Four different administrative data files maintained by the North Carolina Department of Social Services provided data for the analysis files used to examine adoption dissolution, adoption disruption, and adoption assistance in North Carolina. To define the population of children to be included in this study, three data files were merged: (1) summary information on each child receiving adoption assistance, (2) vendor payments made to other post-adoption service providers in the name of adopted children, and (3)records of adoption subsidy checks. The summary child-level file was first merged with the adoption subsidy check record file, and then information on payments to vendors was added into this merged file. The total number of unique ID numbers in this merged file was 12,067. Exhibit 1 summarizes the number of unique ID numbers in each data file.
|Data table||Number of unique IDs||Data range|
|Adoption assistance child summary||11,018||1/18/1973 - 12/01/2001|
|Vendor payments||6,303||5/21/1990 - 6/20/2001|
|Adoption assistance check record||9,848||1/31/1990 - 7/13/2001|
Since there are no records in the payment files prior to January 1990, children with a final order of adoption before January 1, 1990, or who had no information on the date of the final order were identified and deleted from the final analysis file, leaving a possible 8,647 children in this file. Our next step was to assess the quality of this data merge based upon DSS-assigned ID numbers.
Assignment of multiple ID numbers complicates analysis of adoptions.
In North Carolina, children adopted from the public child welfare agency may have up to three different ID numbers. The first ID number, the foster care ID number, is assigned when a child initially enters out-of-home placement. After the adoption decree is final and if the child is to receive adoption assistance, a second ID number is assigned to track these payments. Finally, should an adoption dissolve and a child reenter out-of-home placement, a third ID number may be assigned. To further complicate these analyses, if a child receives adoption assistance in the form of reimbursement for nonrecurring costs prior to the final decree of adoption, the payments are recorded under the foster care ID number. After the adoption is final, the child may receive cash assistance and/or vendor payments recorded under the second ID number. Given the confidentiality of the adoption process, multiple ID numbers for the same child are not linked, making it impossible to assess the experiences of children across the continuum of adoption events.
To address these issues, even though it is possible that adopted children received additional assistance under the foster care ID number, we limit our analyses of adoption assistance to cash assistance and vendor payments that occur after the final decree. Additionally, we developed an algorithm to assess whether there were children with multiple ID numbers in this study population. We identified 182 pairs of ID numbers for children with the same gender, birth date, and date of final adoption decree. Since this number was so small, less than 4 percent of the study population, it appears that our restriction algorithm for study inclusion effectively resulted in a mostly unduplicated count of children. Exhibit 2 summarizes the source of information for the 8,647 children in the final study population. All but 340 children in the final data set had a record of either receiving a cash assistance payment or a vendor payment made in their name. Over half of the children received both.
|Number of children|
|Children with no vendor or cash assistance payment record||340|
|Children with vendor payment but no cash assistance payment||199|
|Children with no vendor payment but cash assistance payment||3,067|
|Children with both vendor and cash assistance payments||5,041|
|Total children with final decree after 12/31/1989 + assistance||8,647|
After using the vendor payment and adoption assistance data files to identify the study population and to analyze patterns of adoption subsidy, we accessed one additional source of information, the North Carolina longitudinal foster care placement data files. These files contain information on all children placed in out-of-home placement in North Carolina since the mid-1980s. The placement data files when linked to the cash assistance data provided the data necessary for studying adoption dissolution. Additionally, the placement data files provided data for the adoption disruption analyses. Exhibit 3 summarizes the source of data for each topical area under study. The results of the analyses are described in the sections that follow.
foster care file
|Patterns of adoption assistance||ü||ü||ü|
"report.pdf" (pdf, 1.44Mb)