of public safety that is
not produced by punishment alone. Robust public safety grows when people have order
and predictability in their daily lives. Crime is just one danger, joining unemployment,
poor health, and family instability along a spectrum of threats to an orderly life.
Public safety is built as much on the everyday routines of work and family as it
is on police and prisons. Any retrenchment of the penal system therefore must recognize
how deeply the prison boom is embedded in the structure of American social inequality.
Ameliorating these inequalities will be necessary to set us on a path away from
mass incarceration and toward a robust, socially integrative public safety.