Background/Context/Setting: In the wake of growing evidence regarding the negative impacts of suspension on student outcomes, states and school districts have implemented disciplinary reforms, restricting or eliminating the use of suspension for minor misbehaviors. Recent evaluations of these reforms suggest that they have the potential to positively impact those students at high risk for suspension (Craig & Martin, 2023; Cleveland, 2023; Pope & Zuo, 2024), yet, the potential for negative peer effects (Pope & Zuo, 2024; Steinberg & Lacoe, 2018) and the use of unintended management strategies (Al Khafaji-King, 2024; Wang, 2022) remain a concern. However, no research has examined the impact of these reforms on an exceptionally vulnerable population: students with disabilities. This lack of research is surprising given the large and growing proportion of SWDs as well as the sensitivity of this group to changes in schooling environments (O'Hagan et al., 2024). Moreover, concurrent research suggests that suspension bans may induce school staff to classify students with disabilities when suspension is no longer permitted (Al Khafaji-King, 2024). However, ex ante, the impact of these ban-induced classifications is unclear. It may be the case that suspension prohibited students from receiving needed services and these students would benefit from service provision (Schwartz et al., 2021). However, and in line with hypotheses of disability critical race theory (DisCrit) (Annamma et al., 2013), it may also be the case that these classifications serve as a mechanism of exclusion (Cruz et al., 2021), and be harmful, especially for Black students (Ballis & Heath, 2021). Purpose/Objective/Research Question: This paper evaluates two potential hypotheses (needed services vs. mechanism of exclusion) as to how SWDs may be uniquely affected by New York City's ban on suspension for minor offenses, first implemented in 2012/13. Understanding the saliency of these hypotheses in the suspension ban context is exceptionally important, given the likelihood that at least some of these classifications are being used as mechanisms of exclusion (Al Khafaji-King, 2024). This work also pays explicit attention to the counterfactual: even if these classifications are intended as mechanisms of exclusion, are they worse than suspension? Understanding how these hypotheses may manifest in these policy environments is essential to creating equitable and efficient disciplinary and disability policy. Specifically, I ask: what is the impact of receiving a presumably ban-induced classification relative to a classification that is less-likely to be ban-induced on test scores and attendance? Data Collection and Analysis: I use data provided by the New York City Public Schools (NYCPS), which contains information on student disability status, disability type, service setting, standardized math and reading scores for grades 3-8, attendance rates, and demographic information (race, sex, etc.). I retain students that are classified in grades 6-12 given that the suspension ban only affected these grades. I group students into three mutually exclusive categories based on their primary disability classification (Fish, 2019): high status, low status, and stratified status disabilities. High status disabilities include autism, other health impairment (usually ADHD diagnoses), and speech or language impairments. Low status disabilities include emotional disturbance and intellectual disabilities. Stratified status composes one classification: learning disabilities. I group schools into those more affected by the ban and those less affected by the ban to provide further treatment contrast. I begin by estimating the suspension ban's differential impact for SWDs' test scores and attendance using an event study triple-difference model. Specifically, I estimate how outcomes for SWDs in above-median suspension schools (more affected by the suspension ban) differentially change relative to outcomes for general education students in above-median suspension schools or SWDs in below-median suspension schools. This model assists in understanding how SWDs' outcomes change on average, not necessarily how receiving a ban-induced classification affects my outcomes of interest. I then use a triple-differences model with individual fixed-effects to compare students classified with either a low, stratified, or high-status classification in a high suspension school pre- and post-classification to students classified in a low suspension school, relative to those students who are never classified. As students are classified in different years, I use the Callaway & Sant'Anna (2021) estimator, which ensures that my comparison group does not compare newly classified students to already classified students ("forbidden comparisons"). I estimate cohort-specific effects, which allows me to compare effect estimates for those cohorts classified prior to the ban to effect estimates for those classified after the ban. For example, if I observe that the difference in the effects of classification between those classified in above-median suspension schools and below-median suspension schools changes post-ban, then I hypothesize that these changes are likely due to the differences in the type of student that is being classified. Findings: I find that, on average, SWDs with low-status disabilities were negatively impacted by the suspension ban, whereas SWDs with high-status or stratified-status disabilities were not differentially affected relative to general education students. Specifically, I find that SWDs with low-status disabilities in above-median suspension schools experienced significant declines in both math and reading scores relative to GENs in above-median schools (Figure 1). General education students experienced slight test score improvements. Notably, I show that test score declines for students with low-status disabilities are likely not due to new, ban-induced classifications actively harming student achievement, that is, the suspension policy environment did not seem to significantly impact the efficacy of service provision (Figure 2). These results underscore the importance of considering unintended consequences and vulnerable groups when employing a seemingly "costless" and popular policy lever to reduce schools' reliance on suspension. Conclusions: This paper provides the first estimates of the effects of special education classification using heterogeneous-treatment effect robust methods, estimates of the impacts of suspension bans on SWDs broadly, as well as how the macro-policy environment (suspension bans) impacts more micro-level policies (special education provision). The results imply that future policy should explicitly consider how SWDs fare when modifying suspension/behavioral policy and how special education can be modified to respond to these macro-level policy changes.