Description and Assessment of State Approaches to Diversion Programs and Activities Under Welfare Reform. C. What States Know About The Effects of Their Diversion Programs


Our conversations with state officials included questions concerning the kinds of data collection efforts associated with the formal diversion programs that are being undertaken, and whether the states have any empirical indicators of the success of their diversion programs and activities. Although these conversations primarily focused on lump sum payment programs as the most structurally formal of these diversion activities, their comments were sufficient to provide a general sense about data collection and data issues.


Most States Have Only Recently Implemented Diversion Programs

As discussed above in Chapters Two, Three, and Four, most states have only recently implemented their formal diversion programs. For example, of the twenty states with lump sum payment programs, sixteen states have implemented their programs within the last four to sixteen months. Similarly, most mandatory applicant job search programs have been gradually implemented since the passage of TANF in 1996. Perhaps because of this, most state officials were not able to report many details about their data collection or management information systems designed to facilitate data collection and analysis of the effects of their diversion programs. Notwithstanding the lack of detail and/or capacity with respect to state data systems, these officials were certainly aware of the need for data collection to: 1) determine whether lump sum diversion program are successful in keeping families off the TANF rolls, 2) measure the effects of other formal diversion programs, and 3) monitor how well these programs and activities are being implemented.

Data Needed to Assess Effects of Diversion Programs

A common set of data elements necessary to document and analyze the effects of diversion programs can be identified. These elements include size of the diverted household including the number of children, general demographic and socioeconomic data, the reasons for eligibility for diversion, the type of diversion assistance provided, what was provided in alternative resources, how much was provided in a lump sum payment, what the lump sum payment was used for, method of lump sum payment (cash or vendor), whether the diverted applicants were receiving other types of assistance, what kinds of employment were secured, with what assistance and how quickly, and how long these families remained off TANF. Data regarding the state's Medicaid program must also be collected including: eligibility criteria,

number of diverted applicants receiving Medicaid, number of joint Medicaid/TANF and Medicaid-only applications, and trends in the Medicaid rolls.

A Wide Range of Data Collection Efforts

State officials described a wide range of data collection efforts that varied from relatively minimal to relatively comprehensive. Where data systems do exist they are most likely to capture information regarding lump sum payments but not necessarily alternative resources or mandatory applicant job search. For example, both Minnesota and Washington report having data management systems in place capable of collecting all of the data elements related to lump sum payments listed in the previous paragraph. Both states will use these data to monitor closely the implementation of their programs and to initiate improvements. Both states are also under a mandate from their legislatures to report to them in 1999 and provide empirical evidence about how their lump sum payment diversion programs are working. On the other hand, Kentucky and South Dakota report somewhat less comprehesive approaches to data collection. South Dakota collects data such as household size, reasons for diversion, the amount pf the lump sum payment and for what purpose, and whether employment was secured. Kentucky collects enough information to monitor the 12-month period of ineligibility but after that period, the records of diverted families may or may not remain in the system. In both of these states, as in many others, it is not clear whether the TANF application of a diverted individual is ever entered into the data system or whether the application is consistently considered denied, incomplete, withdrawn, or "other."

Potential Data Issues and Problems

Several data issues and potential problems became apparent as we reviewed our results. The first involves whether and how the TANF applications of diverted families and individuals are documented and kept in a data system. Recording the existence of a diverted TANF applicant is an essential starting point for data collection efforts sufficient to support analyses of diversion efforts. For numerous states, there was little knowledge or clarity at the administrative level about how these TANF applications are processed. Even if the application is documented, there are different approaches to characterizing the application with implications for how the effects of diversion might be understood, e.g., an application could be categorized as incomplete, denied, complete, or applicant never returned. Another issue concerns how consistently counties within a state are documenting and keeping track of diverted applicants. To the extent that counties are inconsistent in their approaches to documenting diversion efforts, statewide analysis is not possible.

Another issue involves how long a diverted person is kept in the state data system. Many states report removing diverted individuals from their systems as soon as the period of ineligibility expires, which could be as short as three or four months. Such an approach would make it impossible to follow diverted families to determine if they ever returned for TANF assistance. In a similar vein, South Dakota's approach to administering its lump sum payment program presents a potential problem that could occur in other states as the emphasis on work is increased. In this state, the TANF applicant usually must interact with both a DHS caseworker and a Department of Labor (DOL) specialist who could each document the diversion events differently or not at all although both workers share the same computer system in terms of data entry.

Measures of Success for Diversion Programs - Too Early To Tell

For many states, the measures of success for diversion programs appear to include either how many applicants were diverted as a percent of total applications and/or how long applicants stay off the TANF rolls, or whether they ever returned for assistance. In most states, however, there are few data available regarding success. The reasons for this lack of data include: 1) states have not yet begun to collect data, 2) state do not have the capacity to collect, or 3) states have only recently implemented their diversion programs.

For example, both West Virginia and Idaho began implementing their lump sum payment programs in mid-1997 and have seen too few cases to assess success. The Kentucky official reported "having a sense" that the state=s lump sum payment program was working well and that diverted families were more likely to request a series of payments as opposed to one payment; there are currently no systematic efforts to collect data. On the other hand, the South Dakota official expressed pessimism about how frequently that state=s lump sum payment program would be used since this program is seen exclusively as job-support assistance and job prospects are bleak in South Dakota.

Two states, Utah and Virginia, are able to document the success of their lump sum payment diversion programs - it is not a coincidence that these states have the longest-running programs, established in 1993 and 1995 respectively.

Utah has established a strong management information system that supports the tracking of diverted applicants. Utah estimates that 3,000 persons been diverted since January 1993 and that this number represents between 15 and 40 percent of TANF applicants depending upon the site. Utah=s program has been examined and the results suggest that approximately 85 percent of diverted persons do not ever reapply for TANF assistance. Virginia has also established a strong management information. This system supports the collection of data on the number of diverted people by county and the analysis of data to determine how many diverted persons return for TANF assistance and how long after receiving diversion assistance. The state reports that during the first two years of the program 1,004 families received lump sum cash payments and 81 percent of those cases have not reapplied for TANF assistance. The greatest use of lump sum payments in Virginia has been for payment of rent.

Montana implemented its Job Support Program (JSP) in February, 1996 and, as the third oldest lump sum diversion program, collects data on persons participating in this program. Information is collected on household size, income, and other characteristics of diverted applicants. The state is also able distinguish between lump sum payments and alternative resources diversion by virtue of how the data are entered and organized in Montana's computer system.

Most of the states have little empirical evidence about how their formal diversion program are working. While most states have only recently implemented their programs, inadequate data systems also contribute to a lack of empirical evidence. Consequently, it not possible to draw any conclusions about the effectiveness, or effects, of the states' formal diversion programs and activities.