Racial Threat and Punitive
School Discipline
Kelly Welch, Villanova University
Allison Ann Payne, Villanova University
Tests of the racial threat hypothesis, linking the racial composition of place to various measures of social
control, find that where there are greater percentages of blacks, more punitive criminal justice policies are implemented. Just as the criminal justice system continues to get tougher on crime despite stagnant crime rates, it is also
clear that schools are becoming harsher toward student misbehavior and delinquency despite decreases in these
school-based occurrences. However, only a very limited number of studies have been able to partially explain this
trend of intensifying social control in schools. Using a national sample of 294 public schools, the present study
is the first to test the racial threat hypothesis within schools to determine if the racial composition of students
predicts greater use of punitive controls, regardless of levels of misbehavior and delinquency. Results of multivariate analyses support the racial threat perspective, finding that schools with a larger percentage of black students
are not only more likely to use punitive disciplinary responses, but also more likely to use extremely punitive
discipline and to implement zero tolerance policies. They also use fewer mild disciplinary practices and restitutive techniques. In addition, racial threat is more influential when school delinquency and disorder are lower.
Keywords: racial threat, school discipline, punitiveness, racial disparity, student composition.
In Discipline and Punish (1977), Michel Foucault established that the school’s role as an
institution of social control has existed since public schooling began. In recent decades, however, there has been a notable intensification of formal controls in schools (Beger 2002; Skiba
2000). The discipline and punishment of students has become particularly harsh—a trend
that mirrors the mounting punitiveness in the criminal justice system. And, as with falling
crime rates, rates of student delinquency, student drug use, violent school victimization, and
school-related deaths have actually declined since the early 1990s, despite perceived increases
(Beger 2002; Brooks, Schiraldi, and Ziedenberg 1999; Devoe et al. 2005; Dinkes, Cataldi, and
Lin-Kelly 2007).
While there is a range of possible responses to student misbehavior used by teachers
and school administrators, some increasingly restrictive school practices are being utilized,
including more teacher referrals to the principal, more exclusion from class time, and more
detentions, suspensions, and expulsions (Gottfredson and Gottfredson 2001).1 Many of the
1. References to misbehavior include any student violation of school rules (such as talking back to teachers or not
submitting homework), which may or may not also include behaviors that are against the law, such as destruction of
school property or drug use. Law violation, in addition to being categorized as misbehavior, is also considered delinquent.
In the past, schools have dealt with the majority of all in-school misbehaviors internally. But, recently, schools have
responded to certain misbehaviors, like fighting or graffiti, by calling police (Hirschfield 2008).
An earlier version of this article was presented at the 2008 Annual Meeting of the American Society of Criminology, St. Louis, Missouri. The authors would like to extend a special thanks to Denise and Gary Gottfredson for kindly
sharing their National Study of Delinquency Prevention in Schools data for use in this study. They would also like to
express gratitude to the editor and five anonymous reviewers for their very thoughtful comments on this research; their
recommendations and generous assistance have greatly improved this article. Direct correspondence to: Kelly Welch,
800 Lancaster Avenue, Villanova, PA 19085. E-mail: kelly.welch@villanova.edu.
Social Problems, Vol. 57, Issue 1, pp. 25–48, ISSN 0037-7791, electronic ISSN 1533-8533. © 2010 by Society for the Study of
Social Problems, Inc. All rights reserved. Please direct all requests for permission to photocopy or reproduce article content
through the University of California Press’s Rights and Permissions website at www.ucpressjournals.com/reprintinfo/asp.
DOI: 10.1525/sp.2010.57.1.25.
SP5701_03.indd 25
12/21/09 11:33:43 AM
26
Welch/Payne
more punitive practices, like suspension and expulsion, are the direct result of the implementation of zero tolerance policies, which generally refer to federally mandated suspensions for
bringing guns to school, but can also include mandatory consequences for a variety of other
misbehaviors identified by individual institutions (Skiba 2000; Skiba and Peterson 1999).2
Regardless of this ever-escalating severity, school boards, school administrators, and parents
have continued to call for even harsher student treatment (Robbins 2008). It is not clear, however, what is driving this pattern of expanding school punitiveness.
Intensification of School Punitiveness
It is widely acknowledged that American schools—urban public schools in particular—
are defining and managing student discipline with an increasingly punitive approach. Specific
examples of restrictive school policies include the imposition of dress codes or uniforms to
promote conformity (Gottfredson and Gottfredson 2001; Watts and Erevelles 2004), as well
as the use of student ID badges to facilitate immediate identification of troublemakers and
deter would-be rule breakers (Beger 2002; Brooks et al. 1999). Recent statistics indicate that
over 80 percent of schools, and nearly all urban schools, now use some type of security and
surveillance program (Dinkes, Kemp, and Baum 2009). Aspects of these programs include the
use of metal detectors, locked or monitored doors and gates, adult supervision of hallways,
uniformed security guards or uniformed and armed security resource officers (SROs), security
cameras, locker searches, clear school bag requirements, and drug-sniffing dogs (Beger 2002;
Devoe et al. 2005; Giroux 2003; Hirschfield 2008; Watts and Erevelles 2004).
Research suggests that these trends in school punitiveness are not inconsequential. For
students perceived by teachers and administrators to be in the “school to prison pipeline” or
on the “jailhouse track” (Advancement Project 2005), more punitive discipline is sanctioned,
even when the particular violations are not illegal (Hirschfield 2008). Schools are more frequently responding to defiance of behavioral codes and broken laws with in-school suspensions, out-of-school suspensions, and expulsions (Brooks et al. 1999; Fantz 2008; Giroux 2003;
Kupchik and Monahan 2006). Moreover, legal reforms mandate that student involvement in
certain acts that are against the law, such as drug and alcohol use or weapon possession, more
frequently result in referrals to law enforcement rather than being addressed internally, even
when they occur on school property (Hirschfield 2008). These types of responses by teachers,
administrators, and lawmakers further enhance the likelihood that notions of future criminal
justice involvement will be realized. Considering the recent trend toward more restrictive
school practices and punitive student discipline, it is not surprising that the students subjected
to the most intense forms of social control share many of the same characteristics with those
who are most likely to be accused and convicted of crime—they tend to be poor, male, and
members of a racial or ethnic minority (Noguera 2003a; Singer 1996).
As Paul Hirschfield (2008) has argued, there are four general explanations for trends in
school punitiveness, each of which he considers to be inadequate in important ways highlighted here. One explanation is that this approach to discipline is a direct consequence of
popular anxiety about high profile campus shootings (Beger 2002; Watts and Erevelles 2004)
and other notable instances of school violence (DeMitchell and Cobb 2003). However, even
before these events, the public had been concerned about several well-publicized instances
of school violence that did not result in more restrictive policies. Further, this explanation
does not account for the continued expansion of punitiveness toward students despite recent
2. Zero tolerance can be traced to the federal Gun-Free School Act of 1994, which sought to restrict weapons in
schools, but expanded to include nonweapon related and nonviolent behaviors. While zero tolerance has been widely
criticized for unintended consequences, original supporters advocated greater fairness toward students, more consistent
discipline, and violence prevention (Gorman and Pauken 2003; Henault 2001).
SP5701_03.indd 26
12/21/09 11:33:43 AM
Racial Threat and Punitive School Discipline 27
decreases in school violence. This suggests that popular anxiety cannot totally account for
disciplinary trends, many of which did not even address violence.
Increased school accountability for the academic performance of students is another
explanation that has been offered; instead of requiring better teaching and learning, many
schools may have increased their suspensions and expulsions to facilitate the removal of those
students responsible for failing test scores (Bryk et al. 1998; Fuentes 2003). However, by transferring authority away from school officials via the implementation of zero tolerance punishments, underperforming students would not necessarily be targeted for exclusion.
Some argue that concerns about possible litigation for using nonmandatory exclusionary school punishments have led to the involvement of third parties to handle more severe
instances of misbehavior (Arum 2003; Toby 1998). However, Hirschfield (2008) argues this
does not adequately explain why harsher practices are used in urban schools, when affluent students in suburban schools presumably have more access to legal remedies (Arum
2003).
The final explanation pertains to broad support for “getting tough” in a variety of American institutions, including schools that developed a crime-control approach to education. This
“governing through crime” orientation (Simon 2006), however, does not address the changing nature of discipline that is not a result of federal or state initiatives because much discipline
continues to originate with individual teachers and principals, suggesting that school punitiveness is not merely a byproduct of government mandates (Gottfredson et al. 2000).3
Review of Research on School and Student Discipline
A limited number of studies have explored the effects of specific school- and student-level
characteristics on the use of more severe student discipline. Assessments of individual variables
associated with discipline are few, most likely due to the restricted availability of data from the
U.S. Department of Education. Overall, there are certain conclusions about punishment that
may be made from school- and student-level studies. One factor presumed to be closely associated with school punitiveness and disciplinary practice is the level of school crime and disorder (Skiba et al. 2002). Surprisingly, researchers generally conclude that changes in school
punitiveness are not driven by changes in student misconduct and delinquency (Hirschfield
2008; Kupchik and Monahan 2006; Skiba and Peterson 1999; Wu et al. 1982). It is also possible that the implementation of zero tolerance policies is resulting in more punishment of
certain students, even though studies show that there is actually a considerable amount of
individual-level discretion employed in determining which student behaviors are addressed
by them (Ferguson 2000; Vavrus and Cole 2002). For example, one qualitative study found
that mandatory sanctions are often preceded by a series of nonviolent events that culminate
in a singling out by a teacher for a single disruptive behavior (Vavrus and Cole 2002). This
finding indicates that harsher student treatment is not merely a reflection of more violations
punishable by mandated suspensions, but also importantly involves the discretion of teachers
and administrators in identifying and acknowledging those behaviors.
Aside from actual amounts of student delinquency and school disorder, it has been suggested that some measures of crime salience (such as prior victimization and fear, perceived
risk and safety, and concern about crime or delinquency) may predict harsh social control in
schools (Hilarski 2004; Skiba et al. 2002). While there have been no multivariate tests predicting a relationship between crime salience and punitive outcomes for individual students, the
studies examining these crime salience constructs find that where there is greater teacher and
administrative fear of violent victimization, there tends to be more preventive policies used by
3. It is possible that the “governing through crime” approach to education may have expanded indirectly to school
administrators and teachers who adopted this perspective in relating to students (Simon 2006).
SP5701_03.indd 27
12/21/09 11:33:43 AM
28
Welch/Payne
schools (Hilarski 2004). Thus, it seems that the salience of crime and delinquency may also be
manifested in school punishment.
