Predictive Analytics in Action Centre County, PA
Current challenges for child support case management Working the Right Cases Information Overload Visible Results Next Appropriate Action Case Worker Management Lack of a targeted case management strategy
Predictive Modeling Child Support Strategy Child support enforcement has traditionally been a reactive process Shift focus to predicting which NCPs are most likely to fail to pay in the near future Anticipated outcomes: o o o o Prevent arrears Decrease custodial parent complaints Take the right action on a case at the right time Ability to assign the right case workers to the right cases Use analytics to make smarter decisions, do more with less, and improve people s lives.
Pennsylvania s Solution Pennsylvania s Predictive Analytics solution included updates to the existing Performance Improvement Module (PIM) solution and a new application to calculate a predictive analytics score. Solution Payment Score Calculator (PSC) Description Allows a worker to calculate a Score for a case based on 20 variables Score is the likelihood of the defendant to pay 80% towards current support obligations in the next 3 months Information required to generate the Score is easily accessed from the system at the time support order is created or modified Provides a list of upcoming child support establishment conferences with cases assigned to the worker to allow for Score creation prior to conference
Payment Score Calculator Objectives Proactive Case Management Establish better payment pattern More effective NCP meetings Provide new customer outreach methods Identify opportunities for proactive enforcement activities Increase collections Increased Efficiency Maximize performance metrics Minimize costs related to enforcement Improve cost / effectiveness ratios Process Improvement More effective case assignment Establish consistent payment patterns based on tailored successful business process actions for specific scores Provide the right services/right time to encourage compliance with the order Let the NCP know the case is being monitored
Sample Variables Defendant Net Income Active Income Attachment Number of NCP Addresses Number of NCP Employers Distance Between CP and NCP The Payment Scores
Order Establishment Order Establishment Create WSR in conjunction with low financial obligation Sample Business Processes Score 2 Automated Follow-up Early Intervention Letter/ email (15 days) Immediately schedule follow up contact every 2-3 months (enforcement event or work search review with appropriate worker) Missed Payment e-reminder (30 days) Follow-up phone call Alert delivered based on frequency of NFOB Provide statistics of order amounts entered for category 2 payors Manual Follow-up Set temporary order based on guideline calculation with review in 2 months (with review of supervisor.) Review order annually rather than every 3 years. Automated Follow-up Early Intervention Letter/ email (15 days) Schedule review conference 2 months from order issue date Missed Payment e-reminder (30 days) Review order annually rather than every 3 years Manual Follow-up
Implementation Challenges Effective Change Management People Readiness Shift in Worker s Mindset DRS Buy-In Business Process Changes Organizational Readiness Communicating Strategic Intent of PSC SME support for counties More touch-points with users Site visits pre and post implementation Increased presence to discuss business process changes Regular follow-up activities ( Onsite and WebEx meetings, conference calls, etc.)
Outreach Methods WebEx Meetings User Documentation for Self Study SME Support Pre and Post Implementation Instructor led Outreach: Regional Workshops Impact of Payment Score Calculator Post Implementation Plan and Options
Confidence in the Score Payment Score Average Percent of Current Support 1 51% 2 66% 3 77% 4 87% To validate the accuracy of the payment score, a study of 5,000 PACSES cases was completed. For each payment score, the average FYTD percent of current support collected was calculated. A direct relationship between payment score and percent of current paid was found (i.e., the higher the payment score, the higher the percent of current paid).
Crawford County Divided 430 cases between two caseworkers Caseworker A made phone calls to defendant when no payment was made in the month Caseworker B scheduled an enforcement conference for the defendant to come into the office asap
Crawford County 430 cases fell into below payment scores Payment Score 1-67 Payment Score 2-53 Payment Score 3-177 Payment Score 4-133
Crawford County Average Current Support Collected
Lancaster County Divided 123 cases that had a score of 1 or 2 into a control and experimental group Contacted all defendants in the experimental groups
Lancaster County Experimental group 1 & 2 saw an 8% increase in number of cases receiving a payment
Questions? Need More Information? Feel free to contact us. Ann Marie Oldani Director of the Centre County Domestic Relations Section AnnmarieOldani@PACSES.com