More effective use of automation;

Improved location processes;

Enhanced interview techniques;

Improved customer service;

Broadened community linkages.

Colorado's Model Office Project grant (MOP) operated in small, medium and large child support offices. The project involved the participation of the state office as well as the Divisions of Child Support Enforcement in the counties of Denver, Mesa and Archuleta. The various interventions involved actions by personnel at the state and county agencies as well as by private contractors. The project was directed by a steering committee comprised of state and county child support personnel and a representative of the federal regional office. All interventions were assessed by an independent evaluation contractor.

The following describes the interventions that were pursued in Colorado's MOP and the results of the evaluation as of May 1998.

CREDIT BUREAU REPORTING

This project compares 3,000 equivalent delinquent child support cases with and without credit bureau reporting, respectively. The study reveals that credit bureau reporting increases child support payments by about six percent over a nine month period of time. This differential remained constant at sixteen months following credit bureau reporting. Generalized to the 61,139 child support cases in Colorado expected to meet the criteria for credit bureau reporting, the intervention produced approximately $3,056,950 during the first sixteen months following implementation.

Interviews with child support workers and lending industry representatives reveal substantial satisfaction with credit bureau reporting. The evaluation also revealed that: (1) the child support community is divided on the pros and cons of reporting delinquent obligors who are making regular child support payments and the amount of discretion that local child support workers should have in determining the payment status of an obligor; (2) the intervention has little immediate impact on AFDC clients who are intent on avoiding their child support responsibilities; (3) while the lending industry has been extremely supportive of the initiative, the increased reliance on subcontractors to investigate loan applications makes it difficult for child support agencies to contact lenders directly to explain the initiative, explore the feasibility of incorporating child support debt into a mortgage or consumer loan and/or obtain location information for obligors; (4) successful implementation requires outreach to county child support workers about the initiative, workload requirements, expectations and time frames.

For additional information, see Evaluation of Colorado's Credit Bureau Reporting Initiative, a report of the Colorado Model Office Project, April 25,1997.

DRIVER'S LICENSE SUSPENSION

The evaluation of this intervention involved comparing three groups of obligors that qualified for driver's license suspension, with each group receiving a different treatment: credit bureau reporting (N=436); credit bureau reporting and driver's license suspension (N=566); and neither credit bureau reporting nor driver's license suspension (N=1,702). The cases for the three groups were analyzed in terms of certain characteristics of the obligors: number of children, marital status, gender and age, and TANF/non-TANF case status. They were also reviewed in terms of driver's license status, and number and type of citations received. Following the treatment of notification for both compliance and license suspension, the cases were analyzed for patterns of response and payment.

The evaluation revealed that among those who had ever been issued a license, only about 35 percent had a valid license when the notice of suspension was sent. The rest had already been suspended (38%) or had an expired license (23%). Fully two-thirds of the obligors eligible for suspension had been the subject of a license suspension at some time in the past. Only 11 percent had never been ticketed. Overall, each obligor received an average of 10 violations, and the median was 7.

Despite the fact that many delinquent obligors are already suspended and would seemingly not be disadvantaged by the threat of suspension, the DLS intervention did produce child support payments. Overall, obligors in 17 percent of noticed cases established a repayment plan (14% in TANF cases and 28% in non-TANF cases). Those who worked out a repayment plan were twice as likely to have a valid license at the time they were notified as those who did not work out a plan. In the eight months following notification, obligors receiving the DLS treatment paid an average of $72.87 additional in child support. To date, Colorado has notified 22,000 delinquent obligors and suspended 12,000 obligors, suggesting that on an annualized basis, the state has realized $2,404,710 in additional child support due to the DLS intervention.

The DLS treatment is popular with county child support administrators and promises to yield more revenue as it is extended to obligors who only owe past due support and those with commercial licenses.

For additional information, see Evaluation of Colorado's Driver's License Suspension Initiative, a report of the Colorado Model Office Project, May 1, 1998.

