The Michigan Education Achievement Authority (EAA) is a new public school authority currently comprised of 15 schools in Detroit. The EAA is designed to improve academic performance by instituting new financial and administrative management of schools. The creation of a public school system is an untried education initiative for the state of Michigan. This reform method is based on the Louisiana Recovery School District (2003) and was developed around the same time as the Tennessee Achievement School District (2012). All three versions of the statewide public school district have the same vision of turning around underperforming schools to create more opportunity for students. The EAA has the ability to run schools themselves (direct-run) or to delegate authority to charter schools. At this point, the EAA has not been in operation long enough to make a definitive judgment about the program’s effectiveness. However, by examining the outcomes of the Louisiana Recovery School District program, I hope to make conjectures regarding the likelihood of success of the Michigan program. Reform policies in education have high costs, therefore, predicting the success of the EAA by looking at the results of the Louisiana Recovery School District is responsible, prudent, and possibly enlightening. Since the EAA has the potential to expand in the coming years and is in competition with other school districts for state funding, it is critical to monitor this new organization for effectiveness..
My main question for this study is if there is a significant performance difference between the two different treatments: direct-run and charter schools. I will also measure the treatment effect of the entire RSD program without separating out direct-run and charter schools. The data are from schools in the program and surrounding comparison schools from New Orleans and Baton Rouge. The dependent variable is the Student Performance Score (SPS). This is a measure that Louisiana uses to score each school on a yearly basis and standardized test scores are the main determinant. Graph 1 represents the distribution of the data for SPS across 2004, 2010, and 2012. The data for 2004 capture the SPS before the RSD program, and 2010 and 2012 are the SPS after the program. I use both OLS and fixed effects regression to examine the variation in the SPS while accounting for enrollment, percentage of students in special education, percentage of students on free/reduced lunch, year dummy variables, and dummy variables for the program and the two different treatment options.
Based on the data collected from the Louisiana Recovery School District, creating a statewide public school district to serve underperforming schools is successful in improving academic outcomes. As a whole, the schools in the program increased their SPS by a standard deviation after the program intervention. When comparing the two treatment options, the charter schools in the Recovery School District performed higher (on average by 16 SPS points) than the schools run directly by the state. The goal of this study is to make a recommendation for the new Michigan Education Achievement Authority. Therefore, I created projections of the SPS for each of the Detroit schools, shown in Graph 2. There is a clear trend reflected in the projected scores for the Detroit schools. In each school, the score is higher if that school happened to be given over to a charter school rather than a direct-run school. However, the score projections cannot be made with certainty in most schools, so there is not a definitive statistical claim that charter schools would be preferable to direct-run schools in Detroit. The only school where the projection can be confirmed with certainty is Phoenix Elementary-Middle School. The main difference between Phoenix and the other schools is the percentage of students in the Free and Reduced lunch program. The average percentage for the Detroit schools is 20 percent, but Phoenix’s percentage is 68 percent. As poverty in a school increases, so does the certainty that a charter school is the better intervention.
There are four main limitations both the findings and the projections. The first is that it is difficult to generalize from Louisiana to Michigan. The New Orleans and Baton Rouge schools were deeply impacted by Hurricane Katrina. The Detroit schools have been continuously affected by the financial woes of the city of Detroit. The unique experiences of the cities definitely influence the interventions in public education. Specifically, the damage caused by Hurricane Katrina could have left the state unable to rebuild many schools, which could explain to prevalence of charter schools in the program. Also, the installment of a financial manager for Detroit Public Schools could be the reason why Detroit is the only area where the EAA currently operates and would favor continuing the large role of the state in public education. The precedence of state intervention in Detroit Public Schools could have made it easier to enter into the interlocal agreement that removed fifteen schools from the district. The second issue is being able to make a meaningful comparison across charter schools. Charter schools in Louisiana and Michigan are run by separate companies with varying teaching philosophies. Where the state run schools have one unified vision, the charter schools are separate entities. Therefore, when claiming that charter schools improve student performance more than the state run schools it is very difficult to pinpoint exactly why or how. The third issue is the reliability of the reported data. Each school has the discretion to report the test scores and even the percentage of students on Free and Reduced Lunch within the school. This brings up a serious question on if the data collected can be trusted and taken as a reflection of reality. The fourth issue is that the sample size for charter schools is much larger than the direct-run schools. Therefore the validity of the results for the direct-run schools is questionable. Each of these issues must be investigated and considered to increase the reliability of the results and projections.
Since the Michigan program is based off of the Louisiana program, it is the best indication of the direction that the EAA should take. Due to recent legislative actions, the Michigan EAA is due to expand to other parts of the state. Since additional resources will be expended towards this program, it is prudent to examine studies such as these to best serve the students. From the outcome of the Louisiana Recovery School District, it seems that the Michigan Education Achievement Authority should pursue a strategy of granting charters rather than running the schools themselves. Since the current number of direct-run schools in Detroit is twelve, while there are only three charter schools, the EAA is going in the opposite direction. As the Louisiana RSD expanded, it favored charter schools over direct-run schools. With the high number of charter schools already operating in the city of Detroit and with the lifting of the cap on charter schools in the state of Michigan, the EAA would be free to expand the number of charter schools authorized. It will be essential to keep an eye on the growth and outcomes of this program in order to ensure the best intervention for the students of Michigan.
More information is available in this poster: