ABSTRACT
Few evaluations have focused on the supply and demand within the education marketplace in a school choice environment. Because traditional public schools are not subject to the same level of competitive pressures as private schools, we expect that measures of school quality—enrollment, academic achievement, and safety—will be more likely to predict closures for private schools and public charter schools than Milwaukee Public Schools (MPS). We employ survival analysis using data from private, traditional public, and public charter schools in Milwaukee from 2005 to 2016. Data on enrollment trends, demographics, and academic performance from the Wisconsin Department of Public Instruction were combined with data from other sources on school safety and closure. Results from our models suggest: (a) enrollment losses drive school closure in all three sectors, (b) low academic achievement only predicts closure for private schools, (c) families choose schools based on academics in all three sectors, and (d) academics and school safety are positively correlated.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1. School choice programs. Wisconsin Department of Public Instruction. Retrieved from https://dpi.wi.gov/families-students/programs-initiatives/school-choice
2. Legislative Report on Charter Schools 2014–2015. Wisconsin Department of Public Instruction. Retrieved from https://dpi.wi.gov/sites/default/files/imce/sms/Charter-Schools/14-15%20Legislative%20Report%20.pdf
3. Independent (2r or 2x) Charter Schools. Wisconsin Department of Public Instruction. Retrieved from https://dpi.wi.gov/sms/charter-schools/independent
4. Wisconsin Charter School Quick Facts. Wisconsin Department of Public Instruction. Retrieved from https://dpi.wi.gov/sites/default/files/imce/sms/Charter-Schools/Wisconsin_Charter_Schools_Quick_Facts.pdf
5. This is consistent with the previous work of Flanders (Citation2017) who showed that instrumentality charter schools in Milwaukee have very few of the characteristics generally associated with charter schools throughout the country.
6. A Teachable Moment. Wisconsin Policy Forum. Retrieved from https://wispolicyforum.org/research/a-teachable-moment-understanding-the-complexities-of-charter-school-financing-in-milwaukee/
7. Technically speaking, the first voucher program in the United States was launched in Vermont in 1869, followed by Maine’s voucher program in 1873. However, these two town-tuitioning voucher programs are unique in that they allow children that live in rural areas without residentially assigned public schools to attend public schools in another town or nonreligious private schools of their choosing.
8. Children and Youth with Disabilities. National Center for Education Statistics. Retrieved from https://nces.ed.gov/programs/coe/indicator_cgg.asp
9. We standardized this measure because Milwaukee experienced changes in their math tests over the time period of the study. Nonetheless, the overall results are identical when the models use the nonstandardized measure of math proficiency instead.
10. Six percent of the observations were missing for proportion of White students, 12% were missing for the proportion of economically disadvantaged students, and about 21% were missing for the proportions of ELL and SWD students.
11. As shown in of the Appendix, all results hold for both sectors when the log of enrollment and the log of 911 calls are used as independent variables.
12. Different classification practices produce measurement error, which could bias coefficients towards zero and increase standard errors, thereby increasing the likelihood of Type II errors in interpreting results for the SPED control variable.