This site uses cookies to provide you with a more responsive and personalized service. By using this site you agree to our use of cookies. Please read our cookie notice for more information on the cookies we use and how to delete or block them.

Bookmark Email Print this page

Advanced Analytics for Child Support Programs

From reactive enforcement to proactive prevention — Part II


DOWNLOAD  

In Part I of this article, we provided some background on the rise of analytics, discussed the budgetary challenges facing states and talked about the internal and external data required to build a powerful predictive model. In this article we will continue the journey by discussing the basics of child support predictive modeling and how analytics can be leveraged in child support enforcement. We will also outline ways advanced analytics can play a critical role in helping child support agencies enhance the collections process, improve revenue collection, increase the likelihood of meeting performance-based incentive funding, proactively identify the non-custodial parents (NCPs) most likely to go into arrears in the future and help mitigate that arrearage accrual.

Child support modeling basics

Predictive analytics can present a powerful opportunity for child support service agencies to accomplish more with smaller budgets and limited resources. A classic example of child support predictive modeling is to combine internal NCP/case characteristics with external third party data to help calculate a numeric score that segments NCPs along such predicted “target variables” as the likelihood of an NCP’s beginning to pay court mandated child support, the likelihood of an NCP’s becoming in arrears at some point in the future, or the likelihood that 80% or more of the NCP’s accrued amount will be paid in the coming three months. Which target variable (or variables) to estimate and how, are model design issues that experienced statisticians and child support domain experienced practitioners — the architects building on the foundation of good data — work out in advance in consultation with the client.

Leveraging advanced analytics in CSE

Albert Einstein once wrote that “the whole of science is a refinement of everyday thinking.” This is certainly true of predictive analytics. When people make decisions, they typically use their prior experience and domain knowledge to combine various pieces of information to help make predictions or estimate unknown quantities as effectively as they can. In other words, they informally “build predictive models” in their heads.

Bringing the models to life

Ultimately, bringing the predictive model to life typically requires the buy-in of state child support agency leadership and the front line case workers at the forefront of the child support efforts. 

Related links

Share this page

Email this Send to LinkedIn Send to Facebook Tweet this More sharing options

Stay connected