Association for Public Policy Analysis & Management

Student performance has served as the basic metric for school evaluation for decades. But many “poorly performing” schools may actually be functioning extremely well, given the financial and demographic deficits they face.  This is a standard critique voiced by frustrated teachers, administrators and parents. My study builds a more complex measure of school performance, by considering the limits and possibilities facing different schools.  If two or more schools have identical student performance averages, the technique I am using allows an examination of which school is overcoming its deficits more effectively. This efficiency viewpoint for educational policy is valuable in a world of rising expectations, continued underperformance of schools, and budget constraints.  

I conduct a two-stage Data Envelopment Analysis (DEA) of the efficiency of Hawaii public high schools. Hawaii’s public school system is different from the rest of the States. It has only one statewide school district that operates prekindergarten up to grade 12. Hawaii Department of Education (HIDOE) is the official state education agency overseeing all 256 public schools and charter schools, 10,882 teachers and 169,987 students. The purpose of this study is threefold: (1) to introduce the Data Envelopment Analysis method as a primary measurement of educational efficiency to the Hawaii Department of Education; (2) to determine the relative efficiency of public schools in the State of Hawaii and provide school level policy recommendations for improvement and (3) to examine the causes and/or characteristics of school inefficiency in a statewide school district context. This study also employs the DEA Specification Robustness Check to draw robust conclusions about the public education system in the islands where geographic and demographic factors influence the performance and efficiency of schools. 

To the best knowledge of the author and based on literature review, there has been no DEA study carried out before to examine the efficiency performance of schools in Hawaii. DEA has been used in determining the efficiency of schools in various countries but relatively less common in the United States. Hawaii’s single school district and equitable per-pupil funding might make it hard to observe a big difference between simple performance metrics and those adjusted for disparities in inputs between schools. Since DEA takes into account the multiple inputs and outputs of each school, differences between schools can be easily observed. So if the study can be done here, with meaningful results, it will work even better in almost every other U.S. school system.