Studies show that certain student socio-demographic qualities are related to harsh school
discipline. Student economic disadvantage, generally operationalized as whether a student
receives free or reduced-price lunches, is one of the most consistent predictors of harsh school
discipline throughout the last several decades (Brantlinger 1991; McCarthy and Hoge 1987;
Nichols 2004; Skiba et al. 2002). As expected, statistics indicate that poorer students are, in
fact, more likely to be targeted by harsh school practices (Brantlinger 1991; Singer 1996; Skiba
et al. 2002; Skiba, Peterson, and Williams 1997; Wu et al. 1982), while wealthier students
more often receive mild or moderate consequences (Skiba et al. 2002). Similarly, students
whose fathers do not have full-time jobs have been more likely to be suspended (Wu et al.
1982). Student gender is another factor identified as consequential for disciplinary responses:
Male students receive harsher punishments (Artiles 2003; Gregory 1995; Noguera 1995a;
Skiba et al. 2002).
Additionally, various school characteristics have been found to affect the social control
of students. School urbanicity is one feature associated with increased student punishment,
which some have speculated may, in part, be due to student race or socioeconomic status and
levels or types of crime both inside and outside of urban schools (DeVoe et al. 2005; Massachusetts Advocacy Center 1986; Skiba 2000; Wu et al. 1982). Research also shows that
discipline policies are more likely to effectively improve student behavior and accountability
in schools with strong principal leadership (Di Lullo 2004; Lasley and Wayson 1982); good
leadership includes support of teachers, consistent supervision and use of feedback, high visibility and presence, and effective planning and problem-solving (Gottfredson et al. 2000).
Conversely, poor principal leadership is associated with greater use of punitive discipline (Wu
et al. 1982). Discipline training is also related to the quality of a school’s disciplinary response,
such that more training is associated with more use of discipline (Gottfredson et al. 2000; Wu
et al. 1982).
Student Race and Punitive Student Discipline
Racial status is another variable consistently related to student punitiveness, with minority students receiving harsher treatment more often than white students. It is now wellknown that black students are more likely to be subjected to intense school control than white
students (Ferguson 2000; Gottfredson and Gottfredson 2001; Nichols 2004; Welsh 2000), and
are generally given punishments that are more frequent and more punitive for less serious offenses (Brown and Beckett 2006; Nichols 2004; Noguera 2003b; Skiba 2000, 2001; Wacquant
2001). This disparity can be seen with suspensions (APA Zero Tolerance Task Force 2006;
Brooks et al. 1999; Gregory and Weinstein 2008; Vavrus and Cole 2002), expulsions (Gregory
and Weinstein 2008; Meier, Stewart and England 1989; Skiba and Peterson 1999), and even
corporal punishment (Gregory 1995; Shaw and Braden 1990). In addition, studies show that
black students are less likely to receive mild disciplinary alternatives (Skiba and Peterson
2000; Skiba et al. 2002). Minority students are also more likely to be subjected to restrictive
prevention tactics, such as metal detector screening, and mandatory discipline, like zero tolerance punishments (Beger 2002; DeVoe et al. 2005; Noguera 2003b; Skiba 2000; Vavrus and
Cole 2002). Compounding the racial imbalance of discipline, disparities in student treatment
actually increase as punishments become more punitive (Skiba et al. 2002).
Several compelling individual-level explanations for racially disparate discipline have
been advanced, including disproportionate involvement in delinquency (Noguera 2003a;
Skiba et al. 2002; Watts and Erevelles 2004), more involvement in zero tolerance violations
(Ferguson 2000; Vavrus and Cole 2002), perceptions of misbehavior resulting from racial
stereotypes or bias (Nichols 2004; Skiba and Peterson 1999), and the “adultification” of black
kids by school personnel (Ferguson 2000). One of the most frequently cited explanations
SP5701_03.indd 28
12/21/09 11:33:43 AM
Racial Threat and Punitive School Discipline 29
for findings about the effects of student race on discipline is that it is actually socioeconomic
status that influences school punitiveness, and that any associations with race are partially,
if not completely, spurious (Brantlinger 1991; Nichols 2004; Skiba and Peterson 1999; Watts
and Erevelles 2004). However, results from multivariate tests of these explanations show that
they are either insufficient or inaccurate; even while controlling for effects of misbehavior, attitudes, academic performance, parental attention, school organization, and economic disadvantage and poverty, black students receive harsher school punishments (Gregory and Weinstein 2008; McCarthy and Hoge 1987; Skiba and Peterson 1999; Skiba et al. 2002; Vavrus and
Cole 2002).
Racial Threat
While the findings about the factors associated with increased formal control of students
in schools are important, there may be other ways to account for greater school punitiveness. No research to date has assessed the possibility that macro-level racial threat is behind
the punitive trend in school discipline. Racial threat is different from individual student race
that has predicted school punitiveness in prior research. Rooted in the conflict perspective,
the racial threat hypothesis suggests that as the proportion of blacks increases in relation to
whites, intensified measures of control will proliferate in response to the perceived growing
threat derived from closer proximity to minorities. Hubert Blalock (1967) argued that, at the
macro-social level, the presumed threat to the white majority by blacks was both economic
and political, a concept he termed power threat. Research testing this perspective corroborates
that perceived racial competition, a byproduct of a greater number of blacks and the subsequent concerns about limited financial resources (Taylor 1998) and political capital (Jacobs
and Carmichael 2002), leads to the imposition of punitive social controls in order to maintain
dominance of the majority.
Studies of this effect eventually came to include the presumption of black crime as one potential source of threat, an influence identified as social threat (Liska 1992) and, more recently
due to its substantially crime-specific nature, racial threat (Baumer, Messner, and Rosenfeld
2003; Crawford, Chiricos, and Kleck 1998).4 Racial composition of place—the most common
proxy for racial threat—is related to rates of arrest (Mosher 2001), resources and size of both
law enforcement (Chamlin 1989) and corrections (Jacobs and Helms 1999), rates of incarceration (Jacobs and Kleban 2003), and executions (Tolnay, Beck, and Massey 1992). The effect
of racial composition on social control also appears to be mediated by micro-level perceptions
of racial threat, such that widespread associations made between blacks and dangerous predatory criminality are consequential for public support for harsh criminal policies (Chiricos,
Welch, and Gertz 2004) and crime reduction expenditures (Barkan and Cohn 2005).
Contextual Effects of Racial Threat
Blalock (1967) suggested a possibility of conditional effects when noting that “different kinds of persons will not be similarly motivated by the minority percentage variable”
(p. 31). Some research shows that the effects of racial threat are not equally consequential
in all circumstances. Regionally, the percentage of the black population has been associated
with harshness only in racially diverse cities with low segregation (Stolzenberg, D’Alessio,
and Eitle 2004) and in places where concentrated disadvantage is greater (Bontrager, Bales,
and Chiricos 2005), and perceived racial threat has been associated with harshness only in
4. The close association made by many Americans between blacks and crime has been extensively documented
and described (see Welch 2007 for a review). It is the strength and pervasiveness of this link that has led researchers to
conclude that “racial threat” (as distinct from power threat and social threat) is largely crime related.
SP5701_03.indd 29
12/21/09 11:33:43 AM
30
Welch/Payne
the non-South (Chiricos et al. 2004) and Northeast (King 2007). Effects of racial composition
also manifest differently over time (Behrens, Uggen, and Manza 2003). Moreover, effects of
racial threat may depend on the type of crime control assessed, such as the implementation of
habitual offender statutes (Crow and Johnson 2008) and sentencing adjudication (Bontrager
et al. 2005) only for drug offenses.
Some of the results of contextual analyses may be surprising, because racial threat appears to be more influential in situations where one might expect less punitive measures to be
implemented (Stolzenberg et al. 2004). However, it makes sense that greater effects of racial
threat are likely when the effects of other influences on social control are less dominant. For
example, Ted Chiricos and colleagues (2004) found that perceived racial threat increases punitiveness only when crime salience and racial prejudice are lower. Despite some inconsistent
findings, most likely resulting from certain methodological limitations (Parker, Stults, and
Rice 2005), there is relatively strong support for the racial threat perspective in both contextual and noncontextual analyses. Yet, no research has explored its potential implications for
punitive school punishments.
The Present Study
It is plausible that crime-related racial threat is contributing to school punitiveness much
in the same way that it has been found to increase the severity of numerous criminal policies
and garner the public’s endorsement of them. Loïc Wacquant (2001) suggests that it is important to examine the reason that punitive discipline appears to be more intense in schools
that are more heavily populated by disadvantaged black students. Because of the potential
negative consequences of harsh school policies and the dramatic racial disparities in school
punishments, an exploration of this possibility seems particularly critical.
While individual student race has been positively associated with school punishment, it
has never been assessed at the aggregate level for punitiveness across schools. The underlying
threat of black criminality perceived at the local level by policy makers, school district board
members, school administrators, and teachers may lead to the enactment, implementation,
and enforcement of harsher disciplinary policies and practices in schools with greater numbers of black students because of their increased proportions, as would be predicted by the
racial threat hypothesis. In order to test for racial threat within the context of schools, this
study examines the following hypotheses pertaining to punitive disciplinary responses: (1)
Schools with a higher proportion of black students are more likely to use punitive controls,
such as detention and suspension; (2) Schools with a higher proportion of black students are
more likely to implement zero tolerance policies; and (3) Schools with a higher proportion of
black students are more likely to use extreme punitive controls, such as expulsion and calling
the police. Likewise, we expect that racial threat would diminish the willingness of schools
to implement more moderate practices, instead favoring harsh techniques. Pursuant to this
expectation, this study will also test the following hypotheses pertaining to restorative disciplinary responses: (4) Schools with a higher proportion of black students are less likely to use
mild controls, such as parent-teacher conferences and sending students to the counselor; and
(5) Schools with a higher proportion of black students are less likely to implement restitutive
practices, such as community service.
Additionally, we test for possible moderating and contextual effects to determine whether
the racial composition of students is more influential in circumstances in which punitive discipline is less likely, just as previous studies have found stronger racial threat effects in contexts
where less punitiveness is expected. Because prior studies on school discipline have found
such strong and consistent results regarding the predictive qualities of both socioeconomic
disadvantage of students and higher levels of delinquency and drug use, this study will examine the possibility that the effects of racial threat will be stronger under conditions in which
SP5701_03.indd 30
12/21/09 11:33:43 AM
Racial Threat and Punitive School Discipline 31
discipline is not already likely to be high. Therefore, the final two hypotheses this study will
test are that racial composition of students is more influential on discipline in schools with (6) a
smaller percentage of poor students and (7) a lesser prevalence of delinquency and drug use.