WORKER'S COMPENSATION ATTACHMENT

This assessment examined child support collections due to the attachment of worker's compensation insurance in Colorado. Collections were originally projected to match those generated in Massachusetts. These expectations have not been met. One reason for the shortfall in 1996 may have been unforseen complications in computer programming so that certain collections were not appropriately credited to the workers' compensation initiative. CSE staff are confident that these "glitches" have been addressed. What seems to be a more fundamental problem, however, is the state legislation itself. By exempting partial permanent disability benefits from child support, Senate Bill 94-141 significantly limits what can be collected. Thus, it may have been inappropriate to base projections for Colorado on collection patterns in Massachusetts since that state does not exempt partial permanent disability benefits. Despite this limitation, however, the collection of child support from worker's compensation benefits appear to be rising each year, and may yet match predictions.

For additional information, see Worker's Compensation Attachment, a report of the Model Office Project, February 4, 1998.

UNEMPLOYMENT COMPENSATION BENEFITS WAGE ASSIGNMENT

The evaluation of Colorado's Unemployment Compensation Benefits Attachment involved a comparison of UCB collections 39 months prior to and following the initiative. We also gathered data on monthly totals of unemployment insurance benefits paid for the 39 months prior to and following the intervention. To better understand patterns in UCB collections, we reviewed the Labor Department's data on the total number of individuals in the labor force for the period before and after automation, as well as the unemployment rate and total number of individuals filing new claims during this period. Additionally, we conducted telephone interviews with county supervisors to assess the impact of UCB attachments in their counties.

The assessment shows that UCB payments remain steady over time, which can be attributed to the low unemployment rate and rise in the state's labor force. The increased number of individuals in the work force meant that, even with declining unemployment rates, unemployment compensation benefits increased. Child support collections over this period show a relatively steady increase. County supervisors report that the automation of UCB attachments has reduced the work associated with this collection remedy and produces good results.

For additional information, see Evaluation of Colorado's Unemployment Compensation Benefits Attachment Initiative, a report of the Colorado Model Office Project, April 15, 1998.

COMPUTER-BASED TRAINING

With the support of the Model Office Project, Colorado purchased Authorware software and paid for three state child support employees to attend a five-day training program conducted by Policy Studies, Inc. Over the next 15 months, in between many other job duties, these employees developed a comprehensive, multi-media, interactive, computer-based training product (CBT) for child support workers dealing with the location process. Knowledge tests administered to workers exposed to CBT and to conventional, stand-up training on the location process reveal that the two formats are equally effective in imparting child support information. Attitudinal surveys also reveal that both approaches meet with worker approval. To the extent that there are differences, workers tend to find CBT more convenient and entertaining. They are somewhat more apt to find stand-up training responsive to their questions and less apt to find it confusing. Interviews with child support administrators in Colorado reveal that they are eager for the state to distribute the location CBT through the Internet and to develop modules dealing with other topics. Many administrators contend that they cannot afford to send workers to Denver for training and/or to participate in regional training programs. They view CBT as a flexible and affordable training resource that can be used with new and veteran workers in a variety of office settings. Although the process of developing the first CBT module was more complicated than anticipated, the developers believe that future ones will be easier and faster, especially if they have a block of time that they can dedicate to the project. Although stand-up training will still be needed for new initiatives and legislation, workers and administrators believe that many child support training topics are amenable to treatment using CBT formats and that their development should be pursued.

For additional information, see Evaluation of Computer-Based Training, a report of the Colorado Model Office Project, January 14, 1998.

ENHANCEMENTS TO THE AUTOMATED CHILD SUPPORT ENFORCEMENT SYSTEM

In the last quarter of 1996, a number of procedures designed to help county child support workers manage system-generated locate response information were initiated on Colorado's Automated Child Support Enforcement System (ACSES). This included: (1) deleting locate responses on ACSES that were over three years old if there is more current information about the absent parent's residence or employment; (2) displaying on each technician's log-on screen a count of locate responses received and the number of responses that need review and disposition; (3) providing "user friendly" information for processing initiating interstate actions and moving initiating interstate actions from ACSES category 3 (locate) to category 2 (enforce); and (4) generating post office and employer verification letters from locate functions in ACSES.

Following the introduction of these enhancements, the number of locater responses recorded on ACSES dropped from 398,373 to 174,292. An on-line survey completed by 160 county child support workers and interviews with approximately a dozen workers and administrators at the state and county levels reveals considerable satisfaction with these enhancements. Most report that the locate process is easier, more effective and less stressful than it used to be. Many feel that the log-on prompt is a helpful reminder of the number of responses to be reviewed and a way of monitoring locater efficiency. Although workers favor the automatic deletion of old locate information, some feel that the system still contains too much outdated information and hope to see additional refinements of the locate function in the future. They would also like to have on-line access to new location resources such as the state penal system and student loan information. Finally, they report a need for more printers in order to take full advantage of the enhanced document generation capacities of the ACSES.

