server05productnCCPP5-4CPP403.txt unknown Seq 1 13-OCT.docx

server05productnCCPP5-4CPP403.txt unknown Seq: 1 13-
OCT-06 12:41
PUBLIC PREFERENCES FOR
REHABILITATION VERSUS
INCARCERATION OF JUVENILE
OFFENDERS: EVIDENCE FROM A
CONTINGENT VALUATION SURVEY*
DANIEL S. NAGIN
Carnegie Mellon University
ALEX R. PIQUERO
University of Florida
ELIZABETH S. SCOTT
Columbia University
LAURENCE STEINBERG
Temple University
Research Summary:
Accurately gauging the public’s support for alternative
responses to
juvenile offending is important, because policy makers often
justify
expenditures for punitive juvenile justice reforms on the basis
of popu-
lar demand for tougher policies. In this study, we assess public
support
for both punitively and nonpunitively oriented juvenile justice
policies

by measuring respondents’ willingness to pay for various policy
pro-
posals. We employ a methodology known as “contingent
valuation”
(CV) that permits the comparison of respondents’ willingness to
pay
(WTP) for competing policy alternatives. Specifically, we
compare CV-
based estimates for the public’s WTP for two distinctively
different
responses to serious juvenile crime: incarceration and
rehabilitation.
An additional focus of our analysis is an examination of the
public’s
WTP for an early childhood prevention program. The analysis
indi-
cates that the public is at least as willing to pay for
rehabilitation as
punishment for juvenile offenders and that WTP for early
childhood
prevention is also substantial. Implications and future research
direc-
tions are outlined.
* Authors are listed in alphabetical order. This research was
supported by the
MacArthur Foundation Research Network on Adolescent
Development and Juvenile
Justice. Address all correspondence to Alex R. Piquero,
Department of Criminology,
Law & Society, University of Florida, 201 Walker Hall, P.O.
Box 115950, Gainesville,
FL. 32611-5950 (e-mail: [email protected]).
VOLUME 5 NUMBER 4 2006 PP 627–652 R

OCT-06 12:41
628 NAGIN, PIQUERO, SCOTT, & STEINBERG
Policy Implications:
The findings suggest that lawmakers should more actively
consider pol-
icies grounded in rehabilitation, and, perhaps, be slower to
advocate
for punitive reforms in response to public concern over high-
profile
juvenile crimes. Additionally, our willingness to pay findings
offer
encouragement to lawmakers who are uncomfortable with the
recent
trend toward punitive juvenile justice policies and would like to
initiate
more moderate reforms. Such lawmakers may be reassured that
the
public response to such initiatives will not be hostile. Just as
impor-
tantly, reforms that emphasize leniency and rehabilitation can
be justi-
fied economically as welfare-enhancing expenditures of public
funds.
The evidence that the public values rehabilitation more than
increased
incarceration should be important information to cost-conscious
legis-
lators considering how to allocate public funds. Cost-conscious
legisla-
tures may become disenchanted with punitive juvenile justice
policies

on economic grounds and pursue policies that place greater
emphasis
on rehabilitation. They may be reassured, on the basis of our
findings,
that the public will support this move.
KEYWORDS: Public Opinion, Punishment, Rehabilitation,
Juvenile Jus-
tice, Prevention, Crime Policy
Over the past few decades, American juvenile justice policy has
become
progressively more punitive, as evidenced by the increasingly
harsh nature
of the dispositions imposed on juveniles who have been
adjudicated delin-
quent or guilty, as well as by the marked increase in the number
of states
in which juveniles can be tried as adults (Bishop, 2000; Scott
and Stein-
berg, 2003). During the 1990s, in particular, legislatures across
the country
enacted statutes under which growing numbers of youths can be
prose-
cuted in criminal courts and sentenced to prison (Reppucci,
1999; Scott,
2000; Snyder and Sickmund, 1995). Indeed, today, in almost
every state,
youths who are 13 or 14 years of age (or less) can be tried and
punished as
adults for a broad range of offenses, including nonviolent
crimes (Griffin
et al., 1998; Sickmund, 2003). Even within the juvenile system,
punish-
ments have grown increasingly severe (Bishop, 2000; Fagan and
Zimring,

2000).
It is generally accepted that intense public concern about the
threat of
youth crime has driven this trend, and that the public supports
this legisla-
tive inclination toward increased punitiveness (Roberts, 2004).
And yet, it
is not clear whether this view of the public’s attitude about the
appropriate
OCT-06 12:41
PUBLIC PREFERENCES FOR REHABILITATION 629
response to juvenile crime is accurate. On the one hand, various
opinion
surveys have found public support generally for getting tougher
on juve-
nile crime and punishing youths as harshly as their adult
counterparts
(Bureau of Justice Statistics [BJS] Sourcebook, 2003; Moore,
1994; Soler,
2001). At the same time, however, scrutiny of the sources of
information
about public opinion reveals that the view that the public
supports adult
punishment of juveniles is based largely on either responses to
highly pub-
licized crimes such as school shootings or on mass opinion polls
that typi-
cally ask a few simplistic questions (Cullen et al., 2000;
Roberts and

Stalans, 1997). Moreover, evidence from recent research that
seeks to
probe more deeply into adults’ attitudes toward juvenile crime
suggests
that the public response may be more complex than political
rhetoric sug-
gests (Roberts et al., 2003; Schiraldi and Soler, 1998; Stalans
and Henry,
1994). For example, several surveys have found public support
for rehabili-
tation as a goal of juvenile justice policy (Moon et al., 2000;
Roberts, 2004)
and for sanctions and programs that are alternatives to prison
(Krisberg
and Austin, 1993). One survey found that participants thought
that school
discipline, rather than imprisonment, was the best way to reduce
juvenile
crime (Hough and Roberts, 2003). It is plausible that
assessments of public
sentiment about juvenile crime, and the appropriate response to
it, vary
greatly as a function of when and how public opinion is gauged.
An assessment of the public’s support for various responses to
juvenile
offending is important because policy makers often justify
expenditures
for punitive juvenile justice reforms on the basis of popular
demand for
tougher policies. Punitive responses to juvenile crime (e.g., the
incarcera-
tion of juvenile offenders in correctional facilities) are far more
expensive
than less harsh alternatives (e.g., providing juvenile offenders
rehabilita-

tive services in community settings). Furthermore, there is little
evidence
that these more punitive policies are more effective in deterring
future
criminal activity, and some evidence (Bishop et al., 1996) that
overly puni-
tive responses, such as the incarceration of juvenile offenders in
adult
facilities, actually may increase juvenile offending (Fagan,
1997). If politi-
cians’ misreading of public sentiment has led to the adoption of
more
expensive policy alternatives than the public actually wants, tax
dollars are
likely being wasted on policies that are costly and possibly
ineffective, and
that may be less popular than is widely assumed.
In this study, we assess public opinion toward juvenile justice
policy
using an approach that differs from conventional polling, by
measuring
respondents’ willingness to pay (WTP) for various policy
proposals. We
employ a methodology known as “contingent valuation” (CV),
which per-
mits the comparison of respondents’ willingness to pay for
competing pol-
icy alternatives. In our judgment, this approach has three
principal
advantages over conventional public opinion polling.
OCT-06 12:41

First, asking how much respondents as individual taxpayers are
willing
to pay for a specific policy likely yields a more accurate
estimate of their
attitude toward that policy than merely asking whether they
approve or
disapprove of it, because the question requires the respondent to
consider
the cost of the policy as well as its benefits. One of the
shortcomings of
most public opinion polling about policy options is that the
questions
posed seldom situate the hypothetical alternatives in a concrete
economic
context. It is far easier to endorse a particular policy when it is
proposed in
the abstract (e.g., “Do you favor expanding the city’s sanitation
services in
order to clean the streets more frequently?”) than when one is
told the
actual cost of that policy (e.g., “Do you favor expanding the
city’s sanita-
tion services in order to clean the streets more frequently, at an
annual
cost to the city of $1 million per year?”) or what the impact of
that policy
would be on the respondent’s personal tax burden (“Would you
be willing
to pay an additional $100 in property taxes annually in order to
expand the
city’s sanitation services and clean the streets more
frequently?”). As a
consequence, conventional polls may indicate more enthusiastic

public
support for a potentially expensive policy than would likely be
the case if
the actual cost burden of the policy were revealed. Although
asking a
respondent how much he or she would be willing to pay for a
given pro-
gram or intervention is not the same as a formal referendum in
which
respondents are asked to vote up or down a policy where the
cost to the
individual taxpayer is specified in advance, the approach
employed in the
current study likely yields a more accurate estimation of public
opinion
than does conventional polling.
Second, the CV methodology employed here permits a more
direct
comparison of public attitudes toward different policies
designed to
address the same fundamental problem. In conventional opinion
polling,
respondents’ preference for one versus another policy is often
ascertained
(e.g., “Do you favor Policy A or would you prefer Policy B?”),
but the
phrasing of such comparative questions seldom provides
respondents with
information on the relative effectiveness or cost of the proposed
options.
Without knowing what the respondent believes to be the
effectiveness or
cost of each alternative, one is unable to know what the
respondent’s
answer genuinely reflects. Imagine, for example, how different

one’s
responses to a question contrasting Policy A and Policy B might
be if one
were told that the first option had been shown to be only half as
effective
as the second, or that the second cost five times as much as the
first. In the
current study, we use an experimental methodology that permits
us to
compare respondents’ opinions about policy alternatives that are
presented as equally effective. Any observed differences in
respondents’
willingness to pay for two policies of equal effectiveness must
necessarily
indicate a true preference for one over the other.
OCT-06 12:41
The third advantage of the CV methodology is that it permits
one to
calculate a rough estimate of the economic value of a given
policy to the
public. Generally speaking, crime policy is seldom formulated
on the basis
of careful assessments of the economic costs and benefits of
different pol-
icy options or by taking into account their relative value to the
public
(Cook and Ludwig, 2000; Nagin, 2001; Zimring and Hawkins,
1995).
Increasingly, however, legislators want to know whether the

economic
benefits, measured in dollar terms, of a given policy outweigh
its actual
costs. But whereas estimating the costs of a given crime policy
(e.g., man-
dating that correctional facilities provide drug treatment for
individuals
with drug abuse or dependency problems) is often possible (i.e.,
one can
multiply the actual cost of providing treatment by the number of
inmates
with drug problems), estimating the economic benefits of the
same policy
is a different matter, because many such benefits are intangible
(e.g.,
increases in the public’s feelings of safety), and assigning a
monetary value
is difficult. As a consequence, it is often impossible to compute
a
cost–benefit ratio for various policy alternatives based on an
assessment of
the tangible benefits of the policy.
The CV methodology overcomes this problem by estimating the
eco-
nomic value of various policies in the most straightforward way
possible:
by determining how much individuals are willing to pay for
each of them.
This approach differs from conventional economic analyses that
focus
solely on the tangible benefits of interventions (e.g., estimates
of the num-
ber of days of added employment one could gain by successfully
treating
individuals with drug addiction) because it permits the

assessment of
intangibles that are difficult to value economically but that are
neverthe-
less important in the overall assessment of the value of an
intervention
(e.g., the added safety that individuals feel knowing that fewer
individuals
with drug problems are living in their neighborhood). The CV
approach
finesses the difficulties inherent in trying to build an estimate
of total ben-
efits by estimating WTP for each of the component benefits by
asking
respondents for WTP for total package of benefits that may
attend a par-
ticular policy. Thus, if the average taxpayer in a given state is
willing to pay
an additional $100 in taxes each year to implement a particular
policy,1
and there are 5 million taxpayers in that state, the annual value
of that
policy to the taxpayers in that state is $500 million based on the
WTP
criterion. A comparison of this figure with the known cost of
the policy
produces a rough calculation of its cost–benefit ratio and, if
multiple poli-
cies are studied, their relative cost–benefit ratios. Thus, a policy
promises
1. Of course, because there is no real measure of the value of
various policies,
what is available are public perception data.