Methods
Data
The data for this study come from the National Study of Delinquency Prevention in
Schools (Gottfredson et al. 2000). A probability sample of 1,287 public, private, and Catholic
schools was originally selected, of which 848 (66.3 percent) responded to the 1997 Phase One
Principal Questionnaire. Of these 848 schools, 635 (74.9 percent) responded to the 1998 Phase
Two Principal Questionnaire, which asked different questions from those asked in the first
survey. Student and teacher questionnaires were also administered in 1998, although only in
the secondary schools. Data reported by the U.S. Department of Education (National Center
for Education Statistics 2003, 2007) show that serious disciplinary action peaked during these
years. Of the 558 secondary schools involved in phase two, 310 (55.6 percent) participated in
the student survey and 403 (72.2 percent) participated in the teacher survey. When correlations between school and community characteristics and survey participation were examined,
schools located in small towns or rural areas were significantly more likely to have participated.
In addition, schools were less likely to have participated if they are located in urban areas, communities that have more female-headed households with children, and communities where
more households receive public assistance. The response rates and nonrandom attrition from
the study may, therefore, make this a conservative test of racial threat.5
Certain categories of schools are excluded from these analyses. The sample is limited to
public schools because disciplinary policies and norms vary widely from private and religious
schools; assessing these types of institutions would require separate analyses. In addition,
since previous research on discipline and punishment focuses nearly exclusively on public
schools, this study fits better within established theoretical frameworks (Kupchik and Monahan 2006). Alternative schools for disruptive youth are also excluded, because they include a
large number of outliers on several of the study’s variables of interest. Further, it is limited to
secondary schools, because the student and teacher surveys from which many of the control
variables are taken were only administered in middle and high schools. Thus, the final sample
for this research includes 294 public, nonalternative middle and high schools, which represents the population to which this study’s findings are generalizable.
Measures
Items and scales composed from the principal, teacher, and student questionnaires are
described below and descriptive statistics are provided in Table 1. Individual items included in
each scale are detailed in the Appendix. Scales other than those used to measure disciplinary
5. It is unlikely, however, that the basic results of this study would change had more of these schools been included. Exploratory analyses of potential biases introduced by the limited response rates examined participating schools
that were located in similar communities as the majority of nonparticipating schools. These schools were more likely to
have a greater percentage of black students and more punitive disciplinary responses than other schools. Therefore, it
seems likely that the inclusion of the nonparticipating schools would have resulted in actually intensifying the relationships reported in this study. Of course, it is possible that the relationships of interest are not linear in the region of the
distribution in which nonparticipating schools fall, or that some characteristic might alter the relationships established.
However, the linear relationship between community characteristics and nonparticipation, racial composition of the
student body, and the punitiveness of the disciplinary measures seems to indicate that, if anything, the results presented
here will provide conservative estimates of the relationships.
SP5701_03.indd 31
12/21/09 11:33:43 AM
32
Welch/Payne
Table 1 • Descriptive Statistics for Study Variables
Alpha
N
1.00–3.00
.00–1.00
1.00–3.00
1.00–3.00
1.00–3.00
.61
.82
.74
.86
.57
292
283
288
291
292
22.45
.00–99.69
—
294
26.08
18.73
7.08
.37
.16
.35
1.00
.36
.07
.65
.95
.00–100.00
.00–98.11
25.00–96.80
2.11–4.00
.47–1.25
1.00–2.00
-1.84–7.24
.00–4.00
.00–1.00
-1.20–3.00
-2.32–2.39
—
—
—
.91
.84
.90
.86
.94
.61
—
—
222
263
243
285
287
238
243
287
287
281
281
Measure
Mean
Disciplinary response
Punitive disciplinary response
Zero tolerance
Extreme punitive disciplinary response
Mild disciplinary response
Restitutive disciplinary response
2.44
.54
1.91
2.57
1.82
.35
.27
.27
.39
.38
Racial threat
Percent black students
13.52
Control variables
Percent students free/reduced lunch
Percent Hispanic students
Percent male students
Principal leadership
Administrative leadership
Discipline training
Student delinquency and drug use
Perceived lack of safety
Teacher victimization
Concentrated disadvantage
Urbanicity
33.26
10.29
39.11
3.15
.93
1.63
.00
.85
.16
-.15
-.20
SD
Range
responses are taken from the final report of the National Study of Delinquency Prevention in
Schools, and were developed and copyrighted by Gary Gottfredson (Gottfredson et al. 2000).
Dependent Variable. Five degrees of disciplinary response are operationalized using 22
questions from the Phase Two Principal Questionnaire that asks about possible responses to
student misconduct that may be used by school administrators. Three of these scales pertain
to punitive school responses. Extreme punitive disciplinary response is a four-item scale that
includes court action against the student or parent, expulsion from school, calling or notifying
the police, and charging the student with a crime; possible responses to items are “not used,”
“used,” and “used often.” Zero tolerance is a five-item scale representing whether schools
have an automatic suspension policy for possession of tobacco, alcohol, other drugs, a knife,
or a gun; possible responses to items are “yes” and “no.” Punitive disciplinary response is
a five-item scale that includes suspension from school, in-school suspension, after-school
detention, short-term withdrawal of a privilege (e.g., field access, participation in athletics),
and long-term withdrawal of a privilege; possible responses to items are “not used,” “used,”
and “used often.”
The other two scales of disciplinary response are more restorative in nature. Mild disciplinary response is a five-item scale that includes sending students to the school counselor,
conferences with students’ parents/guardians, oral reprimands, notifying parents/guardians
about student behavior, and conferences with students; possible responses to items are “not
used,” “used,” and “used often.” Finally, restitutive disciplinary response is a three-item scale
that includes restitution (requiring a student to repay the school or victim for damages or
harm done), community service, and work duties; possible responses to items are “not used,”
“used,” and “used often.” For all scales, a school’s score is the mean of the principal’s responses
to each item.6
6. A varimax-rotated factor analysis was run using all 22 disciplinary response items, resulting in a five-factor solution illustrating these five scales. Analyses indicate this solution accounts for 56.17 percent of the variance in disciplinary
response. Detailed results of this factor analysis are available upon request.
SP5701_03.indd 32
12/21/09 11:33:44 AM
Racial Threat and Punitive School Discipline 33
Primary Independent Variable. Racial threat is operationalized by the percent of black
students in each school (percent black students), taken from the Common Core of Data, a
program of the U.S. Department of Education’s National Center for Education Statistics that
collects data on public education (Gottfredson et al. 2000). Use of this variable accords with
previous aggregate-level research on racial threat that uses a range of objective measures of
the racial composition of place.7 Results show that, in this study, black students comprise
13.52 percent of an average school’s student body.
Control Variables. Findings of previous research suggest that certain factors influence the
discipline management of schools or the punishment of students. Accordingly, essential variables are included in this study’s models to control for their potential effects on the dependent
measures of school disciplinary response. Socioeconomic disadvantage is represented in this
study as the percent of students receiving free or reduced lunches (percent students free/
reduced lunch), taken from the Phase One Principal Questionnaire. In this study, the average
school offers 33.26 percent of its students a free or reduced price lunch. In addition, since the
relative size of the Hispanic population has been found to affect the severity of criminal justice
practices (Holmes et al. 1996; Jacobs and Carmichael 2002), a higher Hispanic composition
of student populations may be related to harsher school discipline, despite having never been
tested in schools. Thus, this study will include the percent Hispanic students as a control variable; this item is aggregated from responses to the student survey. Hispanic students comprise
10.29 percent of the average school in this sample. The percent of male students in schools
(percent male students) is also included in this research, and is aggregated from responses to
the student survey; in this study, boys compose 39.11 percent of the average school’s student
population.8
Two separate measures of effective principal leadership are included in this study to control for their potential influence on the outcome variables. Principal leadership is a 19-item
scale from the Phase One Principal Questionnaire that includes items regarding principal supervision and feedback, consideration, presence and visibility, and planning. The principals
were asked about their emphasis on a variety of work activities; four possible responses were
“top,” “high,” “some,” and “little.” A school’s score is the mean of the principal’s responses to
each item. The second measure is administrative leadership, which is a 12-item scale from
the teacher survey that measures the teachers’ perceptions of principal leadership. Examples
of these true/false items include “the administration is supportive of teachers” and “teachers
feel free to communicate with the principal.” A school’s score on this scale is the mean across
teachers of the proportion of items endorsed. Both of these scales are expected to have a positive relationship with the disciplinary response outcomes.
This study also includes a measure of discipline training, which is operationalized by a
scale from the Phase Two Principal Questionnaire. Discipline training assesses the quality and
quantity of the training received by school personnel in its disciplinary procedures, and is predicted to be positively associated with each outcome measure. Scale items include questions
like “how much initial in-service training in school discipline procedures was completed?”
and “how much formal follow-up training on school discipline was completed?,” as well as
various yes/no statement items, including “participants practiced applying the principles” and
“participants’ questions and concerns about possible obstacles in applying the principles were
addressed.” A school’s score on this scale was created by collapsing all nonbinary items and
averaging them to form a proportion of items endorsed.
7. Racial threat has been measured at the level of community (Smith and Holmes 2003), city (Mosher 2001),
county (Demuth and Steffensmeier 2004), state (Behrens et al. 2003), and nation (Jacobs and Kleban 2003).
8. The underrepresentation of male students is somewhat expected given their higher rates of both dropping out
and truancy (Cataldi, Laird, and Kewal Ramani 2009; Dinkes et al. 2009). In addition, this measure is aggregated from
responses to the student survey; therefore, the lower mean could reflect lower survey participation by male students
rather than actual underrepresentation in schools.
SP5701_03.indd 33
12/21/09 11:33:44 AM
34
Welch/Payne
Effects of student crime and delinquency are controlled by the inclusion of student delinquency and drug use, a 28-item scale from the student survey inquiring about the number
of different crimes committed by the student and the variety of different drugs used by
the student during the 12 months prior to the survey; possible responses were “yes” and
“no.”9 A school’s score on this scale is the mean across respondents of the proportion of
items endorsed.
Because various measures of crime salience have also been found to predict punitive
discipline in schools, these analyses include two separate measures of it. Perceived lack of
safety is an eight-item scale, which asks teachers how safe they perceive various places in the
school to be. Possible responses for these items were “very unsafe,” “fairly unsafe,” “average,”
“fairly safe,” and “very safe;” the scale is coded so that higher values indicate higher levels
of perceived lack of safety. A school’s score on this scale is the mean across teachers of their
responses to each item. Teacher victimization is measured by a nine-item scale assessing the
types and amount of victimization experienced by teachers during the 12 months prior to the
survey. A school’s score on this scale is the mean across respondents of the proportion of items
to which teachers responded “yes.”