For further information, see Evaluation of Locate Enhancements on the Automated Child Support Enforcement System, a report of the Colorado Model Office Project, March 18, 1997.

PRIVATIZED LOCATION INTERVENTION

Mesa and Archuleta Counties retained private contractors to attempt to locate absent parents in their child support cases in categories 3, 5 and 7. In Mesa County, 1500 cases were randomly assigned to two private vendors (Lockheed, IMS and TMR) and in-office child support staff, respectively. In Archuleta County, cases needing location work were assigned to TMR. Vendors and in-office county staff kept track of their location efforts, the sources they consulted, and the nature of the information they verified. All cases were reviewed six months later for evidence of child support order establishment and payment activity. A comparison group of cases that was not systematically worked by the vendors or in-office staff was studied to determine what happens to cases needing location work in the absence of concerted effort by vendors and/or in-office staff.

The analysis revealed that performance patterns were comparable for the private vendors and in-office staff. All tended to focus their attention on the same types of cases, and despite differences in the use of various locate sources, they were equally likely to verify a new home and employer address. A six month follow-up showed that regardless of the public versus private status of the worker, approximately 20 percent of worked cases were in a paying child support status as compared with 29 percent of cases in the control group. Clearly public workers can do as well as private vendors when they try. The question is whether in-office staff are able to make location work a priority in the normal course of events when they are not competing with private vendors. These findings are consistent with national research on the effectiveness of private child support offices indicating that they perform at least as well or, in some instances, better than public ones.

For further information, see Assessment of Privatized Location in Mesa and Archuleta Counties, a report of the Colorado Model Office Project, February 15, 1996.

STATE ENFORCEMENT UNIT INVESTIGATOR

The State Child Support Enforcement Unit (SEU) retained an experienced location specialist at the state level to help the counties locate obligors. The location specialist was given a random sample of 606 child support cases drawn from the total pool of cases in category 3 in Denver, Mesa, and Archuleta counties. The investigator kept track of the efforts he made to locate obligors for each of these cases. Approximately 7-9 months following this intervention, we checked the Automated Child Support Enforcement System (ACSES) to gauge the location and payment status of each case.

The analysis shows that the efforts of the location specialist clearly made a difference. In approximately 10 percent of the cases, the intervention led to "clean-up" activities that removed the case from Category 3. In nearly 20 percent, the intervention resulted in the case moving to Category 1 indicating full payment, although in more than half the cases the move was only temporary. At the follow-up review, 56 percent of the cases were still in Category 3, 21 percent were in Category 2, 8 percent were in Category 1, and 15 percent were closed or in other categories.

With the automation of the child support system, location work ceased to be a specialized function in Colorado and came to be viewed as a mechanical process to be performed in conjunction with other enforcement and establishment duties. Nevertheless, this intervention and others conducted in the Model Office Project suggest that good locate work requires concerted and persistent efforts by child support workers and that this focus is more important in explaining outcomes than the state, county or private standing of the worker performing the job.

For additional information, see State Enforcement Unit Location Intervention, a report of the Colorado Model Office Project, February 12, 1998.

AUTOMATED STATUS LETTERS

In order to facilitate more regular contact with custodial parents, the child support administrator in Mesa County designed an automated method of generating letters to custodial parents. "Status" letters are intended to apprise custodial parents that the agency needs more information about the noncustodial parent before it can establish or enforce a child support order.

Status letters were sent to 616 custodial parents in Mesa County child support cases needing location information. As a result of the mailing, new information and/or case closings were produced in 19 percent of the targeted cases. ACSES has been programmed to generate automated status letters on a quarterly basis in all Colorado counties. County child support administrators view this as a cost-effective way of communicating with custodial parents on a regular basis and creating a flow information between custodial parents and child support agencies.

For additional information, see Evaluation of the Generation of Automated Status Letters, a report of the Colorado Model Office Project, April 27, 1998.

DIRECT WAGE WITHHOLDING

Direct Wage Withholding (DWW) was initiated in interstate child support cases as a way to get child support payments to the custodial parent faster and to decrease the workload for county workers. The goal of this intervention was to use direct wage withholding as an enforcement remedy of first resort in a cooperative arrangement with several states where Colorado has a significant volume of interstate cases. Texas and Arizona were selected for a pilot project with three counties in Colorado.