OCT-06 12:41
to offer substantial value if the public is willing to pay more
than its actual
cost.
The CV methodology has been widely used in the study of other
policy
arenas (Mitchell and Carson, 1989), but only recently has it
been used in
the criminal justice context to estimate the value of crime-
control pro-
grams in general (Cohen et al., 2004), the value of violent crime
preven-
tion (Atkinson et al., 2005), and such interventions as drug
abuse
treatment programs (Zarkin et al., 2000) and gun control
policies (Cook
and Ludwig, 2000; Ludwig and Cook, 2001) in particular.
The current study employs a CV methodology to compare public
atti-
tudes toward two distinctively different responses to serious
juvenile
crime: incarceration and rehabilitation. The study was carried
out in Penn-
sylvania, a state that is a mixture of urban, suburban, and rural
communi-
ties and that is fairly representative of the U.S. population with
respect to
its political climate (in the 2000 Presidential election,
Pennsylvanians, like
the country as a whole, split nearly evenly between Bush and

Gore).
Importantly, however, Pennsylvania’s crime rates differ
significantly from
those in other parts of the country in ways that might make
residents more
punitive in their preferred response to violent crime. Although
Penn-
sylvania has one of the lowest rates of juvenile property crime
in the coun-
try (1,222 per 100,000 population, compared with the U.S.
average of 1,442
per 100,000), its rate of violent juvenile crime is among the
highest in the
nation (402 per 100,000 in Pennsylvania, compared with the
U.S. average
of 291 per 100,000).
DATA AND METHODS
Telephone interviews were conducted with a random sample of
Penn-
sylvania households (adults over the age of 18) between March
2005 and
August 2005. Individuals, in either English or Spanish, were
selected as
respondents within each household according to the following
script based
on the random sample selection procedure:
Hello, my name is __. I’m calling from the University of
Florida. This
is not a sales call. We are conducting academic research about
crime
in Pennsylvania. This research is being conducted by the
University of
Florida in collaboration with Temple University and we would

like
your opinion. First, I need to know if you are (under 18 years
old or)
18 years old or older. If not, may I speak to someone 18 years
old or
older who lives there? According to the research method being
used
by the University, I have to ask some questions of the ADULT
(age
18 or older) who had the most recent birthday who currently
resides
there. Your phone number was selected at random by computer,
and
only your first name will be used to insure confidentiality. You
do not
OCT-06 12:41
have to answer any question you do not wish to answer. I also
want
you to know that this call may be recorded for quality control
pur-
poses. It should take only 7-10 minutes. May I have your first
name?
The survey followed this brief introduction.
A random digit dial was conducted with an original sample of
7,570 tele-
phone numbers. Of these, 4,231 were ineligible
(business/government, fax,

etc., n = 3,390; language or mental inability, n = 84; answering
machine, n
= 748; and respondent never available, n = 9), leaving an
eligible sample of
3,339. Of these eligible numbers, 1,837 refused leaving a
completed sample
of 1,502. Thus, the response rate, out of the eligible number of
3,339, is
given by 1,502/3,339, or 44.98%, and it is comparable with that
reported in
other similar contingent valuation studies (see Cohen et al.,
2004). With
respect to race and sex, the sample closely mirrored the state’s
population.
Specifically, 86.7% of the sample was white, and 59.7% of the
sample was
female; according to 2000 census data, 85.4% of the state’s
population is
white and 51.7% female. Fifty-percent of the study sample
reported an
income over $50,000, and 50% reported at least some college
experience,
again comparable with the state as a whole. The average age of
the
respondents was 50.18 (range 18–94; median 50).
A survey was developed to examine respondents’ WTP for
rehabilita-
tion and incarceration of juvenile offenders. The survey
instrument was
drafted using an extensive design process that also included
pretesting
among young adults. Additionally, we followed the guidelines
established
by the National Oceanic and Atmospheric Administration
(NOAA) for

studies employing the contingent valuation methodology, and
we modeled
our approach after prior contingent valuation research in
criminology
more specifically (Cohen et al., 2004; Ludwig and Cook, 2001).
The aver-
age time to complete the survey was just under nine minutes.
Respondents were presented with several hypothetical scenarios
and
numerous questions about their background and attitudes. The
basic sur-
vey was the same for all individuals, with one important
exception. One
item, which asked respondents if they would be willing to vote
for a crime
policy proposal requiring each household to pay an additional
amount of
money in taxes, was systematically varied. Half of the sample,
randomly
selected, responded to a proposal to increase the amount of
rehabilitative
services provided to violent juvenile offenders, without any
increase in
their time incarcerated, whereas the other half of the sample
responded to
a proposal to increase the amount of time violent juvenile
offenders were
incarcerated for their crime, without the addition of any
services. Other-
wise, the wording of the two proposals was nearly identical, in
order to
compare responses to each of them.
The text of the added rehabilitation question was as follows:

OCT-06 12:41
Currently in Pennsylvania, juvenile offenders who commit
serious
crimes such as robbery are put in jail for about one year.
Suppose
Pennsylvania citizens were asked to approve the addition of a
rehabil-
itation program to the sentence for these sorts of crimes.
Similar pro-
grams have reduced youth crime by 30%. Youths in these
programs
are also more likely to graduate from high school and get jobs.
If the
change is approved, this new law would cost your household an
addi-
tional $100 per year in taxes.2
After reading this question, respondents were asked: “Would
you be
willing to pay the additional $100 in taxes for this change in the
law?”
Respondents who indicated “yes” were asked an additional
follow-up
question: “Would you be willing to pay $200 for the same
change?”
Respondents who indicated “no” to the original question also
were asked
an additional follow-up question: “Would you be willing to pay
an addi-
tional $50 for this change?” Response options to all questions

were “Yes”
and “No.”
The text of the added incarceration question was nearly
identical:
Currently, in Pennsylvania juvenile offenders who commit
serious
crimes such as robbery are put in jail for about one year.
Suppose
Pennsylvanians were asked to vote on a change in the law that
would
increase the sentence for these sorts of crimes by one additional
year,
making the average length of jail time two years. The additional
year
will not only impose more punishment but also reduce youth
crime by
about 30% by keeping juvenile offenders off the street for
another
year. If the change is approved, this new law would cost your
house-
hold an additional $100 per year in taxes.3
2. The 30% crime reduction figure was obtained from Lipsey’s
(1992) meta-anal-
ysis findings regarding the effect of rehabilitation.
3. Three other points regarding the scenarios are in order. First,
we retained the
30% crime reduction estimate and the $100 dollar amount so as
to maintain rough
comparability with the rehabilitation-added scenario presented
to the other half of the
sample. Second, the rehabilitation-added scenario also suggests
that employment and

educational benefits may result from the expenditure, benefits
that do not follow from
additional incarceration. This statement is based on research
findings that indicate that
rehabilitation programs often provide additional non-crime
benefits (Cullen and Gen-
dreau, 2000). In our discussion of the findings, we examine the
implications of the dif-
ference between the two scenarios. Third, two small matters
might bias the WTP
between punishment and rehabilitation. With regard to the
rehabilitation scenario, the
respondents are asked to consider a proposal to approve the
addition of a rehabilitation
sentence for particular sorts of crimes. They are not specifically
told that the incarcera-
tion time would not increase. It seems possible, therefore, that
at least some respon-
dents interpreted “addition” to mean extended supervision of
some sort (i.e., a longer
sentence) under which a treatment program would be provided.
Thus, for these respon-
dents, rehabilitation and additional punishment would be at
least somewhat conflated.
OCT-06 12:41
The same follow-up questions were asked of respondents who
received
the incarceration scenario as were asked of respondents who
were

presented with the rehabilitation scenario.
Many other questions in the survey were included to validate
respon-
dents’ answers to the WTP question. For instance, to gauge
their overall
preference for punitive responses to juvenile crime, all
respondents were
presented with the following question:
Jason, who is 15 years old, is convicted of robbing a
convenience
store. While a gun was used, nobody was injured. Jason has no
history
of arrests for violent crimes but has been previously arrested for
steal-
ing. Do you think Jason should be sent to jail?
Respondents who indicated that Jason should be sent to jail
were
presented with a follow-up question about how long, in years,
they
believed Jason’s sentence should be.
Similarly, all respondents were presented with one scenario
designed to
gauge their interest in spending additional tax dollars for an
early child-
hood prevention program modeled after a nurse home visitation
program
developed by Olds et al. (1998). The text of the question reads
as follows:
The state of Pennsylvania is considering starting a program of
home
visits by nurses to young mothers in which nurses encourage

healthy
behaviors during pregnancy, good parenting, and the mother’s
own
personal development in terms of education and work. This
program
has been found to reduce the child’s later involvement in crime
and
also cut their use of alcohol during adolescence. In addition, it
cuts
welfare use of the women themselves and reduces the chances
of their
abusing their children. Would you be willing to pay the
additional
$150 in taxes for this change in the law?4
As with the other WTP question, follow-up questions were
asked
depending on the respondent’s initial response. Respondents
who indi-
cated “yes” were asked: “Would you be willing to pay $300 for
the same
change?” Respondents who indicated “no” to the original
question were
asked: “Would you be willing to pay an additional $75 for this
change?”
Second, and suggesting a bias in the opposite direction, the
effectiveness of the punish-
ment policy may be overstated. The extension of sentences by
one year would be
applied to all. Thus, we are not talking about selective
incapacitation. Our goal was to
hold the effectiveness of the hypothetical policy constant across
the two versions of the
survey, even though there is evidence that rehabilitation
outstrips punishment programs

in reducing recidivism. We would like to thank a reviewer for
pointing this out.
4. These were all findings reported in the Olds et al. (1998)
evaluation.
server05productnCCPP5-4CPP403.txt unknown Seq: 10
13-OCT-06 12:41
RESULTS
WTP ESTIMATES
Table 1 arrays WTP by four bid levels: (1) those who said no to
$100 and
no to $50 (to be conservative in our estimate of WTP, coded
$0); (2) those
who said yes to $50 but no to $100 (coded $50); (3) those who
said yes to
$100 but no to $200 (coded $100); and (4) those who said yes to
$100 and
yes to $200 (coded $200).5
TABLE 1. WTP AT BID LEVELS BY CONDITION
Bid Rehab Punish Nurse Visit
No 50–No 100 27.8% 40.8%
Yes 50–No 100 6.9% 7.4%
Yes 100–No 200 36.0% 26.3%
Yes 100–Yes 200 29.4% 25.5%
No 75–No 150 35.0%

