Please Do Not Cite Or Distribute Without Authors Permission

GAPS IN STUDENT DISCIPLINE 1

Gaps in Student Discipline by Race and Family Income

Nathan Barrett
Tulane University
Andrew McEachin
RAND Corporation

Jonathan Mills
University of Arkansas
Jon Valant
Brookings Institution

Last updated: October 24, 2017

Please do not cite or distribute without authors’ permission.

Abstract

This study examines differences in school discipline for black and white, and poor and non-poor, students in the state of Louisiana.We do so in three ways. First, to understand the origins and potential remedies of gaps in student suspensions, we decomposeraw gaps by differences that arise across districts, across schools within the same district, and within schools. We observe gaps arising from each of these sources, including substantial within-school discipline differences for students of different race and family income. Second, we assess discipline gaps in an assortment of panel data regression models. Weconsistently observe that black and poor students are suspended at higher rates and for longer amounts of time than their white and non-poor peers, with differences related to both nonviolent and violent infractions. Third, we testwhether schools punish students differently for similar behaviors by comparing the punishments resultingfromfights between white and black students and fights between poor and non-poor students. We find that black and poor students receive (modestly)harsher punishments for these fights, even controlling for students’ prior records of discipline and other background characteristics.

I. Introduction

In the United States, students of color are suspended and expelled from school at much higher rates than white students.The U.S. Department of Education’s Office for Civil Rights (2016) reports that, compared to white children, black children are 3.6 times more likely to receive an out-of-school suspension in preschool, 3.8 times more likely to receive an out-of-school suspension in grades K-12, and 2.2 times more likely to be referred to law enforcement or subject to a school-related arrest.Among K-12 students, 18% of black boys and 10% of black girls received an out-of-school suspension in 2013-14, compared to only 5% of white boys and 2% of white girls.Although the Office for Civil Rights does not release similar comparisons for poor and non-poor students nationwide, researchers have observed higher suspension rates for Arkansas students from low-income families than their peers (Anderson & Ritter, 2017) and found that black students who attend high-poverty schools are suspended at higher rates than black students who attend other schools (Loveless, 2017).These gaps are concerning both because of the possibility that black and poor students are punished more harshly than other students and because exclusionary discipline is at least correlated with numerous negative outcomes for students (Steinberg & Lacoe, 2017).

Fundamental questions remain about discipline gaps, in part because identifying their causes is empirically challenging.Gapscould arise from actual differences in student behaviors across groups, differences in how schools respond to the same behaviors from different groups, or a combination of the two.Ideally, researchers would be able to observe the true behaviors of many students, observe whether and how schools punish those behaviors, and then assess whetherschools treat students of different backgrounds similarly.Unfortunately, this type of analysis is not feasible at a large scale using conventional data, which tend toinclude information about infractions but notbehaviors.

What is possible, however, is a close examination of student discipline by race and poverty status to isolate where the gaps arise.For example, dogaps arise withinschools, across schools in the same district, oracross districts?To what extent do observably similar student behaviors give rise to disparate punishments?Answers to these questions would helpto diagnose the problems related to discipline gaps and identify their solutions.If thegaps arise from true differences in students’ behaviors, then eliminating gaps might require addressing the root causes of student misbehavior.If gaps arise within schools that punish black and poor students more harshly than other students, then reducing gapsmight require oversight and training.If, on the other hand, gaps arise acrossschools or districts(due to harsher discipline practices in schools with higher proportions of black or poor students), policymakers might need tocompelcertain schools to adopt less punitive practices.

This study uses statewide, student-level data from Louisiana fromthe 2000-01 through 2013-14 school years to examine gaps in exclusionary discipline betweenblack and white students and between poor and non-poor students (as defined by eligibility for free- or reduced-price lunch, or FRPL).[1]These data enable us to explore, within and across schools, whether schools punish black and poor studentsmore often and more harshly than they punish white and non-poor students.Furthermore, the data distinguish between potentiallyviolent or harmful offenses (such as possessing a weapon) and seemingly nonviolent offenses (such as disrespecting authority), which provides an opportunity to assess what types of offenses lead to students missing school and developing records of misbehavior.

