Multiple Choice: Charter School Performance in 16 States

Introduction

As charter schools play an increasingly central role in education reform agendas across the United States, it becomes more important to have current and comprehensible analysis about how well they do educating their students. Thanks to progress in student data systems and regular student achievement testing, it is possible to examine student learning in charter schools and compare it to the experience the students would have had in the traditional public schools (TPS) they would otherwise have attended. This report presents a longitudinal student‐level analysis of charter school impacts on more than 70 percent of the students in charter schools in the United States. The scope of the study makes it the first national assessment of charter school impacts.

Charter schools are permitted to select their focus, environment and operations and wide diversity exists across the sector. This study provides an overview that aggregates charter schools in different ways to examine different facets of their impact on student academic growth.

The group portrait shows wide variation in performance. The study reveals that a decent fraction of charter schools, 17 percent, provide superior education opportunities for their students. Nearly half of the charter schools nationwide have results that are no different from the local public school options and over a third, 37 percent, deliver learning results that are significantly worse than their student would have realized had they remained in traditional public schools. These findings underlie the parallel findings of significant state‐by‐state differences in charter school performance and in the national aggregate performance of charter schools. The policy challenge is how to deal constructively with varying levels of performance today and into the future.

Project Approach

CREDO has partnered with 15 states and the District of Columbia to consolidate longitudinal student‐level achievement data for the purposes of creating a national pooled analysis of the impact of charter schooling on student learning gains. For each charter school student, a virtual twin is created based on students who match the charter student’s demographics, English language proficiency and participation in special education or subsidized lunch programs. Virtual twins were developed for 84 percent of all the students in charter schools. The resulting matched longitudinal comparison is used to test whether students who attend charter schools fare better than if they had instead attended traditional public schools in their community. The outcome of interest is academic learning gains in reading and math, measured in standard deviation units.

Student academic learning gains on reading and math state achievement tests were examined in three ways: a pooled nationwide analysis of charter school impacts, a state‐by‐state analysis of charter school results, and an examination of the performance of charter schools against their local alternatives. In all cases, the outcome of interest is the magnitude of student learning that occurs in charter school students compared to their traditional public school virtual twins. Each analysis looks at the impact of a variety of factors on charter school student learning: the state where the student resides, the school’s grade‐span, the student’s background, time in charter schools, and a number of policy characteristics of the charter school environment.

Summary of Findings 

Charter school performance is a complex and difficult matter to assess. Each of the three analyses revealed distinct facets of charter school performance. In increasing levels of aggregation, from the head‐to‐head comparisons within communities to the pooled national analysis, the results are presented below.

When the effect of charter schools on student learning is compared to the experience the students would have realized in their local traditional public schools, the result can be graphed in a point‐in‐time Quality Curve that relates the average math growth in each charter school to the performance their students would have realized in traditional public schools in their immediate community, as measured by the experience of their virtual twins. The Quality Curve displays the distribution of individual charter school performance relative to their TPS counterparts. A score of “0” means there is no difference between the charter school performance and that of their TPS comparison group. More positive values indicate increasingly better performance of charters relative to traditional public school effects and negative values indicate that charter school effects are worse than what was observed for the traditional public school effects.