Yes 75-No 150 8.4%
Yes 150–No 300 33.7%
Yes 150–Yes 300 23.0%
Average WTP $98.10 $80.97 $125.71
Number of 712 699 1442
respondents
With regard to the rehabilitation-added scenario, 27.8% of the
respon-
dents were unwilling to pay for the service, whereas the rest
were willing
to pay at least $50. As shown, over 60% of the respondents who
received
the rehabilitation-added scenario were willing to pay at least
$100 for the
program. With regard to the punishment-added scenario, 40.8%
of the
respondents were unwilling to pay for the service, a much
higher percent-
age compared with the rehabilitation-added scenario and a
difference that
is significant at p < 0.01. Also, a little over 50% of the
respondents who
received the punishment-added scenario were willing to pay at
least $100
for the program. As for the nurse home visitation program, 65%
of
5. It is important to note here that these amounts may
underestimate some par-
ticipants’ willingness to pay for rehabilitation or punishment.
Presumably, there are
people who would spend more than $200 (the highest figure we
offer as an option for
the rehabilitation and punishment scenarios) and there are
people who might spend

somewhere between $0 and $50 (we score anyone who says
“no” to $50 as being willing
to spend nothing). Thus, the estimates are conservative because
we only know that
respondents would be willing to pay “at least” XX dollars.
Other approaches could use
the midpoint as a WTP.
13-OCT-06 12:41
respondents were willing to pay at least $75 for the program and
56.7%
were willing to pay $150 or more for the program.
Perhaps the most interesting finding from Table 1 concerns the
average
WTP for the rehabilitation-added and punishment-added
scenarios. Here
it can be seen that the average WTP was almost $20 higher for
the addi-
tion of rehabilitation services, $98.10/household, than for the
addition of
an extra year of incarceration, $80.97. This difference is
significant at p <
0.01. It is also noteworthy that the average WTP is still higher
for the
nurse visitation program, $125.71, although the different
response metric
used for this scenario cautions against a direct comparison with
the other
two.

Further analysis indicated that African-American participants
were
more willing to pay for rehabilitation than whites ($102.35 vs.
$97.52) and
less willing to pay for incarceration ($59.31 vs. $84.67).
Women were more
willing to pay for rehabilitation than were men ($102.01 vs.
$92.39). WTP
for both rehabilitation and incarceration generally increased
with income,
as economic theory (Viran, 1992) would predict. For example,
for house-
holds with income greater than $100,000 average WTP for
rehabilitation is
$124.65, whereas for households with income less than $25,000,
it is $85.56.
This correlation between income and WTP increases our
confidence that
expressed WTP is a reflection of real preferences.
Several other findings also increase our confidence that the
responses to
the hypothetical contingent valuation scenarios were based on
actual pref-
erences. We expected that compared with conservatives and
more puni-
tively oriented respondents, liberals and less punitively oriented
respondents would be more supportive of rehabilitation and less
support-
ive of punishment. This is precisely what we found. Support for
rehabilita-
tion was stronger among participants who identified themselves
as liberal
than among those who identified themselves as conservative
($131.47 vs.

$84.11). Not surprisingly, differences were also found between
those who
favored a noncustodial response to the vignette asking the
appropriate dis-
position for the youth who committed an armed robbery over
those who
favored prison. Respondents who recommended a prison
sentence
reported significantly higher average WTP for punishment than
respon-
dents who recommended a noncustodial sentence, $98.16 vs.
$46.83, and
significantly lower average WTP for rehabilitation, $94.25 vs.
$111.29.
Finally, respondents who believe that sending juveniles to jail is
more
effective than rehabilitation have higher WTP for jail than
respondents
with the converse expectations, $88.90 vs. $72.71, and
comparatively lower
WTP for rehabilitation, $81.85 vs. $110.02.6
6. With the exception of the white/black difference in WTP for
rehabilitation, all
13-OCT-06 12:41
Although respondents differ in their responses depending on
their polit-
ical philosophy and attitudes toward punishment, our results
suggest broad

public support for effective rehabilitation. Even more punitively
oriented
respondents expressed substantial WTP for rehabilitation.
Similarly,
although self-identified conservatives reported significantly
higher WTP
for punishment than self-identified liberals, $86.29 vs. $62.76,
and signifi-
cantly lower WTP for rehabilitation, conservatives as well as
liberals
expressed substantial support for public investment in effective
rehabilitation.
COST–BENEFIT ANALYSIS
In this section, we use the results of the contingent valuation
survey to
conduct a “first-cut” cost–benefit analysis. We caution against
placing too
great a weight on the specific calculations because participants
in our sur-
vey responded to a hypothetical question; nonetheless, we
believe the
cost–benefit analysis is informative, particularly regarding the
important
question for juvenile justice policy of striking the right balance
between
the punishment and rehabilitation of offenders (e.g., Moon et
al., 2000),
and for correctional policy decision makers who are faced with
the task of
allocating scarce financial resources (Caldwell et al., 2006).
Dollar cost estimates provide a useful means of comparing
different
types of crime prevention efforts and “reasonable minds can and

do differ
over how best to conduct cost-benefit analyses of [correctional]
policies,
how best to implement the results of such analyses, and how, if
at all, to
fashion or re-orient public policies accordingly” (DiIulio,
1990:51). As do
others, we recognize that the comparison of the benefits and
costs of alter-
native crime-control policies is controversial, but nevertheless
useful and
important. One advantage of cost–benefit analysis and its use of
dollars as
a common metric for analyzing criminal justice policy is that
society
spends dollars to try to prevent crimes (Cohen, 2005:6). The
key question,
of course, is whether the reduced (increased) crime as a
function of differ-
ent crime-control policies is worth its cost. Cohen has
convincingly argued
that a compelling reason to attempt a cost–benefit analysis is
the conse-
quence of not doing so (pp. 6–7):
Whenever a criminal justice or prevention program is adopted
or not
adopted, society is implicitly conducting a benefit-cost analysis
and placing
dollar values on crimes. For example, suppose one program
costs $1 mil-
lion and ultimately will prevent 100 burglaries from occurring.
Whether
made explicit or not, the policymaker adopting that program has
deter-
mined that it is worth spending at least $10,000 to reduce each

burglary ($1
the differences reported in this paragraph and the prior
paragraph are significant at the
p<.01 level.
13-OCT-06 12:41
million divided by 100 burglaries). If another $1 million
program that was
not funded would have prevented 50 serious physical assaults
from occur-
ring, the policymaker is implicitly determining that each assault
is worth
less than $20,000 ($1 million divided by 50). Thus even the
policymaker
who has ethical concerns about placing dollar values on crime
and con-
ducting benefit-cost analysis implicitly makes a value judgment
about the
monetary value of crime.
In short, although cost–benefit analysis does have its
limitations, which
we recognize and admit, we nevertheless believe that the
information
gleaned from such an exercise is a useful piece of knowledge.
The estimates of WTP for incarceration, rehabilitation, and
early pre-
vention at the level of the household provide the basis for

calculating
statewide WTP for each of these programs. According to the
2000 U.S.
Census, there were 4.78 million households in Pennsylvania.
Based on this
scale factor, Table 2 translates our contingent valuation-based
estimates of
average WTP per household into statewide WTP. We should
note again
that the WTP estimates for the nurse visitation program cannot
be directly
compared with those for the other two scenarios, because the
response
options were not the same.
TABLE 2. PROGRAMS BENEFITS
Program Ave. WTP per Statewide WTP
household per year per year
Rehabilitation $98.1 $468 mil.
Longer sentence $80.97 $387 mil.
Nurse visitation $125.71 $601 mil.
In the nomenclature of cost–benefit analysis, statewide WTP
measures
the total dollar value of the benefits of these programs as
perceived by a
representative sample of Pennsylvanians. The contingent value
methodol-
ogy is not designed to provide an accounting of the relative
contributions
of various types of perceived benefits of rehabilitation or
incarceration
that contribute to respondents’ WTP. Given the respondents
willingness to

pay more for the same reduction in crime achieved via
rehabilitation as
that achieved through incarceration, it seems safe to presume
that even if
crime reduction is the largest perceived benefit of
rehabilitation, other
types of benefits such as social productivity of increased
employment and
individual welfare of affected youths likely contribute as well.
In the case
of the longer sentence scenario, respondents likely valued
retribution and
increased public protection in the longer period of
incarceration.
The second key ingredient to a cost–benefit analysis is cost
estimation.
13-OCT-06 12:41
Total annual cost is calculated by multiplying an estimate of the
annual
cost per individual in the target population by an estimate of
size of the
target population. Table 3 reports this calculation for each
program type.
TABLE 3. PROGRAM COSTS
Program Ave. Cost per Size of Target Total Cost
Person-Year Population

Rehabilitation $10,000 2,000 $20 mil.
Longer sentence $50,000 2,000 $100 mil.
Nurse visitation $3,000 100,000 $300 mil.
Consider first the estimates of the annual cost per person. A
2003 study
by the Washington State Institute for Public Policy reports the
cost of a
great variety of treatment programs for juvenile offender
programs (Aos
et al., 2004). The costs vary widely. The three most expensive
rehabilita-
tion programs are multisystemic therapy ($5,681), mentoring in
the Juve-
nile Justice System ($6,471), and Intensive Parole ($5,992). To
be
conservative in our estimation of the benefits of rehabilitation
relative to
cost, we use what we believe is a high-end cost annual estimate
of $10,000
per person. A bulletin from Pennsylvania’s Department of
Public Welfare
(2004) reports per diem rates for confinement in various types
of secure
facilities for juveniles. The average for 2004 was $306/day,
which translates
into an annual cost of $111,000. Here again we err on the side
of caution,
but in this case, to avoid overstating the benefit-to-cost ratio of
rehabilita-
tion relative to incarceration, we use an annual cost estimate of
$50,000
per juvenile. The Washington State study reports that the total
cost of the
Olds’ Nurse Family Partnership for Low Income Mothers is

