Tracking Teacher Effectiveness Gaps Between Charter Schools and Traditional Public Schools

Are teachers working in charter schools more effective in improving student outcomes compared to teachers working in traditional public schools?

This is an important question because charter schools have become the most popular alternative to traditional public schools in the United States over the past three decades, especially in urban school districts. And given the established role of teachers as the most important schooling factor that affects student outcomes, teacher quality is a likely factor in driving charter school effectiveness.

In a new AIR study, we examined the disparities in teacher effectiveness between charter schools and traditional public schools in Florida. To measure individual teachers’ contributions to their students’ test scores, our analysis used the official “value-added scores” calculated by the Florida Department of Education and used throughout the state’s teacher evaluation system.

Behind the Numbers

Our study covered all teachers in Florida with value-added scores in the 2011–12 and 2013–14 school years. Typically, these teachers instructed students in grades that are subject to statewide standardized testing (grades 3 through 10 in reading and grades 3 through 8 in math).

Value-added scores used in our analysis were generated by the Florida Department of Education using the state’s official value-added model. The Florida value-added model measures the effect of teachers on student learning by comparing test scores of their students with those of students in the same grade who are similar, along with prior test scores, subject-relevant courses enrolled, disability status, gifted status, English proficiency, attendance, student mobility across schools, age, and classroom characteristics (such as class size and homogeneity of entering test scores in the class).

These scores are generated by value-added models—statistical techniques currently used in teacher evaluations in many states and school districts. These models have been shown to isolate the contribution of a teacher’s impact on student achievement.

Here’s what we found:

  • Teachers working in charter schools with a higher-than-average proportion of students eligible for free or reduced-priced lunches (which we call high-poverty schools) were more effective than traditional public school teachers working in similar settings. Specifically, the average teacher in high-poverty charter schools in Florida raised student test scores by 3 percent of the standard deviation (equivalent to roughly three weeks of instruction) more than the average teacher in high-poverty, traditional public schools. We found no such difference in lower-than-average poverty settings.
  • The most effective teachers in high-poverty charter schools were significantly more effective than the most effective, traditional public school teachers in similar settings. To put it another way, suppose we have 100 teachers in high-poverty charter schools and 100 in high-poverty, traditional public schools. Our findings suggest that the 10th most effective teacher in high-poverty charter schools raised test scores by 5 percent of the standard deviation (about five weeks of instruction) more than the teacher with the same ranking in the high-poverty, traditional public school. No such difference existed among the least effective teachers in the two sectors.

Are charter teachers in high-poverty schools more effective because they are more experienced or better educated? We examined this question using teacher-level information from Miami-Dade County Public Schools, which houses more than 20 percent of all charter schools in Florida. We found no evidence that cross-sector differences in teacher effectiveness are driven by differences in teacher characteristics.

Instead, we found that early-career teacher effectiveness increased significantly with experience in high-poverty charter schools but not in traditional public schools.

So, why did teacher effectiveness increase with experience in high-poverty charter schools but not in traditional public schools? While the definitive answer was beyond the scope of our study, we explored several mechanisms. One possibility is that charter schools are better at filtering out less effective early-career teachers as a result of less restrictive personnel policies.

Among the teachers we observed in our sample (teachers with value-added scores), we found that those who left high-poverty charter schools were less effective than those who left high-poverty, traditional public schools, providing evidence to support this possibility. Our analysis of Florida educators in the national Schools and Staffing Survey (SASS) also points in this direction: 69 percent of traditional public school principals in Florida (versus 13 percent in charters) reported personnel policies as a barrier to dismissal of poorly performing teachers.

The SASS survey also showed that traditional public school teachers in Florida were significantly more likely than charter teachers to report low parental involvement (62 percent in traditional public schools versus 39 percent in charters) and student unpreparedness (66 percent versus 44 percent) as problems at their schools. Higher rates of charter teachers also reported that they have influence over the content of their classes (56 percent in charters versus 45 percent in traditional public schools), the materials in their classrooms (52 percent versus 35 percent), and their teaching techniques (95 percent versus 82 percent). These findings point to significant cross-sector differences in working conditions, which could explain why early-career teachers improve faster in high-poverty charter schools.

What are the implications for public policy?

States can’t require more parental involvement in traditional public schools, but policymakers could consider ways to give principals more tools to deal with less effective teachers and to give teachers more support overall and more autonomy in their classrooms to help improve classroom conditions, increase teacher motivation, and, in the end, improve student outcomes.

Umut Özek is a principal researcher at AIR affiliated with the National Center for Analysis of Longitudinal Data in Education Research (CALDER). His focus is on the economics of education and public economics.