Although Texas and Arizona share a substantial volume of interstate cases with Colorado, the number of cases subject to direct wage withholding procedures was too small to conduct an accurate evaluation. The project was further hampered by the unfamiliarity of workers and employers with the procedure. While it is true that direct income withholding often works when it is attempted, there are serious underlying problems with it which must be solved for it to become an effective enforcement remedy.

For further information, see Evaluation of Colorado's Direct Wage Withholding Initiative, a report of the Colorado Model Office Project, April 24, 1998.

USE OF AN IN-HOUSE PROCESS SERVER

In an effort to improve performance, the Denver County Child Support Division retained a full-time salaried employee to serve papers associated with the program. An evaluation was conducted to compare the service rate for this employee and contract personnel. It was hoped that the staff employee would work more closely with Division personnel and that the improved communication would result in higher rates of successful service.

Data collected from November 1995 through September 1996 reveal that service rates were virtually identical for in-house and contractual personnel. The two types of workers served 53 percent and 56 percent of the cases they were randomly assigned. Excluding cases with bad addresses, the rate for both rose to 57 and 58 percent. A review of expenditures by the Division for both service sources indicates that the private vendor was paid about $10 for each good service while the in-house process server received an estimated $54 when salary, benefits and mileage were taken into account. Although child support technicians in the Division expressed fewer complaints about the in-house server as compared with contractual personnel, they felt that the performance of both types of workers deteriorated over time and that low rates of successful service may be the "nature of the beast" in the child support arena. The in-house process server thought that rates could be improved if technicians did more aggressive searches for location information on the ACSES, the Department of Motor Vehicles and the Department of Labor, and conveyed more complete information to process servers.

For additional information, see Assessment of an In-House Service of Process Program in the Denver County Division of Child Support Enforcement, a report of the Colorado Model Office Project, May 6, 1997.

IV-A/IV-D INTERVIEW INTERVENTION

In an effort to improve client satisfaction, promote cooperation, and generate more complete information from clients, Denver and Mesa county experimented with two different intake interview formats for new AFDC applicants: joint IV-A/IV-D interviews conducted by a single staff member, and separate interviews conducted by two specialists in the IV-A and IV-D agencies, respectively. Archuleta County used separate interviews. Staff and clients completed brief questionnaires immediately after the conclusion of each interview. In addition, all cases were reviewed six months after the conclusion of the interview to determine their child support and welfare status and the speed with which various child support actions were taken.

The evaluation suggests no consistent benefits associated with either interview approach. Cases with joint and same day but separate interviews looked quite comparable with respect to the information provided by the client, the actions taken, and the overall status of cases within a six-month time frame. Supervisors and technicians in the IV-A and IV-D agencies trying both approaches also had mixed reactions to the two processes. There was no clear consensus that the joint interview did a better job of collecting the information needed to process the public assistance and/or child support application. Clients were perceived to be pleased by having only a single interview, but workers did not agree about whether this translated into benefits for the agency with respect to speed of case processing or quality of information gathered.

One aspect of the joint interview process that was uniformly preferred was the on-line entry of information during the interview. It was credited with reducing repetitive input and errors. It has since been incorporated into the interview procedures at all three sites.

According to workers, several factors other than interview format have had an impact on case processing and outcomes. This includes computerized access to birth certificate information and the changing atmosphere and attitudes engendered by the onset of welfare reform.

For additional information, see Innovative Interviewing Techniques, a report of the Colorado Model Office Project, January 30, 1998.

THE NON-COOPERATION SANCTION

This intervention compared the disposition of non-cooperation cases when the sanction action was taken by the IV-A versus the IV-D agency. Child support staff kept track of 54 cases cited for non-cooperation in Denver and 113 cases in Mesa County during April-November 1996. In February 1997, all these non-cooperation cases were checked for evidence of client cooperation, discontinuation and error. There was no random assignment of cases to IV-A versus IV-D treatment categories: in Denver the sanction process was handled by the IV-A agency, in Mesa County it has handled by IV-D.

The evaluation revealed that more errors occur when the child support technician relies on the IV-A worker to perform the sanction. In Denver, the error rate stood at 22 percent as compared with 5 percent in Mesa County where the IV-D workers performed the sanction. Interviews with child support workers in both counties confirmed that they would like to retain more control over the sanction process.