$9,118 per
child. This cost covers about three years of service, the nine
months of
pregnancy plus two years of follow-up. Therefore, we base our
annualized
cost calculation on an estimate of $3000/year per child.
We turn now to our estimates of the size of the target
population of
youths who would receive added rehabilitation or punishment.
Penn-
sylvania does not report data on the number of juveniles who
are incarcer-
ated for committing serious violent crimes of the type, robbery,
which was
the subject of our WTP scenarios on lengthened sentences and
rehabilita-
tion. In 2003, Pennsylvania juvenile courts placed 5,701
juveniles in facili-
ties outside their home (Pennsylvania Juvenile Court Judge’s
Commission,
2003). However, most of these placements were not to prison-
like facili-
ties. For example, 506 were placed in facilities for drug and
alcohol treat-
ment and 806 were placed in group-homes. Placement in secure
facilities
and boot camps probably comes closest to the dispositions
received by our
13-OCT-06 12:41

target population. In 2003, these totaled 1,234. Another 642
were “wilder-
ness-based” placements. Again to be conservative in our
estimation of
benefit-to-cost ratios, we assumed 2,000 individuals per year
would be the
targets of the enhanced sentence and rehabilitation programs.
As for the nurse visitation program, census data for
Pennsylvania indi-
cate that about 75,000 children two years old or younger live in
households
below the poverty line. This statistic suggests that 35,000 low-
income
Pennsylvania women are pregnant each year. We, thus, estimate
that with
a 100% participation rate, about 100,000 children, either born
or in utero,
would be enrolled in the nurse visitation program. A 100% take-
up rate is,
of course, unrealistic. However, if we were to use a lower rate,
it would
seem only reasonable to factor down the estimated benefits by
the same
factor. As a result, the cost-to-benefit ratio would be
unaffected.
Combining the benefits and costs in Tables 2 and 3 yields very
different
benefit-to-cost ratios by program type. For the rehabilitation
option, the
ratio of benefits to costs is 23.4 (=$468/$20). For the
lengthened incarcera-
tion option, the ratio is 3.87 (=$387/$100), and for the nurse
visitation

option, the ratio is 2.00. All imply that benefits as measured by
WTP sub-
stantially exceed costs. However, the estimated returns per
dollar spent
differ substantially.
For our purposes here, the difference in return between
incarceration
and rehabilitation is of particular interest. Both imply very large
returns,
but the difference in magnitude between rehabilitation, $23.4 in
benefit
per dollar spent versus $3.87 per dollar spent for incarceration
is striking.
This difference is largely attributable to the differences in cost
per person
between rehabilitation and imprisonment because the assumed
size of the
target population is the same in both sets of calculations. In
fact, had we
used the actual current annual cost of incarcerating a juvenile in
Penn-
sylvania based on per diem figures, which are well over
$100,000, the ben-
efit-to-cost ratio for incarceration would be more than halved.
We emphasize the pivotal role of cost (per offender) in
explaining the
difference between the benefit-to-cost ratios for incarceration
versus reha-
bilitation for two reasons. One is that the cost per person-year
estimates
used in the benefit-to-cost calculations are the least speculative
of all com-
ponents of the calculation. We can have some confidence in the
accuracy

of the costs described above of various rehabilitation programs
and of
incarceration of juvenile offenders, and there is no dispute that
rehabilita-
tion programs are far less costly than incarceration.
A second reason for the emphasis on cost per person is that our
esti-
mates of WTP for punishment and rehabilitation, used to derive
our esti-
mate of each policy’s benefit, are roughly comparable in
magnitude. As a
result, the difference in the benefit-to-cost ratio of these two
options is
mostly attributable to the denominator, cost. We judge this an
important
13-OCT-06 12:41
point because it makes clear that according to our contingent
valuation
survey, the public is at least as supportive of effective
rehabilitation of
juveniles as they are of punishment of the juveniles for their
crimes.
Although we acknowledge that the issue of the relative
effectiveness of
punishment and rehabilitation in crime control is far from
settled, the con-
tingent valuation results provide strong support for the
contention that the

public is willing to pay for effective rehabilitation, and it
resonates well
with the finding that more treatment-focused services offer a
better
cost–benefit than more severity-focused sanctions for
delinquent youths
(Fass and Pi, 2002).
DISCUSSION
This study surveyed 1,500 randomly selected Pennsylvania
adults to
elicit their opinions about policies responding to youth crime.
The survey
employed contingent valuation, an innovative methodology that
gauges
preferences on the basis of respondents’ willingness to pay for
public ben-
efits, often through a specified increase in taxes. In our study,
the alterna-
tive policies of increased incarceration or rehabilitation were
presented as
having equivalent effectiveness in reducing juvenile crime.
Participants
were initially asked whether they were willing to pay $100 in
increased
taxes; the amount was doubled or halved depending on whether
their
response was positive or negative.
We found that respondents on average expressed somewhat
greater
willingness to pay for rehabilitation ($98.10) than for longer
incarceration
($80.97) of youths charged with serious crimes—and even
greater willing-

ness to pay for an early childhood prevention program
($125.71). These
results suggest that the public generally is willing to pay for
programs that
promise to reduce youth crime—and more willing to support
and pay for
rehabilitation and prevention programs than for longer periods
of incar-
ceration (see also Cullen et al., 1998). To an extent, the
additional educa-
tional and employment benefits of rehabilitation may account
for the
greater willingness to pay for rehabilitation than for
incarceration. None-
theless, at a minimum, the study finds comparable support for
the two
policy responses to juvenile crime.7 It is noteworthy that even
individuals
who identified themselves as conservative or who supported
punitive poli-
cies in response to attitude questions and the robbery vignette
also indi-
cated substantial WTP for rehabilitation and prevention
programs. This
suggests that rehabilitation and prevention programs as policy
responses
7. It is not possible to evaluate how participants who responded
to the rehabilita-
tion scenario valued various benefits, but it seems likely that
crime reduction was
important. Moreover, the research indicates that rehabilitation,
in fact, offers the addi-
tional educational and employment benefits. See Footnote 2.

13-OCT-06 12:41
to youth crime have substantial public support across the
political spec-
trum, while at the same time evincing crime reduction much
more cost-
effectively than harsher punishments (Greenwood, 2006).
Our study also provides information that permits the comparison
of
rehabilitation and incarceration using cost–benefit analysis, a
standard
mode of policy analysis that has only recently begun to be used
in evaluat-
ing criminal justice policies (Caldwell et al., 2006). Based on
estimates of
the yearly costs per offender of incarceration and of
rehabilitation pro-
grams, and estimates of the number of young offenders
incarcerated in
Pennsylvania, we calculated a cost–benefit ratio for
incarceration and
rehabilitation. This analysis produces strikingly different cost–
benefit
ratios for these two policies, a difference that is the result of
two factors:
(1) A year of incarceration is far more expensive than a year of
rehabilita-
tion and (2) participants’ willingness to pay for the two policies
was
comparable.

Two related limitations of willingness-to-pay methodology
should be
emphasized, both of which speak to the amount of confidence
we can have
in participants’ responses—or, put another way, whether the
amount par-
ticipants stated they were willing to pay is a meaningful figure.
The first is
sometimes described as a demand characteristic or “anchoring.”
Partici-
pants were asked initially whether they would pay $100,
focusing their
attention on that amount as a baseline and, likely, influencing
their ulti-
mate responses. If they had first been asked if they were willing
to pay $25,
their ultimate WTP amount and, consequently, the average for
their sce-
nario may well have been different. The second limitation,
mentioned ear-
lier, is that participants understood that the inquiry was
hypothetical.
Thus, we cannot be sure that their responses would have been
the same in
a tax referendum, where their expressed willingness to pay
might effect
their actual tax obligation. For these reasons, the survey does
not demon-
strate that Pennsylvanians are ready to pay $468 million more
for rehabili-
tation programs or $387 million for incarceration. The actual
amount they
are willing to pay for the program described to them may have
been less
(or more).8

Nonetheless, the cost–benefit analysis provides useful
information for
comparing the publics’ willingness to pay for the alternative
policies. The
factors that might undermine our confidence about the
participants’
8. On this score, we did not explore the “reason” that the public
would be willing
to pay more for one policy over another when they have
equivalent crime-reduction
benefits. Is it because they believe there are other noncrime
benefits such as improved
high-school education/productivity; better long-term outcomes
that are not captured in
the 30% figure; they feel good about setting a youth on the right
course; they dislike
incarceration; etc? This is an important avenue for future
research. We would like to
thank Mark Cohen for this suggestion.
13-OCT-06 12:41
expressions about willingness to pay likely affect responses to
both reha-
bilitation and incarceration similarly. Importantly, there is no
reason to
think that anchoring would affect responses differently; if
participants
were asked initially whether they were willing to pay $25 for
either incar-

ceration or rehabilitation, the average willingness to pay amount
likely
would be affected for both options by a comparable amount.
What is
important in evaluating the two policies is that, although the
dollar
amount that taxpayers are willing to pay for either policy may
be uncer-
tain, participants are willing to pay at least as much for
rehabilitation—
probably somewhat more—as they are for incarceration. This
finding,
together with the external evidence that incarceration is
substantially
more costly than rehabilitation, supports the conclusion that the
returns
per dollar spent on increasing rehabilitation are a better value
than the
returns on increasing incarceration even if the benefit (based on
the pub-
lic’s willingness to pay) is less than the results indicate (see
also Cohen et
al., 2006; Greenwood, 2006). In other words, the benefit-to-cost
ratio for
rehabilitation may be lower than the $23.4 benefit per dollar
spent calcu-
lated on the basis of average WTP of $98.10, but our contingent
valuation
survey suggests it will still be favorable compared with the
cost–benefit
ratio for incarceration.
Our survey challenges the view held by many politicians and the
media
that the public opposes rehabilitation and favors incarceration
of young

offenders. According to conventional wisdom, the driving force
behind the
punitive reforms in recent years has been the public demand for
tough
juvenile justice policies, and politicians frequently point to
public outrage
at violent juvenile crime as justification for sweeping
legislative reforms.
Moreover, some earlier opinion surveys found public support
for policies
that punish juveniles as adults and pessimistic views about
rehabilitation
programs offered by the juvenile system. In contrast, our survey
suggests
that public attitudes about youth crime policy are more
complex. We inter-
pret our results to indicate that members of the public are
concerned
about youth crime and want to reduce its incidence, but they are
ready to
support effective rehabilitative programs as a means of
accomplishing that
end—and indeed favor this response to imposing more
punishment
through longer sentences (e.g., Applegate et al., 1997). This
notion is con-
sistent with findings suggesting that the public is willing to
support
nonpunitive goals such as rehabilitation sometimes to a greater
extent
than policy makers may realize (Gottfredson and Taylor, 1984),
and it is
complementary to Cohen et al.’s (2006) finding that the public
would pre-
fer to spend its next tax dollars on at-risk youth programs.