Finally, characteristics of the surrounding community that may influence punitive responses were measured at the school level using 1990 Census data and school districting
information (Brantlinger 1991; DeVoe et al. 2005; Simonsen 1998; Skiba et al. 1997). Because
influential socioeconomic factors may come from surrounding communities (Welsh, Stokes,
and Greene 2000), this study controls for the poverty, disorganization, and urbanicity of the
census tract in which each school is located.10 Concentrated disadvantage, predicted to increase harsh discipline, is a factor scale that includes the following markers: welfare (average
household public assistance income), female-headed households (ratio of single females with
children under 18 to married couples with children under 18), median income (proportion
of households with income below $27,499), poverty (ratio of persons below the 1.24 poverty
level to persons above), divorce rate (ratio of persons over 15 years who are married to those
who are separated, divorced, or have a spouse absent), and unemployment (proportion of
unemployed males/females in the labor force). Urbanicity, also associated with increased punitiveness in schools, is included as a factor score incorporating the following markers: population size (total population), urban level (city level type), and urbanicity (the proportion of
people living within an urban area). These two scales were created using results from varimax
factor analyses conducted by Simonsen (1998).
Analytic Strategy
The distributional characteristics of these measures were examined first. Both concentrated
disadvantage and urbanicity were trimmed to address three extreme outliers and resulting
skewness.11 Then, five ordinary least squares (OLS) regression models were estimated to test
this study’s first five hypotheses. For these models, each of the disciplinary response outcomes
9. It is plausible that disciplinary punitiveness is related more to student deviance occurring specifically on school
grounds, rather than general student deviance. To examine this possibility, a more specific school delinquency scale was
created that contained only the five deviance items that occurred within the school building, which asked whether a student had committed the following acts during the 12 months prior to the survey: damaged or destroyed school property,
hit or threatened to hit a teacher or other adult at school, hit or threatened to hit other students, stolen or tried to steal
something at school (from classroom, locker, cafeteria, library, etc.), and gone to school when drunk or high on some
drugs. Each regression model was run once with the larger student delinquency and drug use scale and again with this
more specific school delinquency scale. The findings for these two models were similar in both significance and strength.
Results of these analyses are available from the authors upon request.
10. Characteristics of communities in which schools are located have more impact on school crime and disorder
than characteristics of the communities in which the schools’ students live (Welsh et al. 2000).
11. These variables were trimmed by capping the values of the three outliers to three standard deviations above
the variable mean (Wilcox 2004).
SP5701_03.indd 34
12/21/09 11:33:44 AM
Racial Threat and Punitive School Discipline 35
was regressed on percent black students and all control variables.12 For all equations, tolerance
values and variance inflation factors (VIFs) were examined in order to test for multicollinearity. No tolerance values were smaller than .1, and only 5 of 60 VIFs were larger than 2.5, with
the largest of these being 2.943. Thus, multicollinearity is not an apparent problem (Freund
and Littell, 2000). Tests also indicate there is no presence of heteroskedasticity. Due to the directional nature of the hypotheses (positive for the first three and negative for the fourth and
fifth), one-tailed t-values are used to determine statistical significance (Henkel 1976).
To test the final two hypotheses that the effects of racial threat on school punitiveness
vary according to levels of student poverty and student delinquency and drug use, interaction
terms were created by multiplying percent students free/reduced lunch and student delinquency and drug use with percent black students. The five dependent measures of disciplinary response were then regressed on them separately, along with the control variables, to
determine if either interaction term was statistically significant. Each variable comprising the
product terms was mean centered (Cronbach 1987). 13 No violations of regression assumptions
were detected in these models: For each model, only the interaction terms’ tolerance values
were smaller than .1 (ranging from .092 to .098) and only the VIFs associated with the interaction terms and the individual variables that made up these terms were larger than 2.5. These
values are to be expected when using multiplicative interaction terms (Cortina 1993). Again,
one-tailed t-values were used to determine statistical significance.
Finally, to examine the specific effects of the statistically significant interactions, the sample was split at the 50th percentile for each of the conditioning variables used in the product
terms (e.g., greater versus smaller percentage of free/reduced lunch and more versus less delinquency/drug use) and the disciplinary outcome variable was regressed on the percent black
students and the control variables. T-tests were conducted using the unstandardized coefficients and standard errors to test whether the coefficients for percent black students were significantly different in the two equations (t = (b1 – b2)/(standard error12 + standard error22)1/2).
Results
Table 2 presents the results of the OLS regression estimates for the three indicators of
punitive school responses. As predicted in Hypothesis 1, percent black students is significantly and positively related to punitive disciplinary response (b = .007, p < .001), illustrating
that schools with a greater percentage of black students are more likely to use harsh forms
of punishment, such as suspensions, in-school suspensions, and detentions. Not only is this
measure of racial composition statistically significant in this model, it is the strongest predictor
of punitive disciplinary response, with a beta of .404. This is especially notable in light of the
lack of statistical significance of the student delinquency and drug use variable. The only other
significant predictor is greater principal leadership. The R-squared for this equation suggests
that the model explains about one-fifth of the variance in punitive discipline (.205).
12. Some research suggests the possibility of a curvilinear effect of racial threat, in which the racial composition
of a school is positively related to punitiveness up to a certain point and then negatively related to punitiveness after
this point is reached (Hagan, Shedd, and Payne 2005). To examine this idea, a squared percent black students variable
was added to each model while retaining the regular percent black students variable. The squared percent black students
variable was not significant for any of the disciplinary response outcomes, thus not supporting the idea of a curvilinear
effect of racial threat.
13. While some argue that mean centering minimizes the potential effects of multicollinearity in interaction models (Cronbach 1987), it has been a questionable approach to managing this problem (Echambadi and Hess 2007). It is
generally agreed, however, that mean centering is useful for facilitating interpretation of results. In additional tests for
multicollinearity, we found that the independent variables in these models, particularly for the product terms, are stable
and not sensitive to simultaneous inclusion (Jaccard and Turrisi 2003).
SP5701_03.indd 35
12/21/09 11:33:44 AM
36
Welch/Payne
Table 2 • OLS Regression Results for Punitive Disciplinary Response Indicators
Punitive Disciplinary
Response a
Coefficients
Constant
Percent black students
Percent students free/
reduced lunch
Percent Hispanic
students
Percent male students
Principal leadership
Admin. leadership
Discipline training
Student delinquency
and drug use
Perceived lack of safety
Teacher victimization
Concentrated
disadvantage
Urbanicity
Model summary
R-squared
Adjusted R-squared
F statistic
Zero Tolerance a
b
SE
Beta
2.628***
.007***
.002*
.498
.002
.002
.404
.184
.002
.081
.001
.002
.075
.002
-.001
.250***
-.329
.113
.167
.004
.083
.218
.092
.564
-.032
.263
-.153
.114
.027
-.002
-.007
-.065
-.010
.424
.003
.068
.169
.072
.448
-.061
-.011
-.043
-.014
.096
-.001
.062
-.055
.133*
.185
.003 -.042
.068 .083
.178 -.033
.075 .172
.469 .039
.173
-.592
.093
.126
.634
.069
.169
-.106
.159
-.095
-.075
.013
.099 -.132
.490 -.019
.054 .033
.093
-.480
.081
.103 .116
.526 -.108
.056 .178
.024
.034
.065
-.011
.027 -.044
.025
.001
.205
.123
2.499
b
SE
Beta
Extreme Punitive
Disciplinary Response a
.396
.391
.003*** .002 .276
-.001
.001 -.162
.116
.039
1.613
b
SE
1.909*** .407
.003*
.002
.001
.001
.002
.029
Beta
.221
.153
.156
.091
.132
.041
1.455
Dependent variable
a
*p < .05 **p < .01 ***p < .001 (one-tailed tests)
In support of Hypotheses 2 and 3, percent black students is significantly and positively related to both the implementation of zero tolerance (b = .003, p < .001) and the use of extreme
punitive disciplinary response (b = .003, p < .05). Schools with a greater percentage of black
students are more likely to respond even more harshly to misbehavior, such as automatically
suspending students for various violations, expelling students, or even calling the police and
charging students with crimes. Again, the focal independent measure of racial composition
is the most powerful predictor of each of these punitive approaches to discipline (with standardized betas of .276 and .221, respectively). In fact, it is the only significant predictor of the
degree to which schools execute zero tolerance policies. Discipline training is significantly and
positively related to extreme punitive disciplinary response (b = .133, p < .05), demonstrating
that schools in which the adults in the school receive more training in the discipline policies
are more likely to respond very harshly to misbehavior.
Table 3 presents the results of the OLS regression estimates for the two indicators of restorative disciplinary response, which are predicted to negatively relate to greater minority
student composition. The results of the test of Hypothesis 4 reveal that percent black students
is significantly and negatively related to mild disciplinary response (b = -.008, p < .001) and is
also the strongest predictor of it (beta = -.397). As predicted, schools with a larger percentage
of black students are less likely to use milder forms of punishment, such as counselor visits or
oral reprimands, while schools with relatively fewer black students are more likely to respond
to misbehavior in this manner. The percent students receiving free/reduced lunch and administrative leadership are also significantly and negatively related to mild disciplinary response
(b = -.002 and b = -.433, p < .05, respectively), suggesting that schools with a greater percentage of low-income students and in which teachers have a poor perception of their principal’s
SP5701_03.indd 36
12/21/09 11:33:44 AM
Racial Threat and Punitive School Discipline 37
Table 3 • OLS Regression Results for Restorative Disciplinary Response Indicators
Mild Disciplinary Response a
Coefficients
Constant
Percent black students
Percent students free/
reduced lunch
Percent Hispanic students
Percent male students
Principal leadership
Admin. leadership
Discipline training
Student delinquency
and drug use
Perceived lack of safety
Teacher victimization
Concentrated disadvantage
Urbanicity
b
SE
Beta
2.978***
-.008***
.552
.002
-.397
-.002*
.002
.001
.000
.284***
-.433*
.052
.002
.004
.093
.242
.102
Restitutive Disciplinary Response a
b
SE
Beta
.313
-.004*
.522
.002
-.202
-.158
.000
.002
.008
.059
-.009
.266
-.180
.046
.000
.000
.206**
.680***
.039
.002
.004
.087
.228
.097
.001
.003
.210
.305
.037
.254
.629
.037
.712
.591
.112
.222
-.665
.133*
.056
.139
.706
.076
.038
.194
-.106
.202
.139
.000
.994
.071
.057*
.132
.664
.072
.036
.000
.172
.117
.153
Model summary
R-squared
Adjusted R-squared
F statistic
.231
.150
2.873
.186
.102
2.215
Dependent variable
a
*p < .05 **p < .01 ***p < .001 (one-tailed tests)
leadership efficacy are less likely to use these mild forms of discipline. Finally, both principal
leadership and concentrated disadvantage are significantly and positively related to mild disciplinary response (b = .284, p < .001, and b =. 133, p <.05, respectively), indicating that schools
in poorer, more socially disorganized areas and those with principals perceiving themselves to
be effective leaders are more likely to use mild student sanctions.