This paper focuses on three sets of analyses.First, we decompose raw,black/whiteand poor/non-poor gaps indiscipline into across-district, across-school-within-district, and within-school components. Second, to examine possible explanations for these gaps, we estimate the conditional probabilities of various discipline outcomes by race and poverty status, controlling for observable student characteristics, details about the infraction, and various combinations of school, grade, and year fixed effects.Third, to assess whether schools treat students of different backgrounds similarly, we test for gaps in a context in which gaps seemingly should not arise: fights between black and white or poor and non-poor students with similar prior discipline records.

Our findings reveal clear patterns in exclusionary discipline by race and poverty status.First, large discipline gaps exist. Black students account for 46% of the student population during this period but 64% of the student suspensions, while poorstudents account for 62% of the population but 74% of the suspensions.These disparitiesare the product of substantial within-school and across-school gaps in discipline.Within a regression framework, being black and poor are consistently significant predictors of several discipline outcomes—being suspended, being suspended multiple times in the same year, and length of suspension in days—even in models with rich sets of covariates.Among those with no prior suspensions, black and poor students are more likely than their peers to be suspended for both a nonviolent offense and a violent offense, although suspensions for nonviolent offenses are more common. Finally, when black and white students or poor and non-poor students (with similar discipline records) fight each other in school, black and poor students tend to receive longer suspensions than their white and non-poor schoolmates—perhaps the strongest empirical evidence to date of discriminatory discipline practices within schools.

The paper proceeds with a description of the existing research on student discipline gaps.We then describe our data, methods, and results, before concluding with discussion of the study’s implications for diagnosing and addressing these gaps.

II. Background on Discipline Gaps

A. The Incidence and Effects of Exclusionary Discipline

Concerns about the overuse of exclusionary disciplinary practices have persisted for decades, along with concerns that the practices disproportionately affect certain populations.The Children’s Defense Fund (1974, 1975) released reports in the mid-1970s with profiles of suspended and expelled students, noting that the suspension rate for black children was twice as high as the rate for any other group and that many of those suspensions were imposed for non-dangerous, nonviolent infractions.More recently, in the summer of 2011, the U.S. Departments of Education and Justice announced a collaborative project, the Supportive School Discipline Initiative, that would “support the use of school discipline practices that foster safe, supportive, and productive learning environments while keeping students in school” (U.S. Department of Education, 2016).They reported that of the 49 million students enrolled in public schools in 2011-12, 3.5 million received an in-school suspension, 3.45 million received an out-of-school suspension, and 130,000 were expelled.Suspension and expulsion rates, according to their analysis, were three times higher for black students than white students.[2]

Many states and districts have responded to the heightened attention to discipline gaps—and clarification of the related Federal laws—by limiting, postponing, or outright banning exclusionary discipline practices (Anderson & Ritter, 2017).Steinberg and Lacoe (2017) reported that 22 states and the District of Columbia had revised their laws as of May 2015 to deemphasize exclusionary discipline, while 23 of the largest 100 school districts have required non-punitive discipline strategies or limits to the use of suspension.They noted evidence in U.S. Department of Education data of a recent decline in the rates of exclusionary discipline from 2006 to 2011.Such a decline is evident in California, which has emphasized reducing suspensions for relatively minor “disruption” or “willful defiance” incidents that account for a large portion of the gap in black and white suspension rates (Losen, Martinez, & Okelola, 2014; Loveless, 2017).A recent study of student discipline in Maryland also shows exclusionary discipline rates on the decline across student racial categories, although large black/white discipline gaps remain (Porowski, O’Conner, & Passa, 2014).

