If you know me, you know that I love to travel. When I’m planning a trip, I use multiple sources of data and analytics for my decision making. I might consult TripAdvisor reviews to check out lodging or activities. I can look at restaurant ratings, reviews, and even filter by type of cuisine or meal. I also analyze historical flight price trends to help determine the best time to purchase tickets. I want my travel experience to be the best it can be, so I take advantage of the available data created by others.
We use statistics and analytics every day to make decisions in our lives.
 |
| Visualization example: Michigan child support cases by assistance type. |
But of course, the statistics we use in the work we do for the children and families in Michigan is far more important than where we stay on a trip. That said, we must continue to improve our access to the information we need to make decisions in the child support program.
The data needed to make everyday life decisions has been available for decades, but it’s now increasingly easier to access. In the past, when planning a trip, I might have asked friends or read travel guides about restaurants or reviewed box scores for past player performance. Now, I can quickly access the most relevant information I need to make those decisions from the palm of my hand. This is the direction we must continue in to “ensure staff have the data tools and information necessary to be successful” (quoting the Michigan Child Support Program Strategic Plan 2018-2020). When we are successful, we help families achieve well-being and self-sufficiency.
IV-D offices are already doing great work. We are requesting ad hoc queries or running reports in Business Objects, analyzing the data, and producing insights and work lists to improve performance. With the release of Self-Service Reporting in May 2019, we can now create our own reports — giving us even faster access to the data we need.
You may not know this, but OCS has a team dedicated to performance management in the child support program: the Planning, Evaluation, and Analysis (PEA) Section (my team). We work across the program to analyze data and communicate the information and insights to assist our IV-D partners in making decisions. We assist teams with measuring the effectiveness of system or process changes. We are responsible for compiling and submitting the OCSE Annual Data Report and monitoring our performance on the federal incentive factors. With help from our partners, we conduct the Annual Self-Assessment Audit, as required by OCSE. We’re also exploring new ways to use our data, such as predictive analytics and data visualization techniques.
During the fall of 2018, we completed a predictive analytics project with our partners from Accenture. The objectives were to reveal case and demographic attributes that lead to delays in establishing an order and to reveal actions that would move a case through the establishment process. We identified two delay points: court action referrals (better known as CARs) and service of process. We built two predictive models that identify the cases that have the highest risk of delay during the establishment process at those delay points. The models also identify which pieces of information could be added to the case to reduce the risk of being delayed. Communicating these insights will help staff by identifying which data elements are needed to reduce the chance of a delay in the establishment process.
 |
| A predictive model for cases with highest risk of an establishment delay. |
A fully staffed PEA team, including two new positions (a statistician and a communications professional), allows us to expand our services and products. The team is also supported by system staff who are dedicated to reporting and analytics.
While the PEA team tracks our strategic plan and measures the progress made towards each strategic goal, it demonstrates the importance of the last goal area: improve data tools and implement technology. The first three goal areas are about improving: child support processes; the customer experience; and education and outreach. In order to responsibly improve in these areas, we need to make data-driven decisions. Using statistics and analytics to measure the effectiveness of our current efforts allows us to identify changes that need to be made as well as predict results from future efforts.
The child support program’s use of predictive analytics and data visualizations are the next steps on the path, and we must remain focused on improving accessibility to our data. Jeff Weiner, CEO of LinkedIn, summed it up perfectly, “Data really powers everything that we do.”