What explains the differences between our findings and the
conven-
tional view of public opinion about juvenile crime, including
the findings
13-OCT-06 12:41
of some earlier surveys? First, our survey was conducted
recently, and atti-
tudes may have changed since the 1990s, when some surveys
found puni-
tive public attitudes. Juvenile crime rates have declined
dramatically over
the past 10 years (Blumstein and Wallman, 2000), and the
public may real-
ize that the threat has subsided. In general, attitudes (and legal
policies)
toward youth are protective and paternalistic. Perhaps the public
is more
likely to express these traditional attitudes when they do not
perceive
young criminals as a threat.
This suggests another reason that politicians and the media may
have
distorted impressions about public attitudes and crime. Often
public opin-
ion on this issue is gauged when attention is focused on a high-
profile vio-
lent crime by a juvenile, such as a school shooting. In this
context, in part

because of intense media coverage, the salience and magnitude
of the
threat may become distorted in the public imagination. In
contrast, the
attitudes of our participants were probed in a neutral context. It
is plausi-
ble that their responses represent more stable policy preferences
than
those of individuals whose opinion is gauged in the midst of
media cover-
age of a horrendous crime.
POLICY IMPLICATIONS
Any suggestion that our study may offer lessons for policy
formation
must first acknowledge the survey’s limitations. We have
discussed the
possible sources of distortion in the willingness to pay amounts.
Beyond
this, the policy implications of the survey are limited by the fact
that it was
undertaken in a single state, Pennsylvania. A broader sample
from differ-
ent regions of the country might provide a more accurate
measure of pub-
lic opinion about juvenile crime. Nonetheless, as we discussed,
there is
good reason to think that Pennsylvania residents are likely to be
a rela-
tively good proxy for a national sample in their attitudes on this
issue; if
anything, relatively more punitive responses might be expected
in Penn-
sylvania, given its relatively higher rate of violent juvenile
crime (cf.

Baumer et al., 2003).
At a minimum, our findings suggest that lawmakers who are
concerned
about public opinion should consider policies grounded in
rehabilitation
and, perhaps, be slower to advocate for punitive reforms in
response to
public concern over high-profile juvenile crimes. Legislation
enacted in
this climate institutionalizes public fears that are likely short-
lived, and it
may result in laws that do not reflect stable public preferences
about youth
crime policy. Our study suggests that the political risk that
lawmakers face
in resisting public pressure during times of crisis is not as great
as they
might surmise. During calmer times, traditional paternalistic
attitudes
toward juveniles may exert a stronger influence on public
opinion—dis-
sipating enthusiasm for punitive policies.
13-OCT-06 12:41
Our WTP findings offer encouragement to lawmakers who are
uncom-
fortable with the recent trend toward punitive juvenile justice
policies and
would like to initiate more moderate reforms. First, such

lawmakers may
be reassured that the public response to such initiatives will not
be hostile
(e.g., Gottfredson and Taylor, 1984). Just as importantly,
reforms that
emphasize leniency and rehabilitation can be justified
economically as
welfare-enhancing expenditures of public funds. Rehabilitative
programs
are far less costly than incarceration and, often, when informed
by the
principles of effective intervention, more effective (Cullen and
Gendreau,
2000) and less harmful (Clear, 1994). The evidence that the
public values
rehabilitation more than increased incarceration should be
important
information to cost-conscious legislators considering how to
allocate pub-
lic funds.
Increasingly cost–benefit analysis is becoming a useful and
important
policy tool (McDougall et al., 2003) as policy makers seek to
maximize the
value obtained from tax dollars spent. In the realm of criminal
justice pol-
icy, cost–benefit analysis is relatively new (Caldwell et al.,
2006), but cost
concerns have become an increasingly important factor in legal
regulation.
The high cost of punitive sentencing guidelines has become a
considera-
tion in the public debate—long sentences translate into more
prison space,
more staff, and generally higher operating costs. In the past few

years,
legislatures in several states have reduced criminal sentences in
recogni-
tion of the high costs of incarceration (Barkow, 2005). Cost-
conscious leg-
islatures may become disenchanted with punitive juvenile
justice policies
on economic grounds and pursue policies that place greater
emphasis on
rehabilitation and early childhood prevention. If so, they may be
reas-
sured, on the basis of our findings, that the public will support
this move.
REFERENCES
Aos, Steve, Roxanne Lieb, Jim Mayfield, Marna Miller, and
Annie Pennucci
2004 Benefits and Costs of Prevention and Early Intervention
Programs for
Youth. Olympia, Wash.: Washington State Institute for Public
Policy.
Applegate, Brandon K., Francis T. Cullen, and Bonnie S. Fisher
1997 Public support for correctional treatment: The continuing
appeal of the
rehabilitative ideal. The Prison Journal 77:237–258.
Atkinson, Giles, Andrew Healey, and Susana Mourato
2005 Valuing the costs of violent crime: A stated preference
approach. Oxford
Economic Papers 57:559–585.

Barkow, Rachel E.
2005 Federalism and the politics of sentencing. Columbia Law
Review
105:1276.
13-OCT-06 12:41
Baumer, Eric, Steven F. Messner, and Richard Rosenfeld
2003 Explaining spatial variation in support for capital
punishment: A
multilevel analysis. American Journal of Sociology 108:844–
875.
Bishop, Donna
2000 Juvenile offenders in the adult criminal system. In
Michael Tonry (ed.),
Crime and Justice: A Review of Research, Volume 27. Chicago,
Ill.:
University of Chicago Press.
Bishop, Donna, Charles E. Frazier, Lonn Lanza-Kaduce, and
Lawrence Winner
1996 The transfer of juveniles to criminal court: Does it make a
difference?
Crime and Delinquency 42: 171–191.
Blumstein, Alfred and Joel Wallman

2000 The Crime Drop in America. New York: Cambridge
University Press.
Bureau of Justice Statistics
2003 Sourcebook of Criminal Justice Statistics. Washington,
D.C.: U.S. Depart-
ment of Justice.
Caldwell, Michael F., Michael Vitacco, and Gregory J. van
Rybroek
2006 Are violent delinquents worth treating? A cost-benefit
analysis. Journal of
Research in Crime and Delinquency 43:148–168.
Clear, Todd
1994 Harm in American Penology: Offenders, Victims, and
Their Communities.
Albany, N.Y.: State University of New York Press.
Cohen, Mark A.
2005 The Costs of Crime and Justice. London, U.K.: Routledge.
Cohen, Mark A. Roland T. Rust, and Sara Steen
2006 Prevention, crime control, or cash? Public preferences
toward criminal
justice spending priorities. Justice Quarterly. In press.
Cohen, Mark A., Roland T. Rust, Saran Steen, and Simon T.
Tidd
2004 Willingness-to-pay for crime control programs.
Criminology 42:89–110.

Cook, Philip J. and Jens Ludwig
2000 Gun Violence: The Real Costs. New York: Oxford
University Press.
Cullen, Francis T., Bonnie Fisher, and Brandon K. Applegate
2000 Public opinion about punishment and corrections. In
Michael Tonry (ed.),
Crime and Justice: A Review of Research, Volume 27. Chicago,
Ill.:
University of Chicago Press.
Cullen, Francis T. and Paul Gendreau
2000 Assessing correctional rehabilitation: Policy, practice, and
prospects. In
Julie Horney (ed.), Policies, Processes, and Decisions of the
Criminal
Justice System, Criminal Justice 2000. Washington, D.C.:
National Insti-
tute of Justice.
Cullen, Francis T. John P. Wright, Shayna Brown, Melissa M.
Moon, Michael B.
Blankenship, and Brandon K. Applegate
1998 Public support for early intervention programs:
Implications for a
progressive policy agenda. Crime and Delinquency 44:187–204.
13-OCT-06 12:41

DiIulio, Jr., John J.
1990 Crime and Punishment in Wisconsin: A Survey of
Prisoners and an
Analysis of the Net Benefit of Imprisonment in Wisconsin.
Wisconsin
Research Institute Report. Volume 3, Number 7. Milwaukee,
Wisc.
Fagan, Jeffrey
1997 The comparative advantages of juvenile versus criminal
court sanctions on
recidivism among adolescent felony offenders. Law and Policy
18:77–115.
Fagan, Jeffrey and Franklin E. Zimring, (eds.)
2000 The Changing Borders of Juvenile Justice: Transfer of
Adolescents to the
Criminal Court. Chicago, Ill.: University of Chicago Press.
Fass, Simon M. and Chung-Ron Pi
2002 Getting tough on juvenile crime: An analysis of costs and
benefits.
Journal of Research in Crime and Delinquency 39:363–399.
Gottfredson, Stephen D. and Ralph B. Taylor
1984 Public policy and prison populations: Measuring options
about reform.
Judicature 68:190–201.
Greenwood, Peter W.

