Crime and punishment: the role of student body characteristics in schools’ disciplinary behaviours

ABSTRACT

Discretion in schools’ discipline choices can provide an efficient and effective misconduct management structure, but could lead to discipline based on unrelated factors. Consequently, schools’ disciplinary decisions can significantly limit students’ access to education by removing students from familiar learning environments. We investigate schools’ disciplinary decisions for serious misconducts and show that punishments are more severe in schools that do not report misconducts to local law enforcement agencies. Moreover, we show that schools that report fewer misconducts to law enforcement impose more severe punishments when the student body is characterized as having a higher proportion of minority students, lower socioeconomic status students and a higher proportion of students who are below the 15th percentile of standardized test scores. These results suggest that between-school punishment differentials are associated with student body traits.

KEYWORDS:

• Disciplinary behaviour
• punishment disparity
• school administrators
• school discipline
• student characteristics

JEL CLASSIFICATION:

• I20
• J15

Notes

¹ A recent survey of state laws mandating prescribed disciplinary actions (The Civil Rights Project 2000) indicates that most states mandate disciplinary actions for both serious offences (loosely analogous to those that could be tried as felony or criminal cases for adults) and less serious misconducts (such as disobedience, persistent absenteeism, unlawful activity, defiance of authority figures and other vaguely defined behavioural actions). However, despite the existence of such mandates, many states allow school districts to apply disciplinary discretion when composing student conduct protocols.

² A parallel may be drawn to the federal justice system, where sentencing discrepancies have been shown to exist mostly across judges, rather than across decisions made by any particular judge (for example, see Ashenfelter, Eisenberg, and Schwab 1995; Schanzenbach 2005; Iyengar 2011).

³ We limit the responses to those from non-alternative high schools. The exclusion of alternative high schools reduces the sample by less than 5%. Observation counts are rounded to the nearest 10 s to comply with data licence restrictions.

⁴ These offences are considered violent or harmful – the most serious offence category – by a large proportion of US school districts, including three of the largest: Miami-Dade County School Board, New York City Department of Education and Chicago Public Schools.

⁵ Statistics are presented for schools with at least 500 students.

⁶ The SSOCS also provides information about the proportion of students who are American Indian, Asian, or of another race. These three categories characterize 7.5% of the student population in the sample. We do not focus on these groups because there are insufficient data to make confident inferences from the empirical results.

⁷ For robustness, the racial concentration levels were varied between 10% and 50% and showed no qualitative difference in rankings.

⁸ We have also considered controlling for principal characteristics using responses from the Schools and Staffing Survey (SASS) collected by the National Center for Education Statistics. However, relatively few schools in the SSOCS are those for which principal information is available in the SASS and neither the SSOCS school-level nor the SASS principal-level data specify whether the principal or another school administrator is responsible for making disciplinary decisions. Consequently, school-level disciplinary behaviours may be inappropriately modelled if principal information is used. Moreover, Kinsler (2011) shows that differences in principal characteristics do not significantly explain differences in school discipline rates.

⁹ The SSOCS also provides information about a third punishment category, light discipline, which includes sanctions such as detention or no discipline. Consequently, discipline severity trade-offs are made within the context of three punishment categories, implying that the substitution behaviour is not immediately clear (as it otherwise would be if there were only two categories). If administrators decide to use less of one of the disciplinary actions, they will then need to choose how to reallocate their disciplinary decisions among the two remaining categories. A priori, it is unclear which of the two remaining types of punishment will adjust and by how much.

¹⁰ A comparison of the presented results to those from models excluding gang activity indicate that the latter overestimates the effects of a school’s student racial composition. For example, for permanent removals, the marginal effects of increasing a school’s black and Hispanic student body were upward biased by 15% and 22.5% as a result of omitting the gang activity control.