The scope of the cost analysis includes all services provided by or arranged by the demonstration programs, regardless of how they were paid for. Services are included whether they were actually paid for by the program out of the demonstration budget ("program-paid costs") or were provided by other agencies or schools out of their own resources, without charge to the demonstration ("in-kind costs"), to yield the full "resource cost" of the demonstration programs. Costs are thus estimated for services provided directly by demonstration project staff, for payments made to cover child care and other participant expenses, and for services provided to participants by staff of other agencies, either by "loaned" staff at the demonstration site offices or at their own facilities. The cost of public school programs for participants who remained in high school, however, is not included, since we assumed that most high schools have excess capacity and thus the marginal cost of additional time spent in school was equal to zero.
The analysis focuses on the cost of services to participants in the period from July 1988 through June 1989, which is a period approximately in the middle of the overall demonstration term. This period was chosen to avoid the distortions associated with the start-up or close-down stages of the demonstration, and to focus on a period in which the demonstration most closely resembled a fully developed, ongoing program. In the initial and final stages of the demonstration, cost estimates could be distorted if the array of services being provided was incomplete, or if the flow of participants into the program was not yet -- or no longer -- at the level for which the scale of the demonstration staffs had been established.
To estimate the cost of demonstration services we drew on several major sources of data:
- State Demonstration Financial Reports. The starting point was the financial reports on demonstration expenditures issued by the New Jersey Department of Human Services and the Illinois Department of Public Aid. These reports captured the costs of
- Staff working directly for the demonstration program -- site managers, supervisors, case managers and specialists, and clerical staff;
- Services provided by other agencies under explicit contracts with the State or local agency running the demonstration;
- Support payments made for child care, transportation, and other training-related expenses;(5) and
- Materials, supplies, and equipment used at the sites.
- Interviews with Other Agencies. Large numbers of participants received services from other agencies that were not paid for by the demonstration and thus not accounted for in demonstration expenditure reports. These services included job training through the local Job Training Partnership Act service-delivery agencies, General Educational Development (GED) or college courses through community colleges and other institutions, and a variety of workshop and placement services provided at the demonstration offices by staff on loan from other agencies. We consulted at length with demonstration managers and representatives of these other agencies to obtain data needed to estimate the cost of these services.
- Activity Data from Case Tracking Systems. Since cost information on services provided by outside agencies was sometimes stated in terms of either average cost per trainee or student or average cost per month, estimating service cost in some cases required knowing how many demonstration participants utilized these outside services and in some instances for how long. Estimates of utilization were derived from the Teen Progress Management Information System (TPMIS) used by Case Managers in New Jersey to record participant activity, and the corresponding Client Activity Management System (CAMS) in Chicago.
In addition to these major sources of data, information from State AFDC records and the demonstration tracking systems was used to identify periods during which sample members were on and off AFDC, and when they were active or not active in program activities. This information was used as a basis for converting overall cost estimates into unit-cost estimates.