2006 Changing Lives: Delinquency Prevention as Crime-Control
Policy. Chi-
cago, Ill.: University of Chicago Press.
Griffin, Patrick, Patricia Torbet, and Linda Szymanski
1998 Trying Juveniles as Adults in Criminal Court: An Analysis
of State
Transfer Provisions. Washington, D.C.: U.S. Office of Juvenile
Justice and
Delinquency Prevention.
Hough, Mike and Julian V. Roberts
2003 Institute for Criminal Policy Research Study. London,
U.K.: King’s
College School of Law.
Krisberg, Barry and James Austin
1993 Reinventing Juvenile Justice. Newbury Park, Calif.: Sage.
Lipsey, Mark
1992 Juvenile delinquency treatment: A meta-analytic inquiry
into the variabil-
ity of effects. In Thomas D. Cook, Harris Cooper, David S.
Cordray,
Heidi Hartmann, Larry V. Hedges, Richard J. Light, Thomas A.
Louis,
and Frederick Mosteller (eds.), Meta-Analysis for Explanation:
A
Casebook. New York: Russell Sage Foundation.
Ludwig, Jens and Philip J. Cook
2001 The benefits of reducing gun violence: Evidence from

contingent-
valuation survey data. The Journal of Risk and Uncertainty
22:207–226.
McDougall, Cynthia, Mark Cohen, Raymond Swaray, and
Amanda Perry
2003 The costs and benefits of sentencing: A systematic review.
Annals of the
American Academy of Political and Social Science 587:160–
177.
Mitchell, Robert C. and Richard T. Carson
1989 Using Surveys to Value Public Goods. Washington, D.C.:
Resources for
the Future.
Moon, Melissa, Jody Sundt, Francis T. Cullen, and John P.
Wright
2000 Is child saving dead? Public support for juvenile
rehabilitation. Crime and
Delinquency 46:38–60.
13-OCT-06 12:41
Moore, David W.
1994 Majority advocate death penalty for teenage killers.
GallupPoll Monthly

(September, pp.2-3).
Nagin, Daniel S.
2001 Costs and benefits of crime prevention. In Michael Tonry
(ed.), Crime
and Justice: A Review of Research, Volume 28. Chicago, Ill.:
University
of Chicago Press.
Olds, David L., C. R. Henderson, R. Cole, J. Eckenrode, H.
Kitzman, D. Luckey, L.
Pettitt, K. Sidora, P. Morris, and J. Powers
1998 Long-term effects of nurse home visitation on children’s
criminal and
antisocial behavior: 15 year follow-up of a randomized control
trial.
Journal of the American Medical Association 280: 1238–1244.
Pennsylvania Department of Public Welfare
2004 Children, Youth and Families (CYF) Bulletin #00-02-05
(Per Diem Rates/
Locations of Facilities). Harrisburg, Pa: Department of Public
Welfare.
Pennsylvania Juvenile Court Judge’s Commission
2003 Pennsylvania Juvenile Court Dispositions, 2003.
Harrisburg, Pa: Penn-
sylvania Juvenile Court Judges Commission.
Repucci, N. Dickon
1999 Adolescent development and juvenile justice. American

Journal of
Community Psychology 27:307–326.
Roberts, Julian V.
2004 Public opinion and youth justice. In Michael Tonry and
Anthony N. Doob
(eds.), Crime and Justice: A Review of Research, Volume 31,
Youth
Crime and Youth Justice, Comparative and Cross-National
Perspectives.
Chicago, Ill.: University of Chicago Press.
Roberts, Julian V. and Loretta J. Stalans.
1997 Public Opinion, Crime, and Criminal Justice. Boulder,
Colo.: Westview.
Roberts, Julian V., Mike Hough, David Indermaur, and Loretta
J. Stalans
2003 Penal Populism and Public Opinion: Lessons from Five
Countries.
Oxford, U.K.: Oxford University Press.
Schiraldi, Vincent and Mark I. Soler
1998 The will of the people? The public’s opinion of the violent
and repeat
juvenile offender act of 1997. Crime and Delinquency 44:590–
601.
Scott, Elizabeth S.
2000 The legal construction of adolescence. Hofstra Law
Review 29:547–598.

Scott, Elizabeth S. and Laurence Steinberg
2003 Blaming youth. Texas Law Review 81:799–840.
Sickmund, Melissa
2003 Juveniles in Court, 6-10. Washington, D.C.: U.S.
Department of Justice.
Soler, Mark I.
2001 Public Opinion on Youth, Crime, and Race: A Guide for
Advocates.
Retrieved May 30, 2005 on Building Blocks for Youth website:
www.buildingblocksforyouth.org.
13-OCT-06 12:41
Snyder, Howard and Melissa Sickmund
1995 Juvenile Offenders and Victims: National Report.
Washington, D.C.:
Office of Juvenile Justice and Delinquency Prevention.
Stalans, Loretta J. and Gary T. Henry
1994 Societal views of justice for adolescents accused of
murder. Law and
Human Behavior 18:675–696.
Varian, Hal R.
1992 Microeconomic Analysis, 3rd ed. New York: Norton.

Zarkin, Gary A., Sheryl C. Cates, and Mohan V. Bala
2000 Estimating the willingness to pay for drug abuse
treatment. Journal of
Substantive Abuse Treatment 18:149–159.
Zimring, Franklin E. and Gordon Hawkins
1995 Incapacitation: Penal Confinement and the Restraint of
Crime. New
York: Oxford University Press.
Daniel S. Nagin is Teresa and H. John Heinz III Professor of
Public Policy and Statis-
tics, Heinz School of Public Policy and Management, Carnegie
Mellon University. His
research focuses on the evolution of criminal and antisocial
behaviors over the life
course, the deterrent effect of criminal and non-criminal
penalties on illegal behaviors,
and the development of statistical methods for analyzing
longitudinal data. He is the
author of Group-based Modeling of Development (Harvard
University Press, 2005).
Alex R. Piquero is Professor of Criminology, Law & Society
and 2005 Magid Term
Professor in the College of Liberal Arts & Sciences at the
University of Florida, Mem-
ber of the National Consortium on Violence Research, and
Member of the MacArthur
Foundation’s Research Network on Adolescent Development
and Juvenile Justice. His
research interests include criminological theory, criminal
careers, and quantitative
research methods. He is the author of Key Issues in Criminal

Careers Research (Cam-
bridge University Press, 2007). He is recipient of the American
Society of Criminol-
ogy’s Cavan Award and its E-Mail Mentor of the Year Award,
as well as a Teacher of
the Year Award recipient from the College of Arts and Sciences
at the University of
Florida.
Elizabeth S. Scott is the Harold R. Medina Professor of Law at
Columbia Law
School. Between 1988 and 2006, she was on the faculty of the
University of Virginia
School of Law. A 1977 graduate of the University of Virginia
School of Law, she served
as director of the Forensic Psychiatry Clinic of the University’s
Institute of Law, Psychi-
atry, and Public Policy for several years before joining the
faculty. Her area of scholarly
and teaching interest is juvenile and family law. She has written
extensively on juvenile
delinquency, marriage, divorce, child custody, and adolescent
decision making. Much of
her research is interdisciplinary, applying social science
research, developmental theory,
and behavioral economics to legal policy issues involving
children and families. Scott
has worked in collaboration with social scientists on both
empirical and theoretical
research dealing with family issues and has published in both
law reviews and in inter-
disciplinary journals and edited volumes. The American Law
Institute recently adopted
her proposal offered in a California Law Review article that
child custody be allocated
on the basis of past parental roles (Pluralism, Parental

Preference and Child Custody,
80 CAL. L. REV. 615 (1992). She is a co-author with Ira
Ellman and Paul Kurtz of a
13-OCT-06 12:41
widely used casebook in Family Law. She is also a co-author
(with Walter Wadlington,
Charles Whitebread, and Samuel Davis) of a casebook on
Children in the Legal
System.
Laurence Steinberg is the Distinguished University Professor
and Laura H. Carnell
Professor of Psychology at Temple University and Director of
the John D. and Cathe-
rine T. MacArthur Foundation Research Network on Adolescent
Development and
Juvenile Justice. Dr. Steinberg’s research has focused on a
range of topics in the study
of contemporary adolescence, including parent–adolescent
relationships, adolescent
employment, high-school reform, and juvenile crime and
justice. Dr. Steinberg is a
member of the National Academies’ Board on Children, Youth,
and Families and its
Committee on the Science of Adolescent Health and Behavior,
and he has been a fre-
quent consultant to state and federal agencies and lawmakers on
child labor, secondary
education, and juvenile justice policy.

Police Relations with Black
and White Youths in
Different Urban
Neighborhoods
Rod K. Brunson
Southern Illinois University, Carbondale
Ronald Weitzer
George Washington University, Washington, D.C.
Much of the research on police–citizen relations has focused on
adults, not
youth. Given that adolescents and particularly young males are
more likely
than adults to have involuntary and adversarial contacts with
police officers, it
is especially important to investigate their experiences with and
perceptions of
the police. This article examines the accounts of young Black
and White males
who reside in one of three disadvantaged St. Louis, Missouri,
neighborhoods—
one predominantly Black, one predominantly White, and the
other racially
mixed. In-depth interviews were conducted with the youths, and
the authors’
analysis centers on the ways in which both race and
neighborhood context
influence young males’ orientations toward the police.
Keywords: police–community relations; police misconduct;

disadvantaged
neighborhoods
Research has consistently found that age is a predictor of
citizens’attitudes toward and personal experiences with the
police. Young
people have more frequent contacts with the police than adults
due to their
disproportionate involvement in law breaking and their greater
presence on
the streets (amplifying accessibility to the police) (Leiber,
Nalla, and
Farnworth 1998; Snyder and Sickmund 1996). In encounters
with officers,
youths’ age translates into disempowerment, another factor
shaping inter-
actions with the police. Race makes a difference as well:
Minority youth in
Urban Affairs Review
Volume 44 Number 6
July 2009 858-885
© 2009 The Author(s)
10.1177/1078087408326973
http://uar.sagepub.com
858
Authors’ Note: Please address correspondence to Rod K.
Brunson, Center for the Study of
Crime, Delinquency, and Corrections, Southern Illinois
University, 1000 Faner Drive,
Carbondale, IL 62901; e-mail: [email protected]
at SAINT LEO UNIVERSITY LIBRARY on January 14,

2015uar.sagepub.comDownloaded from
http://uar.sagepub.com/
Brunson, Weitzer / Police Relations in Urban Neighborhoods
859
the United States are more likely than Whites to be viewed with
suspicion
and stopped by the police (Black and Reiss 1970; Hurst, Frank,
and
Browning 2000; Leiber, Nalla, and Farnworth 1998; Piliavin
and Briar
1964) and to report negative personal experiences with officers
(Hagan,
Shedd, and Payne 2005; Taylor et al. 2001). Unfortunately,
most studies
documenting age effects do not include respondents younger
than 18 years
of age, leaving out a critical group.
In addition to demographic characteristics such as race and age,
a grow-
ing body of literature shows that citizens’ relations with the
police are also
ecologically structured. Specifically, neighborhood context
shapes both
police practices and police–citizen relations. A popular
perspective in crimi-
nology, social disorganization theory highlights a set of
neighborhood struc-
tures that increase the likelihood of street crime, and some
scholars have
applied the theory to policing as well (Kane 2002; Reisig and
Parks 2000;

