Association for Public Policy Analysis & Management

Exclusionary discipline such as out-of-school suspension (OSS) and expulsion is associated with student outcomes such as lower academic achievement, grade retention, dropping out, and involvement in the juvenile justice system. Further, there is concern about the disparate impacts on students of color and special education students.

In response to these concerns, many school districts and states are moving away from exclusionary discipline toward less punitive consequences. Some of these policy changes explicitly prohibit the use of exclusionary discipline for certain types of infractions (e.g. willful defiance) or for certain groups of students (e.g. pre-K or elementary students), while others limit the length of suspensions, for example. Many of these changes are implemented without causal evidence that either a) the old system was failing or b) the new system will be better.

In one southern state, the state legislature passed a law prohibiting the use of out-of-school suspension as a consequence for truancy. While the law passed in 2013, there was not a full reduction in the use of out-of-school suspension as a consequence for truancy, even by the 2014-15 school year. In fact, while 14% of truancy cases in this state during the 2013-13 school year, this percent was as high as 10% in 2013-14 and 9% in 2014-15.

In this paper, I address three key research questions:

1. What are the characteristics of the schools that do and do not respond to the policy change by reducing and/or eliminating the use of OSS for truancy?
2. What is the impact of this policy change on school-level outcomes (e.g. truancy, disciplinary infractions, and the use of OSS in general)?
3. What is the effect of OSS on student outcomes, using the policy change as an instrument for whether a student received OSS for truancy?

To answer these questions, I use eight years of infraction level data for all K-12 students in public schools in a southern state (2008-09 to 2015-16). To address question one, I use ordinary least squares regression to descriptively assess what types of schools do or do not fully respond to the policy change. To address question two, I utilize interrupted time series, comparing the affected schools those that were previously using OSS for truancy) to otherwise similar, but unaffected schools. Finally, for question three, I restrict the analysis to students with reported truancy incidents and assume the policy change is an exogenous change that affects the likelihood of receiving OSS for truancy. Using this policy change as an instrument for receiving OSS, I estimate a causal impact of OSS on student outcomes (e.g. student test scores and grade retention).

This work seeks to address two key issues in education policy: how schools and school districts actually respond to state-level policy changes, and what impacts these policy changes have on the students in those schools. For example, given the limited response to the law, states must carefully consider the communication and capacity building that are necessary for achieving the desired results.