Finally, as Hypothesis 5 anticipated, percent black students is significantly and negatively
related to restitutive disciplinary response (b = -.004, p < .05). This signifies that schools with
greater percentages of black students are less likely to facilitate practices such as restitution
and community service in response to misbehavior, while schools with smaller percentages of
black students are more supportive of this type of restorative response. Principal leadership,
administrative leadership, and urbanicity are all significantly and positively related to restitutive disciplinary response (b = .206, p < .01; b = .680, p < .001; and b = .057, p < .05, respectively), illustrating that urban schools in which both the principal and the teachers believe
that administrators are effective leaders are more likely to use these practices.
The next set of analyses examined potential conditioning effects of racial threat by regressing the five disciplinary response scales on interactions between percent black students and two
contextual variables: percent students receiving free/reduced lunch and student delinquency
and drug use. Hypotheses 6 and 7 predicted that there would be significant moderated relationships between these variables, such that a higher percent of black students would have
a greater effect on harsh disciplinary responses in schools with a lower percentage of poorer
students and with a lower amount of student delinquency and drug use.
As can be seen in Table 4, most of the interactions are not statistically significant. In fact,
the only significant result observed is the effect that the interaction between percent black
students and student delinquency and drug use has on extreme punitive disciplinary response
(b = -.075, p < .01). This tentatively suggests that for schools with more black students and
SP5701_03.indd 37
12/21/09 11:33:44 AM
SP5701_03.indd 38
Notes: Standard errors in parentheses.
a
Dependent variable
*p < .05 **p < .01 § p = .068 (one-tailed tests)
Model summary
R-squared
Adjusted R-squared
F statistic
Urbanicity
Concentrated disadvantage
Teacher victimization
Perceived lack of safety
Discipline training
Administrative leadership
Principal leadership
Percent male students
Percent Hispanic students
Student delinq/drug use
.206
.116
2.293**
.172
(.566)
.001
(.002)
-.002
(.004)
.251**
(.084)
-.333
(.219)
.116
(.094)
.168
(.128)
-.564
(.646)
.092
(.069)
.025
(.035)
—
Mean-centered delinq/drug use
Percent students free lunch
Mean-centered pct free lunch
.008
(.004)
.002
(.002)
—
2.618**
(.513)
.000
(.000)
—
Mean-centered pct black students
Pct black × delinq/drug use
Pct black × pct free lunch
Constant
.221
.133
2.513**
.001
(.002)
-.002
(.004)
.269**
(.084)
-.401
(.222)
.106
(.092)
.153
(.126)
-.463
(.636)
.079
(.069)
.005
(.036)
-.002*
(.002)
-.583
(.744)
—
-.058§
(.038)
-.013**
(.005)
—
2.681**
(.497)
—
Punitive Disc. Resp. a
.117
.043
.672
.751
(.570)
.003
(.002)
.001
(.004)
.016
(.084)
-.064
(.221)
.024
(.094)
-.150
(.129)
.199
(.650)
.037
(.070)
-.002
(.035)
—
.004
(.004)
-.004
(.002)
—
.193
(.505)
.000
(.000)
—
.117
.043
.677
.003
(.002)
.001
(.004)
.019
(.085)
-.076
(.225)
.020
(.093)
-.151
(.128)
.204
(.646)
.035
(.070)
-.007
(.037)
-.004
(.002)
.593
(.756)
—
-.012
(.038)
.003
(.005)
—
.212
(.505)
—
Zero Tolerance a
.134
.035
1.352
.197
(.471)
.002
(.002)
-.002
(.003)
.063
(.069)
-.060
(.179)
.138
(.076)
.086
(.104)
-.520
(.535)
.080
(.057)
.028
(.029)
—
.004
(.004)
.003*
(.002)
—
1.892**
(.410)
.000
(.000)
—
.175
.081
1.859*
.002
(.002)
.000
(.003)
.036
(.068)
.037
(.178)
.143*
(.074)
.120
(.101)
-.327
(.519)
.100
(.056)
.050*
(.030)
-.002
(.001)
.999
(.602)
—
-.074**
(.030)
.006
(.004)
—
1.844**
(.400)
—
Extreme Pun. Resp. a
.231
.143
2.632**
.249
(.632)
-.001
(.002)
.000
(.005)
.284**
(.093)
-.431
(.243)
.049
(.104)
.226
(.142)
-.689
(.720)
.134
(.077)
.055
(.039)
—
-.007
(.005)
-.002
(.002)
—
2.986**
(.556)
.000
(.000)
—
.234
.146
2.673**
-.001
(.002)
.000
(.004)
.294**
(.094)
-.468
(.248)
.048
(.103)
.213
(.141)
-.603
(.714)
.126
(.077)
.047
(.041)
-.002
(.002)
-.111
(.835)
—
-.028
(.042)
-.011*
(.005)
—
3.004**
(.555)
—
Mild Disc. Response a
Table 4 • Unstandardized OLS Coefficients for Regression of Disciplinary Response Indicators on Interaction Terms
.193
.102
2.116*
.734
(.591)
.000
(.002)
.000
(.004)
.210*
(.087)
.662**
(.229)
.052
(.098)
-.019
(.133)
.909
(.674)
.067
(.072)
.063
(.037)
—
-.008
(.004)
.000
(.002)
—
.273
(.524)
.000
(.000)
—
.187
.095
2.035*
.000
(.002)
.000
(.004)
.210*
(.089)
.665**
(.235)
.037
(.097)
-.005
(.133)
.921
(.673)
.068
(.073)
.053
(.039)
.000
(.002)
.560
(.788)
—
-.012
(.040)
-.005
(.005)
—
.324
(.526)
—
Restitutive Resp. a
38
Welch/Payne
12/21/09 11:33:44 AM
Racial Threat and Punitive School Discipline 39
less delinquency and drug use, extremely harsh discipline is more often implemented. Only
one other term, the interaction between percent black students and student delinquency and
drug use on punitive disciplinary response, approaches significance at the .05 level (b = -.058,
p = .068, t = 1.548). The interaction between percent black students and student delinquency
and drug use is not related to the other three disciplinary response outcomes and the interaction between percent black students and percent students receiving free/reduced lunch is not
related to any of the disciplinary response outcomes.
To further investigate the significant and near-significant interactions, the data were split
at the fiftieth percentile of student delinquency and drug use (median = .110), then extreme
punitive disciplinary response was regressed on percent black students and all control variables (Table 5). In schools with high levels of student delinquency and drug use, percent
Table 5 • OLS Coefficients with Student Delinquency/Drug Use Split at the Fiftieth Percentile
High Student Delinq/Drug Use
B
Coefficients
Constant
Percent black students
Percent students free lunch
Percent Hispanic students
Percent male students
Principal leadership
Administrative leadership
Discipline training
Student delinquency/drug use
Perceived lack of safety
Teacher victimization
Concentrated disadvantage
Urbanicity
Model summary
R-squared
Adjusted R-squared
F statistic
a
Beta
Low Student Delinq/Drug Use
B
SE
Extreme Punitive Disciplinary Response .696
1.627**
.609
.002
.074
.007**
.003
.002
-.103
-.003
.002
.003
.069
.004
.002
.006
-.014
-.002
.004
.104
.110
.076
.102
.262
-.125
.121
.275
.109
.068
.170
.113
.707
-.129
2.149
1.772
.167
-.054
.171
.171
.669
-.099
-.462
1.064
.074
.370*
-.028
.101
.051
.264
.004*
.002
Beta
a
1.908**
.001
-.001
.001
.000
.086
-.212
.056
-.660
-.046
-.417
.160
.075
Model summary
R-squared
Adjusted R-squared
F statistic
Coefficients
Constant
Percent black students
Percent students free lunch
Percent Hispanic students
Percent male students
Principal leadership
Administrative leadership
Discipline training
Student delinquency/drug use
Perceived lack of safety
Teacher victimization
Concentrated disadvantage
Urbanicity
SE
2.593**
.004
-.003
.001
.004
.162
-.501*
.106
.048
.083
-1.068
.104
-.009
.183
.028
1.046
.226
.044
1.436
.779
.003
.002
.004
.007
.112
.282
.118
.762
.181
.760
.080
.055
Punitive Disciplinary Response a
2.134**
.839
.225
.010**
.003
-.282
-.001
.003
.068
.003
.003
.089
-.008
.006
.186
.534**
.141
-.264
-.028
.378
.117
.094
.155
.008
4.472*
2.430
.086
.403
.236
-.227
-.433
1.409
.215
-.002
.137
-.028
.070
.060
.244
.083
1.505
.415
-.305
.357
-.059
.108
.074
.231
.173
.232
-.093
-.055
.357
.594
-.003
.181
-.194
-.518
-.012
.087
.243
.378
-.063
-.003
.164
.341
.190
2.160*
Dependent variable
**p < .01 *p < .05 (one-tailed tests)
SP5701_03.indd 39
12/21/09 11:33:45 AM
40
Welch/Payne
black students has no significant effect on extreme punitive disciplinary response. However,
in schools with low levels of student delinquency and drug use, percent black students has a
significant positive effect on extreme punitive disciplinary response (b = .007, p < .01), as predicted. The results of the t-test confirm this significant difference (t = -1.664, p < .01). Thus, for
schools with lower levels of delinquency and drug use, those that have a greater percentage of
black students are more likely to respond in a very harsh manner to student misbehavior.
Finally, similar findings can also be seen for the effect of the interaction between percent
black students and student delinquency and drug use on punitive disciplinary response (also
shown in Table 5), which had approached significance at the .05 level. As before, in schools
with high levels of student delinquency and drug use, the effect of percent black students on
punitive disciplinary response is nonsignificant. However, in schools with low levels of student delinquency and drug use, percent black students has a significant and positive effect on
punitive disciplinary response (b = .010, p < .01). The results of the t-test show that the difference approaches significance at the .05 level (t = -1.414). Again, for schools with lower levels
of delinquency and drug use, those that have a greater percentage of black students are more
likely to respond in a harsh manner to student misbehavior. These findings indicate that the
percentage of black students in a school has a stronger effect on punitive discipline practices
in schools with lower levels of student deviance.
