Association for Public Policy Analysis & Management

Research suggests that mobile students should be of special concern because of the generally negative effects that changing schools has on their learning and achievement. Mobile students are disproportionately poor and minority and tend to have lower average academic achievement and decreased probability of graduating from high school. Two recent meta-analyses calculated that changing schools during the elementary grades costs mobile students the equivalent of 3-4 months of learning in both reading and math.

These disadvantages come on top of the fact that these students have basically been left out of federal education policy. Under No Child Left Behind (NCLB), students are required to be enrolled in a school for a “full academic year” (FAY) in order to count on that school's accountability rating. Most states require students to be enrolled continuously in a school from a specific date in the fall (referred to as the “snapshot date”) through the date of testing to be included on a school’s rating. This means that mobile students who move after the snapshot date do not count in a school’s rating.  Schools’ lack of accountability for mobile students may have exacerbated these students’ disadvantage, as schools may have focused resources on other students who did count for their rating. Research suggests that due to the sanctions attached to failing to meet NCLB’s requirements, schools react to the incentive structure of NCLB by allocating resources where they will most affect their ratings.

Because students who enroll after the snapshot date do not count for the school’s rating, schools facing accountability pressure may shift resources to students who count and away from those who do not. The use of a specific snapshot date in determining who counts makes regression discontinuity (RD) an ideal method for investigating schools’ reactions to mobile students. This is strict RD because treatment (counting towards accountability) is based entirely on enrollment date and there are no cross-overs (students who enrolled before the cutoff date who did not count for their school or students who enrolled after the date and did count for their school).

To examine the effects of enrolling after the cutoff date, I will use test scores as a proxy for schools’ resource allocation to students and compare the average test score for students who enrolled in schools just before the snapshot date with students who enrolled just after the snapshot date. The dataset I will use for the analysis is an administrative dataset containing all 6^th-12^th graders in one large urban district in a Southern state. The data includes state test scores, student enrollment and withdrawal information, and student demographics.

Knowing if schools react strategically to the incentives of the snapshot date is relevant as Congress currently debates the reauthorization of NCLB. Certain subsets of mobile students (e.g., migrant and homeless students) are given special protection and funding under federal policy, although that doesn’t extend to all mobile students. The results of this analysis could help inform policymakers as they determine the future of federal education policy.