Restorative practices statistically replicate racial inequality in discipline, even while reducing overall school suspension rates and improving school climate. To understand why, the researcher conducted a year-long, multi-case ethnography examining restorative practices at three public secondary small schools in New York City. All three school sites enrolled majority students of color and mostly White staff, thus reflecting the norm for urban school districts across the United States. Data were collected through 2700 hours worth of observations; document analysis; and semi-structured interviews with students, teachers, and administrators (900 hours per school). Data were coded using the en vivo method of coding (Saldana, 2013). Codes were grouped into subcodes, and then into four main themes, which became the findings of the larger study. This article examines one major finding from the study: while restorative practices are intended as an alternative to zero tolerance policies, they are in part legitimized by the presence of zero tolerance policies in the districtwide discipline code. Administrators invoked suspension to earn credibility in the eyes of students, parents, teachers, and fellow principals in the buildings where their schools were collocated (i.e., sharing space).