Discussion
This research is the first to test—and support—the racial threat hypothesis in the school
setting. Specifically, it examines whether schools with a higher proportion of black students
are more likely to use punitive and extremely punitive discipline, more likely to implement
zero tolerance policies, and less likely to use mild and restitutive approaches. The results of
OLS regression estimates show that each of the first five hypotheses of this study are supported in the expected directions: Schools with a larger composition of black students are
more likely to respond to student misbehavior in a harsh manner and less likely to respond
restoratively.
In addition, this study examined the potential moderating effects of both student economic disadvantage and student deviance, predicting that the proportion of black students would
have a greater influence on discipline in schools with a smaller percentage of poor students
and a lower amount of student delinquency and drug use. The results of OLS regression estimates show that these hypotheses were supported only for student delinquency and drug use,
and then only for the outcomes of punitive and extremely punitive discipline. The percentage
of black students has a stronger effect on punitive and extremely punitive discipline practices
in schools with lower levels of student deviance. Thus, in schools where students engage in
less delinquency and drug use, it appears that schools may be more likely to respond harshly
and extremely harshly to misbehavior partly because of the racial composition of the student
body. In conditions where delinquency and drug use are high, it is likely that student punishment is already so intense that there is less opportunity for other factors, including the racial
composition of students, to have an appreciable effect. This finding corroborates what was
expected based on prior research analyzing contextual differences in racial threat. In addition,
even though the hypothesis predicting an interaction between racial composition of students
and student poverty was not supported, results do confirm that the effect of racial threat on
harsh school discipline remains present, regardless of the socioeconomic status of the students.
This inconsistency of the presence of contextual effects accords with previous research.
The racial threat perspective is supported here in large part because this study was able
to control for several important crime-specific influences. It is clear that there is a residual
effect of racial composition of schools on disciplinary policies even after controlling for the
effects of school delinquency, drug use, teacher victimization, and teacher perceptions of lack
SP5701_03.indd 40
12/21/09 11:33:45 AM
Racial Threat and Punitive School Discipline 41
of safety—variables that were measured irrespective of race. This indicates that there is some
race-specific factor that persists in increasing punitiveness, regardless of levels of crime and
crime salience in schools. It is plausible that the relationship found between racial composition and school discipline may be attributable to unmeasured individual-level intervening
influences. Specifically, it may actually be that among school officials, perceived threat associated with blacks is higher in institutions in which proportionally more black students are matriculated, which may then contribute to the intensification of discipline. For example, racial
composition may influence race-related factors, such as stereotypes, bias, or discrimination.
In addition, racial composition may influence factors related to both race and crime, such as
the fear of black crime or the stereotyping of blacks as criminals, which may then facilitate the
expansion of punitive discipline. Based on the findings of previous research on the effects of
micro-level factors on punitive controls reviewed earlier, these possible explanations for the
established effect of racial composition on disciplinary policies are worth exploring.
This study’s findings are important because they demonstrate that the effects of racial
threat extend beyond criminal justice institutions traditionally found to be influenced by it.
Racial threat is not only associated with harsh sanctions for criminal offenders, but with the
punitive treatment of students in school. This also suggests that the effects of racial threat may
be operating even beyond these two social institutions.
The findings reported here also have substantial implications for school discipline. It is
striking that the implementation and use of harsh school policies appears to at least be partially attributable to the relative size of schools’ minority populations. If racial threat is to some
degree responsible for the intensified punishment—and sometimes even criminalization—of
students, then it has contributed to what has been identified as “a heavy toll on students,
teachers, and the entire school community” who are suffering the consequences of diminished focus on quality education (Noguera 1995a:189). By being subjected to more punitive
discipline, like suspensions, expulsions, and arrests, certain students are not equally benefiting
from education—if they are not in class, they will not have learning opportunities commensurate with that of their peers, the result of which is further academic disadvantage. Not surprisingly, research confirms that suspensions are linked to academic failure, grade level retention,
and dropping out of school (Skiba and Peterson 2000). This not only affects future prospects of
employment, but also the likelihood of involvement in crime and the criminal justice system
(Gregory and Weinstein 2008; Hirschfield 2008). Furthermore, although this study does not
specifically measure effects for particular students, consequences of racial threat in schools
may be presumed to be most significant for black students, since harsher discipline is practiced
in schools where they are present in larger percentages.
Study Limitations and Directions for Future Research
It is prudent to consider the limitations of this study, some of which derive from methodological shortcomings. The cross-sectional nature of the data makes it impossible to truly
determine causality in the associations found.14 Also, because the sample is restricted to public, secondary, nonalternative schools, the findings are only applicable to these types of educational institutions. In addition, sample response rates are potentially problematic, although
14. For instance, while the positive association between the percent of black students in a school and the existence
of punitive disciplinary responses could suggest the effects of racial threat as found here, it is also possible that a shift
in the composition of student populations occurred after the creation of certain discipline policies, thus challenging this
study’s conclusions. However, prior research using time-series data, for example, also supports the racial threat perspective in relation to severity of criminal policy (Kent and Jacobs 2004), so the use of more longitudinal data would likely
add important dimensions to the findings revealed here. Further, it should also not be assumed that racial threat is responsible for the trend of increased punitiveness that began in the early 1980s, because the cross-sectional nature of this
data make that untestable. In addition, a significant relationship between other variables, such as student delinquency
and discipline, may be suppressed by the use of cross-sectional data.
SP5701_03.indd 41
12/21/09 11:33:45 AM
42
Welch/Payne
some researchers have shown that studies with low participation can actually demonstrate
more or equally accurate results than studies with higher response rates (Keeter et al. 2006;
Visser et al. 1996). Because schools located in urban areas, in communities with more femaleheaded households with children, and where more households receive public assistance were
less likely to have participated in the study, the results of this research may not generalize as
well to schools located in such communities, although results would likely not have diminished had more of these schools been included.
Another possible limitation of this research is that the relationships herein identified are
attributable to certain key constructs that are not included in the regression estimates. For
example, this study was not able to control for or contextualize by the potential influence of
regional variation or poor academic ability of students, which may be associated with harsh
student discipline (Jacobs and Carmichael 2004; Wu et al. 1982). It might be useful to control
for or contextualize by more specific details related to the delinquency and disorder in schools,
as discipline likely varies according to levels of violent, property, and drug offenses as well as
less serious acts of student mischief. Additionally, school climate and structure may influence
disciplinary practices (Payne 2008).
Finally, a greater understanding of how racial threat operates in schools may be derived
by future research that examines the micro-processes involved in creating the effects of racial
threat. For example, this study was not able to determine whether it is black students who
were being punished more harshly within schools. While establishing individual-level relationships is neither a necessary, nor common, component of tests of racial threat, studying
various dimensions of racially disparate applications of discipline, racial stereotypes, bias, and
discrimination among local policy makers, administrators, and teachers may partially explain
how it affects school policies. Previous studies have noted the importance of racial beliefs and
prejudice to student treatment (Brown and Beckett 2006; Ferguson 2000; Giroux 2003; Noguera 2003b; Skiba and Peterson 1999; Skiba et al. 2002; Watts and Erevelles 2004), but none
have connected it by multivariate analyses with punitive disciplinary outcomes in schools. It
may also be instructive to test whether perceptual indicators of racial threat are influential
within schools and whether other school characteristics, such as gender distributions, moderates the relationships found in this research.
Conclusion
This is the first multivariate test of the effects of racial composition of students on punitive
school discipline. This research supports the racial threat hypothesis by showing that schools
are more punitive and less restorative when there are more black students enrolled in them.
This is independent of the types and levels of delinquency in those schools and independent
of other influences that have been shown to affect discipline practices, such as student economic status and gender, school crime salience, school urbanicity, and the training of faculty
and administrative staff. In addition, while the proportion of black students in schools affects
punitive and extremely punitive punishments even more when school delinquency and disorder are low, the effects of racial threat are otherwise not subject to contextual differences.
Thus, racial threat remains a powerful explanation for greater use of harsh discipline and reduced use of restorative discipline, regardless of the amount of school misbehavior or student
economic disadvantage. These findings suggest that not only do the trends in school policies
and punishment mirror those that are sanctioned by the criminal justice system, but that part
of the reason is also similar. The conclusions of this study reveal important findings about the
broad extent of racial threat and the applicability of its influence in various settings. They also
expose notable flaws in school discipline policies and practices and call for serious consideration of ways to rid them of their racially disparate application.
SP5701_03.indd 42
12/21/09 11:33:45 AM
Racial Threat and Punitive School Discipline 43
Appendix. Items Included in Study Scales
Disciplinary Responses
Punitive Disciplinary Response
Suspension from school (exclusion of student from membership for periods of 30 days or less)
Brief exclusion of student from attendance in regular classes (e.g., in-school suspension)
After-school detention
Short-term (5 days or less) withdrawal of a privilege (e.g., playground, athletic participation)
Long-term (more than 5 days) withdrawal of a privilege (e.g., playground, athletic participation)
Zero Tolerance (automatic suspension if . . . )
Possession of alcohol
Possession of other drugs (e.g., marijuana, LSD, cocaine)
Possession of a knife
Possession of a gun
Extreme Punitive Disciplinary Response
Court action against student or parent
Expulsion from school (exclusion of student from membership for periods of time over 30 days)
Calling or notifying the police
Charging student with a crime
Mild Disciplinary Response
Sending student to school counselor
Conferences with student’s parents/guardians
Oral reprimand
Notifying parents about student’s behavior
Conference with student
Restitutive Disciplinary Response
Restitution (requiring a student to repay the school or a victim for damages or harm done)
Community service
Work duties, chores, or tasks as punishment
Control Variables
Principal Leadership
Tour the school to establish my presence
Observe teachers’ instruction and classroom management practices
Formally assess the needs or problems of the school
Use reason or passion to generate staff commitment to tasks
Plan staff meetings
Discuss quality of work performance with staff members
Evaluate the effectiveness of existing school practices
Check with teachers before making changes that may affect them
Assign responsibilities to teachers
Review teacher performance with individual teachers in a formal evaluation
Discuss alternative plans for school improvement with staff, district personnel, or community
Praise teachers or recognize effective staff performance
Mention observed strengths and weaknesses in performance to teachers at the time of observation
Establish policies or standard operation procedures to cover most day-to-day decisions
Be patient with and helpful to faculty
Communicate performance expectations
Review progress on improvement plans with individual staff members
Set schools improvement goals, taking into account such things as time, resources, obstacles, cost
Offer support or sympathy when a staff member experiences a difficulty
Teacher Victimization (This year in school have any of the following happened to you in this school?)