Experts have seen these rates, and the associated gaps between student subgroups, as problematic in part because of the associations between exclusionary discipline and negative outcomes for students.Being suspended from school is correlated with lower academic achievement (Arcia, 2006; Beck & Muschkin, 2012; Raffaele Mendez, Knoff, & Ferron, 2002; Raffaele Mendez, 2003; Skiba & Rausch, 2004), lower probability of on-time graduation (Ekstrom, Goertz, Pollack, & Rock, 1986; Raffaele Mendez, 2003; Suh, Suh, & Houston, 2007; Wehlage & Rutter, 1986), and greater contact with the juvenile justice system via the “school-to-prison pipeline” (Morgan, Salomon, Plotkin, & Cohen, 2014; Fabelo et al., 2011; Nicholson-Crotty, Birchmeier, & Valentine, 2009; Skiba, Arredondo, & Williams, 2014).However, these relationships are largely correlational, leaving unanswered questions about whether suspended students’ outcomes would have been better had they not been suspendedor expelled (Steinberg & Lacoe, 2017).Yet the lack of causal evidence does not mean that causal relationships do not exist, and there remains an intuitive argument that students are more likely to learn when they are present in school.[3]

B. Possible Explanations for Gaps

While certain groups of students are suspended and expelled at higher rates than their peers, basic questions about the causes of those gaps remain unanswered.Understanding gaps in student discipline requires attention to student behaviors, infractions, and punishments, and the relationships between them.We refer to behaviorsas what students do in school,infractionsasbehaviorsthat schools document as misconduct, and punishmentsas the penalties associated with thoseinfractions.In this context, discipline gaps could arise through some combination of three basic pathways.First, there could be actual differences in the behaviors of black or poor students and their peers.Second, there could be differences in the ways that schools translate behaviors to infractions for black and poor students relative to their peers (e.g., if schools more commonly write up black students for the same behaviors).Third, there could be differences by race and income in the waysthat schools translate infractions to punishments (e.g., if schoolssuspend black students for more days than white students for similar infractions).

The first possible explanation for discipline gaps—not exclusive of the others—is that studentsof different races and family incomes simply behave differently from one another. These differences, if they exist, could arise within schools (if poor and minority students behave differently from their schoolmates) or across schools (if students in schools with higher proportions of black or poor students behave differently from students in other schools).

Discipline gaps that arise from true differences in behaviors may not be attributable to discriminatory practices by schools. If, for example, black and white students are suspended at different rates or lengths of time but those differences are proportionate to these groups’ actual rates and severity of misbehavior, then schools might be responding in nondiscriminatoryways to the behaviors they observe. Of course, this exonerates neither schools nor broader societal forces from contributing to varying levels of misbehavior. School culture can affect student behavior, perhaps especially for poor and minority students (Haynes, Emmons, & Ben-Avie, 1997), and black and poor students might behave differently from their peers due to their disproportionate experiences with poverty and the challenges that accompany it.This includes challenges related to physical health (Chen, 2004), stress and anxiety (Bradley & Corwyn, 2002), exposure to violence (Gorman-Smith & Tolan, 1998), and varying norms and home environments (Leventhal & Brooks-Gunn, 2000).A review of the related literature by Gregory, Skiba, and Noguera (2010) concluded that poverty-related factors likely account for part of thediscipline gaps observed but that black/white gaps often persist after controlling for socioeconomic status (SES).These black/white differences among students of similar SES might reflect actual differences in behaviors, with a variety of possible explanations that include cultural mismatches between students and teachers (Monroe, 2005), perceived pressures to “act white” (Fordham & Ogbu, 1986; Fryer, 2006), or students’ reactions to other forms of bias in schools (Gershenson, Holt, & Papageorge, 2016).True behavior differences are hard to identify empirically, since researchers seldom have data on students’ actual behaviors beyond whether schools recorded those behaviors as infractions.

