Can Court Diversion Improve School Attendance among Elementary Students? Evidence from Five School Districts

Abstract

We examined the effect on attendance of a truancy court-diversion program for elementary students. Truancy court-diversion programs represent a shift from a law-and-order approach toward a public health model to address school absenteeism. Instead of directly referring parents of truant elementary students to child protection services or juvenile court, most court-diversion programs educate parents about the importance of school attendance and create an attendance contract that includes social service referrals. Despite being widely implemented, truancy court-diversion models have not been rigorously evaluated. Using 10 years of administrative data from multiple state and local agencies from Minnesota, we constructed a counterfactual of students from schools that did not implement truancy court-diversion. We used difference-in-differences methods to test whether parent education and attendance contracts improved attendance relative to direct referral to child protection services or court. We failed to reject the null hypothesis of no program effect. Most truancy program evaluations use a pre- and post-mean comparison, but our analysis suggested that the magnitude of the bias in such studies is substantial because absenteeism showed a pattern of regression to the mean.

Keywords:

• Educational neglect
• chronic absenteeism
• truancy
• matched sampling
• difference-in-differences

Notes

1 The results from pre-, post-mean different t-tests are also presented in tables in Appendix.

2 We used the following model to estimate the native OLS estimates: $Y_{it}=\alpha +\beta {\mathit{Post}}_{it}+{\mathit{X}}_{\mathit{it}}\mathit{\delta }+{\epsilon }_{it}$ where student i was in the set of students in the program referred group, Post = 0,1 indicates pre-, post-referral period. Similar to the sample described in the earlier section, pre-referral period included year and year prior to the referral, and post-referral period included the 4 years after the referral. Similar to Equation (1)$$\mathit{Bias}=E\left(y|X,d=1\right)-E\left(y|X,d=0\right)=B_1+B_2+B_3$$, ${\mathit{X}}_{\mathit{it}}$ includes the vector of time-varying student characteristics including free-lunch eligibility, disability status, child welfare involvement, and number of school transfers.

Additional information

Funding

This work was supported by Award No. 2014-IJ-CX-0010, awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. The opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect those of the Department of Justice.