The current study grows out of earlier research--conducted by one of us (with a collaborator) and published in book form (1) in 2006--on the following problem: In many public programs that seek to improve social conditions and increase opportunity for low-income people, a relatively small number of participants, whom the study termed "bad apples," (2) so severely disrupt those programs that they prevent the rest of the participants from gaining the programs' full benefits. The public school system is the most important venue in which the bad apples problem occurs, and the current study seeks to increase our understanding of how the system tries to manage this problem.
Roughly 1.1 million students are enrolled in the New York City public schools each year. (3) Over the period spanning the 2005-06 and 2009-10 school years, the schools issued 323,680 suspensions, or about 64,736 a year. (4) This works out to significantly less than 5.9% of the total enrolled students in the system during those five years because many of the suspended students were repeat offenders. These suspended students are responsible for a great deal of the time, energy, money, and other resources that the system expends on student discipline and classroom management. Even more importantly, they are presumably responsible for much of the disruptive behavior that impedes classroom learning for all students.
Past studies have found that the New York City school system has higher suspension rates for students of color, students who are male, and students with special needs. (5) The system also may fail to ensure that all students, even the more disruptive ones, receive the support they need to succeed academically or to at least complete high school. This study attempts to probe more deeply into the data on suspensions in New York City and to learn how schools use suspensions (to improve behavior, to aid classroom management, to improve test scores, etc.) and, thus, to consider whether and how they can use suspensions more effectively--for example, by issuing them more or less frequently or in different ways.
The introduction that follows, taken verbatim from an earlier project, (6) presents some necessary background for the current study:
Widespread, intense concern about the harmful effects of disruptive
students on their peers is evident in public opinion polls and
surveys of both teachers and students. This finding is hardly
surprising. The externality and public good aspects of classroom
education mean that just as students can learn from each other as
well as from their teachers, so too can one student's misconduct
quickly cascade through the classroom, thereby reducing learning
for all. (7)
For purposes of this discussion, we call such chronically disruptive students "bad apples."
Although disruptive behavior and discipline problems can be defined in different ways, data sets in this area often distinguish among four crude categories of misconduct: disruption, nonserious violence, violence, and criminal violence. (8) According to one report, more than seventy-seven percent of public elementary and secondary schools suffered at least one violent incident (as defined in the report) during the 1999-2000 year, and students were more likely to fear being attacked at school than being attacked when away from school. (9) Schools that reported a large number of serious discipline problems were more likely to experience violent criminal incidents, including rape, sexual battery, physical attacks, and robbery. (10)
Not surprisingly, the risk of encountering violence in school is concentrated in poor areas and is greatest for economically disadvantaged students. (11) Some believe that noncriminal incidents, which are of course far more common than violent crime, are more responsible for disrupting the educational environment and making it more difficult for students to learn. …