Untangling the School Segment of The School-to-Prison Pipeline
Nancy López and Jane Hood gave this PowerPoint presentation on August 9 2009 at the annual meeting of the Society for the Study of Social Problems in San Francisco. Our presentation reports on results from the RWJF Center for Health Policy - funded research on the effects of race, class, gender, and special education status on high school discipline for all freshmen in a large New Mexico school district. In addition, we have included some illustrative quotations from an ongoing qualitative study of 10th graders in a large high school in another Although this presentation focuses on the place of school discipline in the school-to-prison pipeline, our research has many direct and indirect links to health and health policy, the most direct of which is the link between school success and life chances. The results of our multivariate analysis find a circular relationship between basic skills acquisition, school success, and discipline. Students who do not learn basic skills are more likely to get into trouble than those who have mastered these skills. Since we are studying freshmen only in this research and have data only on students who were enrolled in school, we cannot estimate the effect of discipline on drop-out rates. However, other research shows a direct link between school suspension and the probability of dropping out of school (Wald and Losen, 2003).
Summary: With the exception of studies by Russell Skiba and his colleagues, research on the school segment of the school-to-prison pipeline has rarely included multivariate analyses of school district data. Instead we have multiple studies that show gender, race, class, and sometimes special education disparities in school discipline. Consistently, black students are suspended at two to three times the rate of white students, and Skiba’s research demonstrates that the gap persists even after one controls for seriousness of offense. However, previous studies have shown not only that race, class, gender, special education status, and academic failure are related to disciplinary outcomes, but also that all of these variables are related to each other. Therefore, without a multivariate analysis including the most important of these measures, it is hard to know why students of color get in trouble more often than others. Our analysis of basic skills testing and disciplinary data for 6846 southwestern high school freshmen allows us to focus on Hispanics and English Language Learners while also looking at African Americans, Native Americans, and Asians. Having observed in a previous classroom observation study that gifted students seemed to be shielded from disciplinary consequences, we include gifted as well as non-gifted special education status in our analyses. And, because previous studies (Wald and Olson 2003) have found that reclassified freshmen had as disproportionate share of discipline problems, we control for age as well as basic skills scores in our analyses.
Logistic regression results show that, indeed, our independent variables are related both to each other and to school discipline in complex ways. For example, non-white students are actually less likely than whites to be placed in special education once class (free lunch) and basic skills scores as well as the interaction between class and race are taken into account. Although our analysis does not find African American students to be more likely than whites to get disciplinary referrals, once black students do get a referral, they are 2.74 times more likely than whites to get at least some out-of school suspension even after controlling for sped/gifted status, age, gender, and seriousness of infraction. However, for gifted/SPED status, class, and gender, the reverse is true. Here differences show up at the referral level with gifted students 1/3 as likely as others to get referrals, males 1.56 more likely than females to get them, and students receiving free lunch 1.36 times as likely as paid lunch kids to get referrals. However, once they get referrals, males are surprisingly not more likely than females to be suspended. Finally, we found that overage and SPED students are more likely to both get referrals and to be suspended than are their counterparts.
Our data analysis does not, however, show a direct connection between school discipline and juvenile incarceration patterns. For example, Hispanics predominate in the juvenile justice population, but Native Americans are the ones most likely to be arrested at school. The missing segment may well be what happens outside of school once kids stop going to school.
Using insights from focus group interviews we are doing in a similar school district and some observations done in a previous qualitative study of the study district, we suggest some ways in which the dynamics of race, gender, pedagogy, and school engagement may help to explain these patterns.
References
Losen, D.J. and G. Orfield (2002). Racial Inequity in Special Education. Cambridge, MA: Havard University Press.
Skiba, Russell J. et al. 2002. "The Color of Discipline: Sources of Racial and Gender Disproportionality in School Punishment," The Urban Review. 34 (4): 317-342.
2006. "Disparate Access: The Disproportionality of African American Students with Disabilities Across Educational Environments," Exceptional Children. 72 (4): 411-424.
Wald, J. and D J. Losen (2003). Deconstructing the School-to-Prison Pipeline, Fall 2003 Issue of New Directions for Youth Development (San Francisco CA: Jossey-Bass).