MacDonald et al. 2007; MacDonald and Stokes 2006; Sampson
and Bartusch
1998; Velez 2001). Social disorganization theory holds that
certain neighbor-
hood conditions (poverty, unemployment, single-parent
households, etc.)
weaken social ties between residents and decrease their
willingness to engage
in social control over offenders, hence increasing neighborhood
crime rates.
Neighborhood disadvantage and disorganization also appear to
be associ-
ated with certain patterns in police operations and in residents’
orientations
toward the police. First, police presence in economically
distressed, high-
crime neighborhoods is typically greater than in middle-class
and affluent
areas (Kane 2002; Klinger 1997; Terrill and Reisig 2003),
which increases
the likelihood of recurrent, unwelcome police attention and the
potential for
tenuous police–citizen relations. Second, the amount of street
crime charac-
teristic of these communities offers plenty of opportunities for
police corrup-
tion and other forms of malfeasance. Scholars have found that
aggressive
policing strategies are disproportionately concentrated in
disadvantaged,
crime-ridden communities (Fagan and Davies 2000; Kane 2002;
Reisig and
Parks 2000; Smith 1986; Terrill and Reisig 2003). Living in
such neighbor-
hoods can have adverse outcomes for residents, whom police

tend to distrust
and treat less respectfully than residents of tranquil
neighborhoods (Klinger
1997; Smith 1986; Werthman and Piliavin 1967).
In addition to greater opportunities for abuses of police power
in disadvan-
taged communities, the residents also have little capacity to
hold officers
accountable. The same conditions that increase crime in a
neighborhood may
also foster police misconduct (Kubrin and Weitzer 2003).
Specifically, com-
munity disorganization undermines not only residents’ capacity
to intervene
860 Urban Affairs Review
against crime and disorder, but also increases residents’
powerlessness in the
face of abusive police practices (Kane 2002). Stated differently,
residents of
socially disorganized neighborhoods have low “collective
efficacy” (Sampson
1997), insofar as they lack the mutual trust and collective
willingness to inter-
vene in response to both law breakers in their midst and to
abusive police
officers. Residents of more affluent communities, however, can
draw on orga-

nizational resources and connections to local elites that can be
leveraged in
demands for police accountability (Weitzer 1999, 2000).
If neighborhoods with high levels of socioeconomic
disadvantage and
disorganization are typically areas that register higher rates of
police mis-
conduct toward residents, this raises the question of whether
racial compo-
sition makes a difference. In other words, do socioeconomically
similar but
racially distinct neighborhoods differ in terms of the quality of
police
services or the frequency of police abuses of citizens? This
question has
rarely been examined by scholars.
A few quantitative studies suggest that neighborhood
socioeconomic
status may be more important than racial composition (Reisig
and Parks
2000; Sampson and Bartusch 1998; Velez 2001). However,
these studies are
too few to draw definitive conclusions, and more important, in
most
American cities, the number of Whites living in highly
disadvantaged neigh-
borhoods may be insufficient to sustain multilevel models
comparing Whites
with disadvantaged Black neighborhoods and individuals
(MacDonald et al.
2007). Moreover, by aggregating the data and identifying
predictors across
many neighborhoods (of the same general type), these large-
scale studies

mask potentially important neighborhood-specific patterns in
police–com-
munity relations. This points to the need for contextualized,
qualitative
research comparing disadvantaged communities of different
racial composi-
tions. The closest any studies have come to comparing these
particular con-
texts are Jacob’s (1971) research on one working-class White
and one poor
Black neighborhood in Milwaukee (this was a limited
quantitative analysis)
and Foster’s (1989) comparison of one White working-class and
one disad-
vantaged Black neighborhood in London (examining police
behavior, not
residents’ attitudes). In neither investigation were the study
sites well-
matched disadvantaged neighborhoods, perhaps because of the
lack of a
comparable disadvantaged White neighborhood in either city.
To further
explore the ecology of race, the present study examines the
perceptions and
experiences of young residents of three similarly disadvantaged
communi-
ties that differ in racial composition—White, Black, and mixed.
861

Study Setting and Methodology
Most of the literature is quantitative, and it is important to
complement
these studies with qualitative research to document complex and
nuanced
citizen understandings of police practices. Only a handful of
qualitative
studies of either adults or youth exist (Brunson 2007; Brunson
and Miller
2006a, 2006b; Carr, Napolitano, and Keating 2007; Ellison
2001; Sharp
and Atherton 2007; Weitzer 1999, 2000). The present article
contributes to
this body of literature.
Data for this article are drawn from in-depth interviews with
male adoles-
cents living in three highly disadvantaged neighborhoods in St.
Louis,
Missouri: one Black (Barksdale), one White (Mayfield), and one
racially
mixed (Hazelcrest, 40% Black, 50% White).1 Table 1 provides
census data on
respondents’ neighborhoods and for St. Louis. The three
neighborhoods are
almost perfectly matched on socioeconomic indicators: poverty
rate (26%),
median household income ($23,000-$25,000), and
unemployment rate (12%-
15%). (The racially mixed community was selected from a set
of census
tracts comprised of nearly equal numbers of Black and White
residents.) In
other words, the study sites allowed us to hold neighborhood

socioeconomic
status constant while varying on racial composition—a design
advantage not
present in most other qualitative studies of police–community
relations (e.g.,
Carr, Napolitano, and Keating 2007; Sharp and Atherton 2007).
The three neighborhoods also fit the criminogenic profile
identified
by social disorganization theory. All three neighborhoods
register higher
rates than the city average for at least some index crimes (see
Table 2).
Street crime and disorder were also described in our youths’
accounts
Table 1
Neighborhood Racial Composition and Socioeconomic Profile
Median
Household % Families in % Female-headed
% White % Black Income Poverty Householda % Unemployed
Mayfield 85.6 7.4 $22,861 26.3 37.8 11.8
Barksdale 1.5 97.2 $24,099 26.5 44.7 15.2
Hazelcrest 50.2 39.8 $24,933 26.1 55.4 11.8
St. Louis 43.9 51.2 $27,156 20.8 47.5 11.3
Source: U.S. Bureau of Census (2000).
a. Single-parent household with children under 18 years of age.

and in our interviewers’ descriptions of neighborhood
conditions. In
addition to serious physical decay and social disorder in
Hazelcrest,
respondents told us that robberies, assaults, shootings, and car
thefts
were not uncommon. Similar problems afflicted Barksdale, but
it stands
out in its high homicide rate. Mayfield was noted for certain
kinds of
street crime (vandalism, drug use) and disorder manifested in
unsuper-
vised youth, stray animals, abandoned vehicles, debris, vacant
lots, and
derelict buildings.
Although the article is focused on citizens, some aspects of
policing in
St. Louis can help contextualize our study. Community policing
is fairly
limited in the city: Monthly neighborhood meetings are held,
and a public
affairs officer is assigned to each of the nine police districts.
Traditional
policing is the norm, which means that police work largely
consists of
responding to calls and proactive patrolling rather than
developing
police–community partnerships to prevent crime and deal with
disorder. In
our three study sites, in particular, residents reported that
officers were

mainly involved in patrolling and surveillance and engaged in
frequent
stops and searches of pedestrians and motorists. Finally, there
appears to be
some racial disparity in police stops of citizens in St. Louis.
According to
police department data, African-American motorists in the city
are some-
what more likely to be stopped by police than White motorists
(Missouri
Attorney General 2005-2007). It should be noted, however, that
most of our
Table 2
Neighborhood and City Crime, 2005-2006
Aggravated Auto
Homicide Rape Robbery Assault Burglary Larceny Theft
Number
Mayfield 2 8 34 82 138 738 145
Barksdale 18 10 104 325 405 522 302
Hazelcrest 6 29 378 607 881 1390 842
Ratea
Mayfield .66 2.67 11.35 27.37 46.07 246.41 48.41
Barksdale 3.43 1.90 19.82 61.96 77.21 99.52 57.57
Hazelcrest .34 1.68 21.94 35.24 51.15 80.71 48.89
St. Louis City
Number 260 613 6,112 9,941 15,723 46,409 16,791
Ratea .74 1.76 17.56 28.55 45.16 133.30 48.23
Source: St. Louis Metropolitan Police Department, 2005 and
2006.

a. Mean rate per 1,000 population in 2000.
study participants who were stopped were too young to drive at
the time
and were instead stopped as pedestrians.
The current study is based on information obtained from 45
male ado-
lescents living in the three study sites. The interviews were
conducted
between fall 2005 and spring 2006.2 Respondents range in age
from 13 to
19, with a mean age of 16. Participation in the study was
voluntary. The
young men were paid $25 and promised confidentiality.
Three community organizations working with adolescents at
neighbor-
hood recreation centers helped to recruit study participants—a
strategy
used in some other studies (Carr, Napolitano, and Keating 2007;
Sharp and
Atherton 2007). Our data collection strategy was designed to
enlist a wide
range of youth. One of the centers was run by the St. Louis
Public School
District and provided after-hours GED classes to school
dropouts. Another
center was jointly operated by a grassroots community
organization and the

public schools. This site hosted an “open gym,” but the
sponsors’ ultimate
goal was to deliver social services to at-risk neighborhood
youth. The
remaining center was funded and operated by the city’s Parks
and Recreation
Department.
Sampling was purposive: Counselors were asked to identify and
approach young males for participation in the study, persons
who were
known to live in the neighborhood. The goal was to interview
young Black
and White males who live in comparable disadvantaged
neighborhoods and
are at risk of or involved in delinquent activities, as these
youths would
likely have more contact with police. In other words, sampling
was
designed to include youths who may have had personal
experiences with
the police, but we did not specifically select youths known to
have had bad
experiences or who had previously expressed animosity toward
local law
enforcement. As it turned out, about half (46%) of our sample
reported that
they had been arrested by the police at some time, and 22% had
been
arrested within the previous six months.
Young males are the focus here because research has identified
them as
the group for whom involuntary police contacts are most
frequent in the
United States (Hurst, Frank, and Browning 2000). Few studies

have offered
an in-depth examination of the nature of young men’s
experiences with the
police and their perceptions of these interactions. Our study
allows for a
detailed examination of these issues with both Black and White
youth. The
interviews explored not only youths’ attitudes toward the police
but also
their personal experiences and observations of officers,
including detailed
accounts of their encounters. Reliability was enhanced by cross-
checking
and probing the youths’ responses to the interview questions.
Interviews
863
lasted approximately one hour, and all except one were
conducted in
private offices at each location.3 The tapes from those
interviews were
transcribed and analyzed by the authors.
The data are limited to respondents’ self-reports—their attitudes
toward
the police and their accounts of encounters with and
observations of
officers. In the analysis, we selected statements that illustrated

