Frequently Asked Questions: CSDE Public Use Data

What questions does the CSDE address?

The Child Support Demonstration Evaluation (CSDE) looks at both direct effects of the child support pass-through policy on child support paid and received, and a wide range of potential secondary effects. Specifically, the evaluation looks at the effects of the pass-through on:

• Orders and paternity establishment
• Formal and informal child support collections
• W-2 and related program costs
• Employment and income of the custodial parent
• Employment and income of the noncustodial parent
• Noncustodial parent's involvement with the child and
• Child well-being.

Why is this study important?

The results of the CSDE show that Wisconsin's full pass-through has been able to increase the amounts of child support received by members of an economically vulnerable group, to increase the amounts of child support collected, and to achieve a variety of other positive effects, at little cost to the government. These results suggest that government retention of child support, the current policy in most states, may have detrimental effects on families while offering little fiscal benefit. In contrast, the benefits to children of establishing paternity and setting a pattern of child support payments are potentially more enduring.

What kinds of data were collected?

Data were collected from two major sources:

• Administrative data from the State of Wisconsin on public assistance use and participation, child support payments and receipts, employment and wage histories.
  
  
• The Survey of Wisconsin Works Families, a panel study of mothers who participated in W-2 and of the legal fathers of a randomly selected child, conducted in two waves and focusing on families' experiences during 1998 and 1999.

What data are currently available?

Administrative data are available for a random 75 percent sample of the original CSDE population; the sample consists of approximately 17,500 cases. Data are available on the custodial parent, the children, and the noncustodial parent(s) associated with those children. Administrative data are currently available in several files, including demographic data for the custodial parent, children, and noncustodial parent; child support payment and receipt histories, employment and earning summaries, and benefit receipt histories. Data from the various files can be linked.

Survey data are available for approximately 2400 custodial mothers and 650 noncustodial fathers of a randomly selected focal child. Four files are available, for mothers and for fathers in each of the two time periods. The four files can be linked.

Data for individual cases are not available from either data source. Administrative data and survey data cannot be linked. Detailed descriptions of the data, as well as codebooks, are accessible through the links on the CSDE data page.

How is the privacy of research participants being protected?

Users will not be permitted to run programs that provide individual-level data, nor to view values with small cell sizes. Specific restrictions are detailed in the SAS program rules. Programs and output will be examined both electronically and manually to insure that these rules are followed. In addition, data available for access have been prepared in such a way as to minimize the possibility of identifying any individual or case, by rounding numbers, top-coding extremely high values, aggregating related categories of data, and dropping some identifying information. As mentioned above, files include data for only a partial sample of participants to decrease the possibility that any individual could be positively identified using demographic data. Users are required to sign a confidentiality agreement, and to notify IRP immediately if they identify any individual-level data.

How are these data accessed?

Users must apply and be approved for access to the data. After payment of a fee, users will be provided with instructions on how and where to submit SAS programs. CSDE data can only be accessed by submitting SAS computer programs; actual copies of the data files are not provided.