Damage to personal property worth more than $10.00
Theft of personal property worth less than $10.00
Theft of personal property worth more than $10.00
Was physically attacked and had to see a doctor
Was physically attacked but not seriously enough to see a doctor
Received obscene remarks or gestures from a student
SP5701_03.indd 43
12/21/09 11:33:45 AM
44
Welch/Payne
Was threatened in remarks by a student
Had a weapon pulled on me
Perceived Lack of Safety (At your school during school hours, how safe from vandalism, personal
attacks, and theft is each of the following places? (reverse coded))
Your classroom while teaching
Empty classrooms
Hallways and stairs
The cafeteria
The restrooms used by the students
Locker room or gym
Parking lot
Elsewhere outside on school grounds
Administrative Leadership
The school’s administration makes it easy to get supplies, equipment, or other arrangements.
In your opinion, how well do teachers and administrators get along at your school?
Administrators and teachers collaborate toward making the school run effectively.
There is little administrator-teacher tension in this school.
Our principal is a good representative of our school before the superintendent and the board.
The principal is aware of and lets staff and students know when they have done something well.
Teachers or students can arrange to deviate from the prescribed program of the school.
Teachers feel free to communicate with the principal.
The administration is supportive of teachers.
It is hard to change established procedures here.
The principal of our school is informal.
The principal of our school is open to staff input.
Discipline Training
If there was in-service training in discipline, which of the following describe the training?
The presentation was clear and organized.
Principles to be followed were presented.
Principles were illustrated with examples.
Participants practiced applying the principles.
Participants received feedback on their performance in applying the principles.
Participants’ questions and concerns about possible obstacles were addressed.
How much formal follow-up training on school discipline was completed by the average individual
who manages discipline?
None
One occasion
Two occasions
Three or more occasions
Student Delinquency and Drug Use (In last 12 months, have you . . . )
Purposely damaged or destroyed property belonging to a school?
Purposely damaged or destroyed property that did not belong to you?
Stolen or tried to steal something worth more than $50?
Carried a hidden weapon other than a pocket knife?
Been involved in gang fights?
Hit or threatened to hit a teacher or other adult at school?
Hit or threatened to hit other students?
Taken a care for a ride (or drive) without the owner’s permission?
Used force or strong-arm methods to get money or things from a person?
Stolen or tried to steal things worth less than $50?
Stolen or tried to steal something at school?
Broken into or tried to break into a building/car to steal something or just to look around?
Belonged to a gang that has a name and engages in fighting, stealing or selling drugs?
Sold marijuana or other drugs?
Smoked cigarettes?
Used smokeless tobacco?
Drunk beer, wine, or “hard” liquor?
SP5701_03.indd 44
12/21/09 11:33:45 AM
Racial Threat and Punitive School Discipline 45
Gone to school when you were drunk or high on some drugs?
Sniffed glue, paint, or other spray?
Smoked marijuana (weed, grass, pot, hash, ganja)?
Taken hallucinogens (LSD, mescaline, PCP peyote, acid)?
Taken sedatives (barbiturates, downers, Quaaludes, reds)?
Taken amphetamines (uppers, speed, whites)?
Taken tranquilizers (Valium, Librium)?
Taken heroine (horse, smack)?
Taken cocaine (coke)?
Used crack?
Used other narcotics (codeine, Demerol, dilaudid)?
Taken steroids?
References
Advancement Project. 2005. “Education on Lockdown: The Schoolhouse to Jailhouse Track.” Retrieved
January 15, 2009 (www.advancementproject.org/reports/FINALEOLrep.pdf).
American Psychological Association (APA) Zero Tolerance Task Force. 2006. “Are Zero Tolerance Policies
Effective in the Schools? An Evidentiary Review and Recommendations.” Washington, DC: American Psychological Association Council of Representatives.
Artiles, Alfredo J. 2003. “Special Education’s Changing Identity: Paradoxes and Dilemmas in Views of
Culture and Space.” Harvard Educational Review 73:164–202.
Arum, Richard. 2003. Judging School Discipline: The Crisis of Moral Authority. Cambridge, MA: Harvard University Press.
Barkan, Steven E. and Steven F. Cohn. 2005. “Why Whites Favor Spending More Money to Fight Crime:
The Role of Racial Prejudice.” Social Problems 52:300–14.
Baumer, Eric, Steven F. Messner, and Richard Rosenfeld. 2003. “Explaining Spatial Variation in Support
for Capital Punishment: A Multilevel Analysis.” American Journal of Sociology 108:844–75.
Beger, Randall R. 2002. “Expansion of Police Power in Public Schools and the Vanishing Rights of Students.” Social Justice 29:119–30.
Behrens, Angela, Christopher Uggen, and Jeff Manza. 2003. “Ballot Manipulation and the ‘Menace of
Negro Domination’: Racial Threat and Felon Disenfranchisement in the United States, 1850–2002.
American Journal of Sociology 109:559–605.
Blalock, Hubert. 1967. Towards a Theory of Minority Group Relations. New York: Capricorn Books.
Bontrager, Stephanie, William Bales, and Ted Chiricos. 2005. “Race, Ethnicity, Threat, and the Labeling of
Convicted Felons.” Criminology 43:589–622.
Brantlinger, Ellen. 1991. “Social Class Distinctions in Adolescents’ Reports of Problems and Punishment
in School.” Behavioral Disorders 17:36–46.
Brooks, Kim, Vincent Schiraldi, and Jason Ziedenberg. 1999. School House Hype: Two Years Later. San Francisco: Center on Juvenile and Criminal Justice.
Brown, Lionel H. and Kelvin S. Beckett. 2006. “The Role of the School District in Student Discipline:
Building Consensus in Cincinnati.” The Urban Review 38:235–56.
Bryk, Anthony, Penny Bender Sebring, David Kerbow, Sharon Rollow, and John Q. Easton. 1998. Charting
Chicago School Reform: Democratic Localism as a Lever for Social Change. Boulder, CO: Westview Press.
Cataldi, Emily F., Jennifer Laird, and Angelina KewalRamani. 2009. “High School Dropout and Completion Rates in the United States: 2007.” Washington, DC: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Retrieved November 18, 2009
(http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2009064).
Chamlin, Mitchell B. 1989. “A Macro Social Analysis of Change in Police Force Size 1972–1982: Controlling for Static and Dynamic Influences.” Sociological Quarterly 30:615–24.
Chiricos, Ted, Kelly Welch, and Marc Gertz. 2004. “Racial Typification of Crime and Support for Punitive
Measures.” Criminology 42:359–89.
Cortina, Jose M. 1993. “Interaction, Nonlinearity, and Multicollinearity: Implications for Multiple Regression.” Journal of Management 19(4):915–22.
Crawford, Charles, Ted Chiricos, and Gary Kleck. 1998. “Race, Racial Threat, and Sentencing of Habitual
Offenders.” Criminology 36:481–511.
SP5701_03.indd 45
12/21/09 11:33:45 AM
46
Welch/Payne
Cronbach, Lee J. 1987. “Statistical Tests for Moderator Variables.” Psychological Bulletin 87:51–57.
Crow, Matthew S. and Katherine A. Johnson. 2008. “Race, Ethnicity, and Habitual-Offender Sentencing:
A Multilevel Analysis of Individual and Contextual Threat.” Criminal Justice Policy Review 19:63–83.
DeMitchell, Todd A. and Casey D. Cobb. 2003. “Policy Responses to Violence in Our Schools: An Exploration of Security as a Fundamental Value.” Education & Law Journal 2:459–84.
Demuth, Stephen and Darrell Steffensmeier. 2004. “Ethnicity Effects on Sentence Outcomes in Large Urban
Courts: Comparisons among White, Black, and Hispanic Defendants.” Social Science Quarterly 85:994–1011.
Devoe, Jill F., Katharin Peter, Margaret Noonan, Thomas D. Snyder, and Katrina Baum. 2005. Indicators
of School Crime and Safety 2005. Washington, DC: U.S. Department of Education and the U.S. Department of Justice.
Di Lullo, Louis. 2004. “Appropriate Conduct.” Principal Leadership (High School Ed.) 4:27–29.
Dinkes, Rachel, Emily Forrest Cataldi, Wendy Lin-Kelly. 2007. “Indicators of School Crime and Safety:
2007.” NCES 2007-NCES 2008, U.S. Department of Education and U.S. Department of Justice.
Washington, DC: Government Printing Office.
Dinkes, Rachel, Jana Kemp, and Katrina Baum. 2009. Indicators of School Crime and Safety: 2008. Washington DC: National Center for Education Statistics.
Echambadi, Raj and James D. Hess. 2007. “Mean-Centering Does Not Alleviate Collinearity Problems in
Moderated Multiple Regression Models.” Marketing Science 26:438–45.
Fantz, Ashley. 2008. “Children Forced into Cell-Like School Seclusion Rooms. CNN.com, December 17.
Retrieved December 17, 2008 (www.cnn.com/2008/US/12/17/seclusion.rooms/index.html).
Ferguson, Ann A. 2000. Bad Boys: Public Schools in the Making of Black Masculinity. Ann Arbor: University of
Michigan Press.
Foucault, Michel. 1977. Discipline and Punish: The Birth of the Prison. New York: Pantheon Books.
Freund, Rudolf J. and Ramon C. Littell. 2000. SAS System for Regression. Cary, NC: SAS Institute.
Fuentes, Annette. 2003. “Discipline and Punish: Zero Tolerance Policies Have Created a ‘Lockdown Environment’ in Schools.” The Nation 277:17–20.
Giroux, Henry A. 2003. “Racial Injustice and Disposable Youth in the Age of Zero Tolerance.” International
Journal of Qualitative Studies in Education 16:553–65.
Gorman, Kevin and Patrick Pauken. 2003. “The Ethics of Zero Tolerance.” Journal of Educational Administration 41:24–36.
———. 2001. “What Schools Do to Prevent Problem Behavior and Promote Safe Environments.” Journal
of Educational and Psychological Consultation 12:313–44.
Gottfredson, Gary D., Denise C. Gottfredson, Ellen R. Czeh, David Cantor, Scott B. Crosse, and Irene
Hantman. 2000. A National Study of Delinquency Prevention in School: Final Report. Ellicott City, MD:
Gottfredson Associates.
Gottfredson, Gary D., Elizabeth M. Jones, and Thomas W. Gore. 2002. “Implementation and Evaluation
of a Cognitive-Behavioral Intervention to Prevent Problem Behavior in a Disorganized School.” Prevention Science 3:43–56.
Gregory, Anne and Rhona S. Weinstein. 2008. “The Discipline Gap and African Americans: Defiance or
Cooperation in the High School Classroom.” Journal of School Psychology 46:455–75.