themes con-
sistently found throughout the data. The quotes used were not
atypical, with
the exception of a few issues that we indicate only a few
respondents
mentioned. We do not assume that young men’s version of
events is accu-
rate or that in all cases they have provided full disclosure. In
fact, citizens
may misconstrue police actions and intentions. What is most
important for
the current study, however, is specifically how respondents
characterize
their experiences with and perceptions of the police.
Nonetheless, we
attempted to strengthen the validity and reliability of the data
by asking
study participants about their views several times during the
interview, by
inquiring about their personal encounters and direct
observations of police
officers, and by requesting detailed narratives of incidents. The
data analy-
sis was conducted with great care to make certain that the
themes we dis-
covered accurately represented young men’s descriptions. This
was
accomplished using grounded theory methods, by which
recurrent topics
were identified along with less common but important issues
(Strauss
1987). Each author independently coded the data and
subsequently catego-
rized it into themes and subthemes. In the presentation of study
findings,
we consider young men’s racial backgrounds and draw

comparisons based
on where they live. Any conclusions from these comparisons,
however,
must be regarded cautiously, given the modest numbers of
respondents
across the three study sites. Nonetheless, future research
comparing Black,
White, and mixed neighborhoods can be used to corroborate our
results.
Results
The data point to both racial and neighborhood differences on
the expe-
riential questions. In general, (1) White youth had a less
troubled relation-
ship with and more positive views of the police than Black
youth, and (2)
police treatment of residents appeared to be less problematic in
the White
neighborhood (Mayfield) and more problematic in the Black
neighborhood
(Barksdale), with the mixed neighborhood (Hazelcrest) falling
in between.
Black youth in Barksdale were much more likely than Whites in
Mayfield
and Hazelcrest to report personal and vicarious experiences
with police

abuse. In addition, within Hazelcrest, White residents fared
better than
Blacks, and Blacks in Barksdale were more likely than Blacks
in Hazelcrest
to report personal negative experiences with officers, while
Black youth in
Hazelcrest were more likely to know someone who had been
mistreated by
police. The vast majority of Barksdale and Hazelcrest
respondents reported
that policing in their neighborhoods primarily consisted of
pedestrian and
vehicle stops by patrol officers and specialized units. While a
handful of our
youths thought that proactive police actions were intended to
address par-
ticular neighborhood problems, respondents also considered this
style of
policing as overly aggressive and confrontational. The
following discussion
examines youths’ accounts of their own experiences as well as
the experi-
ences of others that were either directly observed by or
communicated to
them.
Unwarranted Stops
Unwarranted stops of citizens by officers are those where the
officer
lacks any indication of illegal conduct or those based entirely
on a vague
intuition that something may be awry. There is no way of
knowing what
proportion of stops is proper and what proportion unlawful or

otherwise
unwarranted. We do know, however, that African-American
motorists are
somewhat more likely to be stopped by police than White
motorists in St.
Louis and much more likely to be searched following a stop
than White
motorists (Missouri Attorney General 2005-2007). Research
indicates that
citizens hold varying definitions of whether a stop is justified,
both in the
abstract and regarding their own experiences (Weitzer 1999).
What matters
for the present analysis is whether citizens believe they have
been stopped
without due cause.
Several of our respondents reported being routinely stopped in
situations
in which they believed there was no basis for suspicion. Jamal
recounted
such a stop: “The [police] stopped me and they ran my name
and said I
needed an [identification card] ’cause I wasn’t in the system.
[The officer]
was like, ‘I’m not arresting you, [but] can I put you in
handcuffs and run
your name?’” Jamal’s account illustrates what many young
males consider
arbitrary police decisions to stop and question them. And even
though he
realized that he was not under arrest, Jamal said that he took
particular
exception to being placed in handcuffs “like a common
criminal” while the
officer processed his information. Similarly, David reported,

“Me and my
friends was walkin’ and I guess [the police] thought we was
hangin’ on the
corner. [The police] rode up and pulled us over. First thing they
said was,
865
‘Get on the hood [of the patrol car].’ . . . They told us to spread
our arms
and legs and then searched us.” Police were depicted as
overreaching:
Officers “lock you up for anything” (Darius) and “harass us
constantly” for
“no reason” (Derek). Will described one incident:
A friend of mine and me were in the community, we were
outside and it was
a late night and I guess the officer that approached took us as
gang bangers
or whatever. He asked us what we were doing and we looked at
each other
and we said, “Nothing,” and he [said] that we looked real
suspicious. . . . He
used the excuse that we had drugs in our mouth[s] and told us to
take what-
ever we had in our mouths out. We had grills [decorative dental
molds] in our
mouth[s] and he made us take them out, we showed them to him

in our
hand[s] and [the officer] smacked ’em out and when they [hit]
the ground, he
stomped on them and laughed. But he was showing us that he
had more
power, authority over us at the time, so there was nothing we
could do or say.
Respondents were particularly annoyed at unwelcome police
attention
when engaged in law-abiding behavior. Todd and his friends
were detained
by officers as they walked home from school: “The police got
out of the car
and were like, ‘What ya’ll doing?’ I said, ‘We’re coming home
from
school.’ [The officer] was like, ‘What’s in the book bags?’ He
came over
and started checking but couldn’t find nothing but books.” And
Martez and
his friends were subjected to a series of physically intrusive
searches:
We was playin’ basketball and [my friend] put a wristband in
his gym bag.
. . . The police thought it was some crack so they stopped him
and was harass-
ing him, like, “Where its at?” He was like, “I ain’t got nothin’.”
After they
checked him, they checked all of us. Only thing they found was
wristbands,
white wristbands. . . . [The police officers] took all six of us in
[to the station]
and was checkin’ our mouth[s] and [other body parts] . . . to see
if we have
drugs and they found out [that] we didn’t.

The majority of Black youths said that they routinely attracted
police atten-
tion regardless of whether they were involved in criminal or
suspicious
activities.
White youths reported fewer experiences, overall, with police
stops. In
Mayfield, White youths reported few contacts of any kind with
police offi-
cers. But White youths’ risk of being stopped was heightened in
three spe-
cific situations: (1) while in the company of young Black males,
(2) when
in racially mixed or majority-Black neighborhoods, or (3) while
dressed in
hip-hop apparel. It has been argued that Whites enjoy a “racial
halo effect”
that reduces the chances of being viewed with suspicion by
police officers
(Weitzer 1999), but this halo appears to dim in these three
situations of guilt
by association.
Such incidents were reported by several White youths. For
example, Ed

explained, “If White [police officers] are pulling over Whites
they are prob-
ably nicer than if they are pulling over Blacks, but if they are
pulling over
Whites and Blacks together . . . a White cop is going to treat a
White kid
with Black kids just as bad or worse than he would treat Black
kids.” Ed
had been stopped while driving through a Black neighborhood
with Black
friends: “I probably stood out because I was a White driver and
had two
Black guys and a Black girl in the car with me.” And Toby
explained how he
and his friend were treated when the police encountered them in
a majority-
Black neighborhood:
[We] was on a corner during school hours and a cop talked to us
about what
we were doing, and then took us back to school. We got in
trouble for it at
school, it sucked. . . . The cop that stopped us was being a dick
at first.
He kept asking us if we were going on a booty call together.
You know, like
we were gay. Then kept making jokes about booty calls and then
ask[ed] if
we left school because of the “brothers.” Then he asked if we
were scared of
the “brothers” and if that is why we left school or if the
“brothers” booty
call[ed] us and that is why we left. The cop finally quit giving
us shit, took
us back to school, and we got three days of in-school
suspension.

While Toby and his friend were offended by the officer’s
comments, they
decided against reporting him because they “kinda got off
easy.”
Toby’s account is consistent with those provided by other White
youths
in Hazelcrest and Mayfield. Several Whites reported that when
officers
found them in majority-Black neighborhoods, they initially
expressed con-
cern for their safety. That is, police wanted to make sure that
White kids
understood the potential danger of such settings. However,
White youths
who frequented certain Black neighborhoods were more likely
to be viewed
suspiciously by officers. For instance, Kyle observed,
[The police] asked me what I was doing in a Black
neighborhood, ’cause I’m
a White boy. They said, “You ain’t sellin’ drugs are you?” and
he tried to
plant drugs on me. I didn’t have weed on me ’cause I never sold
it or smoked
it but the [officer] put it in my pocket, put his hand in my
pocket and pulled
it out. I felt his thumb was folded under his hand, there’s a
lump. And then
when he touched me I felt a bag under his thumb, he pulled it
out and then
said, “Aha, what’s this?”
867

In addition, police officers appeared to be bothered by White
youths’ fond-
ness for urban clothing and argot. As James observed, “It don’t
depend on
what race you are, it depends on what you wear. . . . [If] a
White guy is wear-
ing hoodies, brand new Jordans, a tank top, [the police] gonna
harass him too.”
Nate agreed that “it’s the way we dress and talk. [Police] pretty
much stereo-
type people. . . . They think if kids do saggin’ pants and grills,
gold [teeth] in
they mouth, [that] we punks or we ain’t no good.” Kyle, who
had grown up in
a Black neighborhood, said that police think “we look thuggish,
so they treat
us like thugs. . . . But if you grew up in a perfect neighborhood,
the [police]
treat you like you’re a human being.” And finally, Maurice
noted,
[The police] assume you run the streets, steal cars, or smoke
weed because
you dress a certain way, like baggy pants or a long T-shirt and
Nike brand
shoes. They consider you as a gang member just because of
what you were
wearing or how you talk.

This is a perfect illustration of Skolnick’s (1966, 45-48)
“symbolic
assailant”—an individual whose mere attire, demeanor, or
language is con-
strued by police as a cue that the person is a potential threat or
involved in
illegal activity.
The White respondents did not necessarily claim that police
treated them
the same as their Black associates, who remained officers’
primary targets,
but they made it clear that they were treated differently in these
situations
because of guilt by association with Black kids, Black
neighborhoods, or
Black youth culture. Such experiences occurred when officers
encountered
White youth in Black or mixed neighborhoods. In Mayfield,
White youth
reported having little contact with officers, primarily observing
them on
patrol. Limited contact meant fewer occasions for unpleasant
encounters,
with the result that Mayfield youth were more likely than other
youth to
believe that most police stops were justified. In fact, some
respondents
expressed a desire for a greater police presence in Mayfield and
for more
face-to-face contact with officers—also a finding in a
Philadelphia study of
youth (Carr, Napolitano, and Keating 2007). One youth said
officers should
“ask residents how they are doing,” and another said, “Cops
should get to

server05productnCCPP5-4CPP403.txt unknown Seq 1 13-OCT.docx

server05productnCCPP5-4CPP403.txt unknown Seq 1 13-OCT.docx

Recommended

Recommended

More Related Content

Similar to server05productnCCPP5-4CPP403.txt unknown Seq 1 13-OCT.docx

Similar to server05productnCCPP5-4CPP403.txt unknown Seq 1 13-OCT.docx (20)

More from lesleyryder69361

More from lesleyryder69361 (20)

Recently uploaded

Recently uploaded (20)

server05productnCCPP5-4CPP403.txt unknown Seq 1 13-OCT.docx