(WITHDRAWN) Impacts of Training Mandated Reporters on Children's Child Welfare Involvement

Background/Purpose:

This study estimates impacts of a 2015 California law requiring county offices of education and school districts to annually train all staff on reporting child abuse. Previously, the state encouraged schools to provide training, without specifying content or frequency. The new law was passed following high-profile cases where staff knew of child abuse but did not always report it.

Implementing this law has required significant resources. For a sense of scale and intensity, the state developed online training modules that staff could use at no cost to counties and districts. The modules are currently 7 hours for first-time users and 3 hours each subsequent year. As of 2019, over 200,000 staff had completed the online training. This is the first study to estimate impacts of training mandated reporters.

Method:

The challenge for estimating impacts is identifying a plausible comparison group when the policy was implemented state-wide. The study addresses this by focusing on young children with different exposure to school staff due to their age. The data comprises counts of children involved with counties’ child welfare agencies, by age and reporter type (California Child Welfare Indicators Project at UC Berkeley). A difference-in-differences strategy is used to compare county-level outcomes for children who were reported by educational staff at ages 3-4 (less exposure) or ages 5-6 (more exposure), before and after the policy. This rigorous matched comparison group design accounts for county-specific trends in reports that would influence children ages 3-4 and ages 5-6 similarly. Additional county-level education data are used to account for other factors that could influence child abuse reports differently by age group (California Department of Education DataQuest).

Results:

The study finds that annual training led to an average of 13 additional 5-6 year olds being reported by school staff as subjects of suspected child abuse to their county’s child welfare agency per year. Results are robust to inclusion of county-time-varying covariates and fixed effects for county-year, county-age, and age-year. For 5-6 year olds, impacts will be discussed for child abuse reports, number of investigations conducted, cases opened, and racial disparities in each of these outcomes. Pre-post findings will be presented for other children.

Conclusions/Implications:

Annually training education staff did not increase child abuse reports for children as anticipated by the law’s sponsors, who suggested, “The absence of training is a failure of our system that leaves millions of students at risk every single day.” The study findings lead to three main implications. For child welfare agencies, these estimates could be a benchmark for projecting how other efforts to increase child abuse awareness and reporting will influence operations. For policymakers, identifying and targeting sources of under-reporting with administrative data could be more cost-effective than repeated universal training. That could also be less likely to cause over-reporting, which adds strain to child welfare agencies’ limited resources and can be traumatic to the reporter, child, and family. For researchers, the study highlights how existing administrative data can be used to identify impacts of policies to improve child welfare.