Gregory, James F. 1995. “The Crime of Punishment: Racial and Gender Disparities in the Use of Corporal
Punishment in U.S. Public Schools.” Journal of Negro Education 64:454–62.
Hagan, John, Carla Shedd, and Monica R. Payne. 2005. “Race, Ethnicity, and Youth Perceptions of Criminal Injustice.” American Sociological Review 70:381–407.
Henault, Cherry. 2001. “Zero Tolerance in Schools.” Journal of Law & Education 30: 547–53.
Henkel, Ramon E. 1976. Tests of Significance. Newbury Park, CA: Sage Publications.
Hilarski, Carolyn. 2004. “How School Environment Contribute to Violent Behavior in Youth.” Journal of
Human Behavior in the Social Environment 9:165–78.
Hirschfield, Paul J. 2008. “Preparing for Prison? The Criminalization of School Discipline in the USA.”
Theoretical Criminology 12:79–101.
Holmes, Malcolm D., Harmon M. Hosch, Howard C. Daudistel, Dolores A. Perez, and Joseph B. Graves.
1996. “Ethnicity, Legal Resources, and Felony Dispositions in Two Southwestern Jurisdictions.” Justice Quarterly 13:11–30.
Jaccard, James and Robert Turrisi. 2003. Interaction Effects in Multiple Regression. 2d ed. Thousand Oaks,
CA: Sage.
Jacobs, David and Jason T. Carmichael. 2002. “The Political Sociology of the Death Penalty: A Pooled
Time-Series Analysis.” American Sociological Review 67:109–23.
SP5701_03.indd 46
12/21/09 11:33:45 AM
Racial Threat and Punitive School Discipline 47
———. 2004. “Ideology, Social Threat, and the Death Sentence: Capital Sentences across Time and Space.”
Social Forces 83:249–78.
Jacobs, David and Richard Kleban. 2003. “Political Institutions, Minorities, and Punishment: A Pooled
Cross-National Analysis of Imprisonment Rates.” Social Forces 80:725–55.
Jacobs, David and Ronald Helms. 1999. “Collective Outbursts, Politics, and Punitive Resources: Toward a
Political Sociology of Spending on Social Control.” Social Forces 77:1497–1524.
Keeter, Scott, Courtney Kennedy, Michael Dimock, Jonathan Best, and Peyton Craighill. 2006. “Gauging
the Impact of Growing Nonresponse on Estimates From a National RDD Telephone Survey.” Public
Opinion Quarterly 70:759–79.
Kent, Stephanie L. and David Jacobs. 2004. “Social Divisions and Coercive Control in Advanced Societies:
Law Enforcement Strength in Eleven Nations from 1975 to 1994.” Social Problems 51:343–61.
King, Ryan D. 2007. “The Context of Minority Group Threat: Race, Institutions, and Complying with Hate
Crime Law.” Law & Society Review 41:189–224.
Kupchik, Aaron and Torin Monahan. 2006. “The New American School: Preparation for Post-Industrial
Discipline.” British Journal of Sociology of Education 5:617–31.
Lasley, Thomas J. and William W. Wayson. 1982. “Characteristics of Schools with Good Discipline.” Educational Leadership 40:28–31.
Liska, Allen E. 1992. Social Threat and Social Control. Albany: SUNY Press.
Massachusetts Advocacy Center. 1986. The Way Out: Student Exclusion Practices in Boston Middle Schools. Boston: Massachusetts Advocacy Center.
McCarthy, John D. and Dean R. Hoge. 1987. “The Social Construction of School Punishment: Racial Disadvantage Out of Universalistic Process.” Social Forces 65:1101–20.
Meier, Kenneth J., Joseph Stewart, and Robert E. England. 1989. Race, Class, and Education: The Politics of
Second-Generation Discrimination. Madison: University of Wisconsin Press.
Mosher, Clayton. 2001. “Predicting Drug Arrest Rates: Conflict and Social Disorganization Perspectives.”
Crime & Delinquency 47:84–104.
National Center for Education Statistics. 2003. “Indicators of School Crime and Safety, 2002.” Washington, DC: U.S. Department of Education. Retrieved February 5, 2009 (http://www.ojp.usdoj.gov/bjs/
pub/pdf/iscs02.pdf).
———. 2007. “Indicators of School Crime and Safety, 2007.” Washington, DC: U.S. Department of Education. Retrieved February 5, 2009 (http://www.ojp.usdoj.gov/bjs/pub/pdf/iscs07.pdf).
Nichols, Joe D. 2004. “An Exploration of Discipline and Suspension Data.” The Journal of Negro Education
73:408–23.
Noguera, Pedro A. 1995a. “Preventing and Producing Violence: A Critical Analyses of Response to School
Violence.” Harvard Educational Review 65:189–212.
———. 1995b. “Reducing and Preventing Youth Violence: An Analysis of Causes and an Assessment of
Successful Programs.” Pp. 25–43 in Wellness Lectures, edited by California Wellness Foundation. Oakland: California Wellness Foundation and the University of California, Berkeley.
———. 2003a. “Schools, Prisons, and Social Implications of Punishment: Rethinking Disciplinary Practices.” Theory into Practice 42:341–50.
-———. 2003b. “The Trouble with Black Boys: The Role and Influence of Environmental and Cultural
Factors on the Academic Performance of African American Males.” Urban Education 38:431–59.
Parker, Karen, F., Brian J. Stults, and Stephen K. Rice. 2005. “Racial Threat, Concentrated Disadvantage, and Social Control: Considering the Macro-Level Sources of Variation in Arrests.” Criminology
43:1111–33.
Payne, Allison A. 2008. “A Multilevel Analysis of the Relationships among Communal School Organization, Student Bonding, and Delinquency.” Journal of Research in Crime and Delinquency 45:429–55.
Robbins, Christopher G. 2008. Expelling Hope: The Assault on Youth and the Militarization of Schooling. Albany,
NY: SUNY Press.
Shaw, Steven R. and Jeffery B. Braden. 1990. “Race and Gender Bias in the Administration of Corporal
Punishment.” School Psychology Review 19:378–83.
Simon, Jonathan. 2006. Governing Through Crime: How the War on Crime Transformed American Democracy and
Created a Culture of Fear. New York: Oxford University Press.
Simonsen, April A. 1998. “The Effects of Community Disorganization on School Administrative Practices:
Implications for Delinquency Prevention Practice.” Masters Thesis, Department of Criminology and
Criminal Justice, University of Maryland, College Park.
Singer, Simon. 1996. Recriminalizing Delinquency. Cambridge, UK: Cambridge University Press.
SP5701_03.indd 47
12/21/09 11:33:45 AM
48
Welch/Payne
Skiba, Russell J. 2000. “Zero Tolerance, Zero Evidence: An Analysis of School Disciplinary Practice.” Policy
Research Report. Bloomington: Indiana Education Policy Center.
———. 2001. “When is Disproportionality Discrimination? The Overrepresentation of Black Students in
School Suspension.” Pp. 176–87 in Zero Tolerance: Resisting the Drive for Punishment in Our Schools: A
Handbook for Parents, Students, Educators, and Citizens, edited by W. Ayers, B. Dohrn, and R. Ayers. New
York: The New Press.
Skiba, Russell and Reece L. Peterson. 1999. “The Dark Side of Zero Tolerance: Can Punishment Lead to
Safe Schools?” Phi Delta Kappan 80:372–82.
———. 2000. “School Discipline at a Crossroads: From Zero Tolerance to Early Response.” Exceptional
Children 66:335–47.
Skiba, Russell J., Reece L. Peterson, and Tara Williams. 1997. “Office Referrals and Suspension: Disciplinary Intervention in Middle Schools. Education and Treatment of Children 20:295–315.
Skiba, Russell J., Robert S. Michael, Abra Carroll Nardo, and Reece L. Peterson. 2002. “The Color of Discipline: Sources of Racial and Gender Disproportionality in School Punishment.” The Urban Review
34:317–42.
Smith, Brad W. and Malcolm D. Holmes. 2003. “Community Accountability, Minority Threat, and Police
Brutality: An Examination of Civil Rights Criminal Complaints.” Criminology 41:1035–63.
Stolzenberg, Lisa, Stewart J. D’Alessio, and David Eitle. 2004. “A Multilevel Test of Racial Threat Theory.”
Criminology 42:673–97.
Taylor, Marylee C. 1998. “How White Attitudes Vary with the Racial Composition of Local Populations:
Numbers Count.” American Sociological Review 63:512–35.
Toby, Jackson. 1998. “Getting Serious about School Discipline.” The Public Interest 133:68–84.
Tolnay, Stewart E., E. M. Beck, and James L. Massey. 1992. “Black Competition and White Vengeance:
Legal Execution of Blacks as Social Control in the Cotton South, 1890 to 1929.” Social Science Quarterly 73:627–44.
Vavrus, Frances and KimMarie Cole. 2002. “’I Didn’t Do Nothin’: The Discursive Construction of School
Suspension.” The Urban Review 34:87–111.
Visser, Penny S., Jon A. Krosnick, Jesse Marquette, and Michael Curtin. 1996. “Mail Surveys for Election
Forecasting? An Evaluation of the Columbus Dispatch Poll.” Public Opinion Quarterly 60:181–227.
Wacquant, Loïc. 2001. “Deadly Symbiosis: When Ghetto and Prison Meet and Mesh.” Pp. 82–120 in Mass
Imprisonment: Social Causes and Consequences, edited by David Garland. London, UK: Sage Publications.
Watts, Ivan E. and Nirmala Erevelles. 2004. “These Deadly Times: Reconceptualizing School Violence by
Using Critical Race Theory and Disability Studies.” American Educational Research Journal 41:271–99.
Welch, Kelly. 2007. “Black Criminal Stereotypes and Racial Profiling.” Journal of Contemporary Criminal
Justice 23:276–88.
Welsh, Wayne N. 2000. “The Effects of School Climate on School Disorder.” The ANNALS of the American
Academy of Political and Social Science 567:88–107.
Welsh, Wayne, Robert Stokes, and Jack R. Greene. 2000. “A Macro-Level Model of School Disorder.”
Journal of Research in Crime and Delinquency 37(3)243–83.
Wilcox, Rand R. 2004. “Power: Basics, Practical Problems, and Possible Solutions.” Pp. 65–86 in Handbook
of Psychology, Volume 2: Research Methods in Psychology, edited by John A. Schinka, Wayne F. Velicer, and
Irving B. Weiner. New York: John Wiley & Sons.
Wu, Shi-Chang, William Pink, Robert Crain, and Oliver Moles. 1982. “Student Suspension: A Critical
Reappraisal.” The Urban Review 14:245–303.
SP5701_03.indd 48
12/21/09 11:33:45 AM