For further information, see The Non-Cooperation Sanction, a report of the Colorado Model Office Project, February 26, 1997.

THE GOOD CAUSE EXEMPTION TO CHILD SUPPORT COOPERATION REQUIREMENTS

Pursuant to the Family Violence Amendments to the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 ("PRWORA"), states are urged to consider the impact of domestic violence on applicants. In this intervention, we examined the experiences of victims of domestic violence when they apply for public assistance and pursue child support intake procedures. We also examined their experiences when requesting waivers from child support cooperation requirements through conventional good cause application procedures. Our study is based on interviews with 20 victims of domestic violence in Denver County who applied for public assistance and a review of 69 applications for good cause.

For additional information, see Child Support Policies and Domestic Violence: A Preliminary Look at Client Experiences with Good Cause Exemptions to Child Support Cooperation Requirements, a report of the Colorado Model Office Project, January 1997. A shortened version of this report appears in "Child Support Policies and Domestic Violence," Public Welfare, Winter 1997 and Domestic Violence Report, Vol. 2, No. 4, June/July 1997.

Following this assessment, we designed and conducted training programs on domestic violence for self-sufficiency and child support workers in Denver, Mesa, and Archuleta Counties. Screening procedures, simple notices and data collection forms were developed, and intake workers began to systematically assess applicants for domestic violence. Those who disclosed were apprised of the good cause option. Those who were uninterested in good cause were interviewed about their reasons for not pursuing it. Those who were interested in good cause are referred to local domestic violence professionals for assistance with the application and documentation requirements. All victims were given information on relevant community resources.

The project yielded information on 1,082 female, new applicants for public assistance. Across the four office sites in which the screening was conducted, 40 percent disclosed current or past abuse. Disclosure rates for individual offices ranged from 28-49 percent. Most of the abuse reported by women involved former partners (74%), who were the fathers of one or more of their children (75%). Nearly all the women (81%) reported being hit or beat up with half characterizing it as occurring more than a few times and half saying it last happened within the past two years. Despite the seriousness of the abuse reported by many victims, only 2.7 percent of all public assistance applicants or 6.7 percent of the women who disclosed domestic violence said they would be interested in applying for good cause. The rest said that they wanted child support and/or that the pursuit of child support would not put them at additional risk. Only a third of the women who expressed an interest in good cause wound up getting it; two-thirds had their applications denied. These women were generally unable to produced documents to support their claim of domestic violence and threats of harm.

For additional information, see Child Support and Domestic Violence: The Victims Speak Out, a report of the Colorado Model Office Project dated March 6, 1998, that is forthcoming in the journal Violence Against Women.

A PRIVATIZED CUSTOMER SERVICE UNIT IN DENVER COUNTY

In an effort to improve client satisfaction and relieve technicians of frequent telephone interruptions, the Denver Division of Child Support Enforcement contracted with Lockheed Martin IMS to handle all its phone calls and simple case actions effective April 1, 1996. The evaluation of the intervention consisted of: telephone interviews with 242 parents who phoned the Division prior to and 249 who phoned following privatization; interviews with 22 child support technicians six and twelve months following implementation; the conduct of focus groups with custodial parents and customer service representatives; and review of various indicators of child support performance for Denver County before and after the implementation of the Unit.

The assessment revealed that while many parents had opposed the creation of a special team of workers to handle telephone duties, those who phoned the privatized Unit were extremely satisfied with the service they received. Callers appreciated reaching a human being rather than an answering machine and characterized representatives as knowledgeable and helpful. Half of the callers who remembered phoning the agency both prior to and following privatization characterized service as "much" or "somewhat" better.

Child support technicians and supervisors also reported being more pleased with the Unit than they had expected to be. Nearly all interviewed staffers rated the Unit favorably and supported its continuation. Despite fears that representatives would lack the skill to help clients, most believed that the Unit had dramatically cut the number of calls they received and as a result they were able to clean up and close old cases, initiate wage assignments, and prepare cases for court action. Although there is some question about the precise level of referrals to technicians, a recent assessment conducted by Lockheed IMS revealed that only 20 percent were referred to technicians for further action and that representatives were handling 80 percent of calls to the Division entirely on their own.

