In this systematic literature review, we examine the corpus of empirical studies in education that use administrative data (i.e., population-level data) to describe and estimate the impacts of service delivery models for specially designed instruction on outcomes for students identified with special education needs. We focus on studies that use quantitative data analysis--either descriptive or causal--to answer questions about the relationship between special education service delivery models and student outcomes. We analyze seven studies, each of which finds a positive relationship between more time spent in general education classrooms and outcomes for students with disabilities (SWDs). In our analysis, we discuss the affordances and limitations of this type of analysis and opportunities for the field to expand data collection and analysis of population-level data in a way that better illuminates the state of special education services, both in the present and longitudinally, for SWDs.