An increasing proportion of cases that come to the attention of state child support enforcement programs need paternity establishment before child support can be ordered. Strong fiscal incentives exist for states to establish paternity. Given dwindling resources, agencies need effective and efficient methods for increasing the number of children with paternity established in the state. This research asks: To what extent do a predictive-modeling based paternity establishment approach and enhanced outreach efforts affect changes in paternity establishment rates and field staff and hospital staff responses across a 20-month period? To what extent does Geographic Information Systems (GIS) mapping of pre- and post-interventions help stakeholders visualize data and make meaning of results?

This 2011-2014 collaboration between Iowa’s Child Support Recovery Unit (CSRU) and researchers at Iowa State University involved: baseline assessments; design, implementation and monitoring of innovative methods for both paternity establishment procedures and community outreach; and evaluation. First, historical paternity establishment data reported by Iowa hospitals and regional CSRU offices, and statewide demographics were mapped using GIS technology. Baseline surveys of CSRU field staff and hospital staff involved in paternity establishment were conducted electronically. Maps and quantitative survey results were used to identify areas for improvement at baseline.

Second, using regression modeling to analyze historical Title IV-D case characteristics, a target-setting model was developed to predict paternity establishment for current cases. This model was used to identify a viable pool of cases with children needing paternity establishment and to set targets equitably among field offices.

Third, impacts of the target-setting model and enhanced outreach to hospital staff and community organizations that served unmarried couples before and after the birth of their child have been assessed using follow-up surveys of field and hospital staff.  Four key expected outcomes are tracked over 20 months: 100% of field offices meeting paternity establishment targets; 10% increase in proportion of paternity affidavits completed in hospitals, 10% decrease in ratio of judicial to administrative paternities; and 5% increase in ratio of hospital affidavits to paternity orders. Final results are reported through charts and maps to visualize quantitative impacts of interventions. Overall, the modeling approach narrowed the viable pool of paternity establishment cases from 3,425 to 1,305, allowing more efficient use of field workers’ time. CSRU staff report a much better understanding of the target-setting process. Hospital staff survey results reinforce the need to concentrate outreach to unmarried couples prior to arriving at the hospital at the time of the birth.

The paper illustrates effective use of GIS technologies for understanding child support enforcement data and for highlighting regional differences, as well as the potential for transference of an innovative modeling approach to setting paternity establishment targets in other states.