Association for Public Policy Analysis & Management

Newark now has one of the most highly concentrated urban charter school sectors in the nation. Since 2003, the charter sector has grown from 10 schools serving about 5 percent of the city’s students to 43 schools accounting for about 30 percent of the city’s enrollment. While the effects of charter schools appears to vary dramatically by locality, prior research using a matched comparison group model has found that Newark’s charter sector is among the most effective in the nation at improving student academic performance (CREDO 2015).

An interesting feature of Newark’s charter sector that is not found in many other localities is that since 2014 the district has centrally assigned students to both charter and traditional public schools according to a unified enrollment process based on a deferred acceptance algorithm. Each spring, parents send to the central office a form ranking in order up to nine schools in which they would like to enroll their child the following year. The matching procedure is designed in such a way that the optimal strategy for the parent is to list their true preference order for each school.

The purpose of this study is to describe parental preference for charter schooling within a historically underperforming urban public school system with a rapidly expanding charter school sector. The analysis will use student-level administrative data with information about student demographic characteristics, residential information, test scores, enrollment, and the rank order preference for schools from 2014 through 2017. The basic approach of the analysis is to estimate OLS regressions relating observed student and school characteristics to the rank-order preference that students state for charter schools in the city.

The results of the analysis will provide important and policy-relevant information about the demand for charter schooling relative to traditional public schools within an expanding urban school choice environment. Further, the analysis will also discover the extent to which student attributes are associated with a preference for charter schooling. For instance, the results from analyzing stated parental preferences will provide information about the extent to which charter schools enrolling a smaller proportion of students with disabilities or students who speak English as a second language is due to parental preferences for schooling options.

Finally, I plan to supplement the data with information about the type of educational environments offered in each charter school acquired from the National Alliance for Public Charter Schools. This analysis will provide information on whether parents of children with particular characteristics are more likely to state a preference for schools with particular attributes. For instance, the results will provide information about whether there is a relationship between a student’s academic achievement and the probability that the student lists a preference to attend a school that follows the No Excuses approach to education, or other schools that provide very different educational environments.