Association for Public Policy Analysis & Management

The BICS team worked with the California Division of Child Support Services in Sacramento and San Joaquin Counties to test whether a behavioral intervention would increase NCP engagement in the child support order establishment process. The main objective was to increase the number of NCPs who submitted Answer forms in response to service. By submitting an Answer form, NCPs ensure that a hearing is scheduled. In addition, this prompts child support agency staff to call parents to discuss the establishment process and answer any questions NCPs have.

Using insights from behavioral science, the BICS team designed an intervention that included a new cover page (an “explainer” sheet) that was put on top of the package of papers in the summons and moved the Answer form directly behind it as well as scripts that staff followed when speaking to NCPs who had been served and those who had filed an Answer; the scripts aimed to help NCPs understand their options, make a plan, and stay motivated. These components were delivered by specially trained BICS case managers who were funded by the grant. NCPs were randomly assigned to receive this new form of outreach or business as usual over an 8.5-month period.

The test measured whether the intervention could increase the percentage of NCPs who submitted an Answer form and decrease the percentage of orders established by default (by increasing orders established at a court hearing or by stipulation). The test also measured whether the intervention could improve collections on child support. The team also hypothesized that decreasing the number of orders established by default may increase the overall collections rate for the agencies and the children they serve given the correlation between default orders and lower payment rates.

Results showed that more NCPs in the BICS intervention group were served compared with NCPs in the control group. This impact was unanticipated and larger in San Joaquin (77.4 percent served in the control group vs. 85 percent in the intervention group) than Sacramento (65.6 percent vs. 69 percent). This unexpected “service effect” likely drove the statistically significant impacts on Answers filed (27.6 percent vs. 30.6 percent) and payments ($351 vs. $444 paid). There is some evidence to suggest that intervention components (rather than the service effect alone) may have contributed to the statistically significant impacts on stipulations (6.6 percent vs. 8.7 percent). There were no statistically significant impacts on orders established by default. The results varied by county and the language spoken by the NCP in ways that are discussed below.