A second explanationfor disciplinegapsis that schools might translate behaviors to infractions differently for students of different races or family income.Whether a student is referred to the office for punishment depends on both the student’s behavior and the teacher’s perception of, or tolerance for, that behavior (Skiba & Williams, 2014).If a black student and white student engage in the same behaviors, the black student could be more likely to receive an infraction if teachers or administrators disregard, forgive, or never notice the white student’s misbehavior.This would be more plainly reflective of systemic discrimination. It could arise across schools—without any particular school or person treating different groups differently—if schools with predominantly poor or minority students respond to misbehavior differently from other schools. It also could arise within schools as a direct result of discriminatory behavior.

Two recent studies provide suggestive evidence.Gilliam, Maupin, Reyes, Accavitti, and Shic (2016) asked 135early childhood teachers to observe videos of four children—ablack boy, black girl, white boy, and white girl—to watch for “challenging behavior in the classroom,” noting to the teachers that “sometimes this involves seeing behavior before it becomes problematic.”The researchers selected videos with, in their view, no signs of challenging behavior, and then tracked participants’ eye movements to see whether teachers tended to monitor some students more than others when looking for misbehavior.They found that teachers focused a disproportionate share of their time gazing at black children, and particularly black boys.They accompanied this task with a vignette experiment to test whether teachers’ attitudes differed toward students with stereotypicallyblack and white names.They found that teachers rated white students’ misbehavior as more severe, which they interpreted as consistent with a “shifting standards” hypothesis (Biernat & Manis, 1994).They found no relationship between the children’s race and whether teachers thought they should be suspended or expelled.Okonofua and Eberhardt (2015) experimented with names in a similar context.With a group of K-12 teachers as study participants, they described students who had misbehaved twice, randomly assigning participants to see either stereotypicallyblack or whitestudent names.They found that teachers felt more troubled by black students’ misbehaviors and were more inclined to regard the black students as troublemakers.A second experiment, designed similarly to the first, showed that teachers were more likely to see themselves suspending the black students in the future.

If, as these studies suggest, teachers monitor and assess similar-behaving black and white students in their classrooms differently, we might expect within-school racial gapsin how schools translate behaviors to infractions.At the same time,across-schoolgaps could arise if schools with higher proportions of black students tend to write up students for behaviors that would not yield write-ups in other schools.An assortment of theories and findingssupport this possibility.Welch and Payne (2010) looked to schools to test the “racial threat” theory (Blalock, 1967) that authorities use more aggressive forms of control in settings with higher proportions of blacks relative to whites.Using survey responses from 294 public middle and high schools, they found that leaders of schools with higher proportions of black students reported being more likely to use punitive and extremely punitive discipline, with those differences persisting even after controlling for an assortment of school characteristics.[4]This stricter enforcement in schools with higher proportions of black students seems consistent with a trend toward “no excuses” philosophies in urban schools (Thernstrom & Thernstrom, 2004). It is also consistent with findings that black/white gaps in office referrals often arise not from severe or objectively clearmisbehaviors but rather from behaviors like loitering and disrespectful conduct that are more subject to the discretion of school officials (Shaw & Braden, 1990; Skiba et al., 2002).

Anderson and Ritter (2017) used longitudinal, student-level infraction data from Arkansas to study exclusionary discipline gaps within and across schools. The authors conditioned their analysis on a sample of students who were referred to the principal’s office for a behavioral infraction. Discipline outcomes were coded either exclusionary (i.e., expulsion or out-of-school suspension) or inclusionary/no discipline (i.e., in-school suspension or no discipline). Usinglogistic regression models that controlled for student and infraction characteristics (but omitted school fixed effects), they found that black students are approximately 2.4 times as likely as white students—and FRPL-eligible students are 1.2 to 1.5 times as likely as non-FRPL students—to receive exclusionary discipline. While the FRPL results persisted in models with school fixed effects, the race results did not, leading them to conclude that the key driver of black/white discipline gaps was likely related to differences in the types of schools that black and white students attend.[5]