TANF agencies generally track basic caseload trends over time, such as total families or families of certain types.
We have all heard the refrain "one size does not fit all" and it certainly applies to public policy: matching the right services to the right populations is not always easy. TANF agencies collect a wealth of data that can be used to better target services. This post talks about some simple approaches that agencies can use, based on what researchers have been using to create TANF subgroups for many years. You can use this approach to target services to your customers based on their characteristics.
A key principle in segmenting your TANF caseload is to find risk indicators which correlate with key outcomes such as employment and long-term benefits use. After many years of experience, MDRC researchers have found three indicators to be predictive and operationally feasible because most TANF agencies collect the data. The three risk indicators are:
- A customer does not have a high school diploma
- A customer has not worked in the past year
- A customer has been on welfare for at least two years previously.
The most disadvantaged are usually defined as individuals with all three of these risk indicators. The least disadvantaged are defined as individuals with none of these risk indicators. Likewise, individuals are considered moderately disadvantaged if they had one or two risk indicators. These indicators have been useful in several programs and can be modified as needed based on your caseload.
This post just scratches the surface. If you want to see an example of how researchers have used this kind of analysis in the past, check out this report which examined subgroups across 20 welfare-to-work programs. There are also much more advanced ways to create subgroups and indices, but we've found these 'tried and true' approaches to be very effective and can work with you to help implement them in your office if you become a TDC pilot agency. Depending on interest, future posts will cover regression-based subgroups and how to use analysis to find individuals who are at the tipping point between good and bad employment trajectories. Please send your comments/suggestions to email@example.com.