Although there is no evidence that child support collections or other enforcement activities have increased markedly since the creation of the Customer Service Unit, Division personnel are confident that performance patterns will improve over time. Most of the problems associated with the Unit are a function of higher-than-expected call volumes and low staffing levels. As a result, over half of interviewed parents report having to wait a long time to speak with someone and approximately 27.4 percent of all calls to the Unit are abandoned before a representative gets on the line. Technicians also feel that representatives could do better quality work, handle more duties and refer even fewer calls to them if they were not so pushed for time. Division personnel are split on the future of the Unit with about half favoring an in-house effort and half favoring the continuation of the privatized approach. Those who favor the in-house effort stress the importance of representatives having easy access to Division personnel. Those who favor the privatized approach feel that it provides a needed measure of accountability and reduces opportunities for collusion between representatives and technicians.

For information on the implementation of the privatized customer service intervention and early reactions to it, see Preliminary Assessment of the Privatized Customer Service Unit for the Denver County Division of Child Support Enforcement, a report of the Colorado Model Office Project, June 16, 1996. For information on reactions to the Unit approximately one year after its implementation, see Ongoing Assessment of the Privatized Customer Service Unit for the Denver County Division of Child Support Enforcement, a report of the Colorado Model Office Project, April 4, 1997.

AN IN-HOUSE CUSTOMER SERVICE UNIT IN MESA AND ARCHULETA

In an effort to improve client satisfaction and relieve technicians of frequent telephone interruptions, the Mesa County Division of Child Support Enforcement hired a worker to handle all its phone calls. Callers to Archuleta County were referred to the Mesa County customer service representative. The evaluation of this intervention consisted of mailed questionnaires completed by 38 parents in Mesa and 20 parents in Archuleta prior to the implementation of the customer service intervention, and telephone interviews with 150 parents in Mesa and Archuleta who phoned the Division following implementation. In addition, child support technicians were interviewed before and after the intervention. Finally, we examined child support performance patterns for the two counties to gauge whether collections and other indicators of child support activity had improved.

The evaluation shows that the in-house customer service program enjoyed high approval ratings by both clients and staff. A higher proportion of parents in 1997 than in 1996 reported receiving professional treatment, were more apt to characterize the customer service representative as knowledgeable about the child support system, and said that the representative tried hard to be helpful. They were also more satisfied with actions taken and case follow-up. Nearly one-half of the callers who contacted the Division in both 1996 and 1997 considered service to be improved. Child support workers were also pleased with the customer service representative and said that they had experienced significant reductions in direct calls. A comparison of performance measures for technicians before and after introduction of the customer service program revealed an upward trend in total wage assignments, total verified locates, and collections due to wage assignment. However, performance in most categories varies monthly and it was difficult to discern a trend. While technicians reported that they spend more time doing child support duties that will result in increased collections over time, so far the intervention has only allowed them to catch up on work. In addition, they report new demands on their time from the state's mass case processing initiatives such as credit bureau reporting and driver's license suspension.

For additional information, see Assessment of an In-House Customer Service Program in Mesa County and Archuleta County Divisions of Child Support Enforcement, a report of the Colorado Model Office Project, May 27, 1997.

EMPLOYER BROCHURE

We designed a brochure for Colorado employers that explains their role in the child support program including new hire reporting, medical support, wage withholding, and direct income withholding across state lines. The brochure was mailed to approximately 15,000 smaller employers along with wage withholding bills sent by the Family Support Registry. We also mailed the brochure directly to more than 15,500 larger employers with 20 or more employees.

DEBT AND RETROACTIVE SUPPORT

One of the last interventions of the Model Office Project is the Debt and Retroactive Support intervention. This intervention experiments with ways of promoting the payment of current child support. Specifically, it explores the feasibility and impact of dropping debt/retroactive support obligations to promote the payment of current child support orders.

A random sample of non-treatment and treatment cases was generated in two Colorado counties: Mesa and Jefferson. In the non-treatment group, cases receive the usual calculation and notification with respect to debt and retroactive support. In the treatment group, debt and retroactive support is waived. The payment behavior for each of these two groups is measured over a one-year period to determine if the establishment of debt and retroactive support changes the noncustodial parent's payment of current support. In addition, the qualitative aspect of this evaluation focuses on interviews with child support enforcement workers and noncustodial parents. Currently, data forms and parent survey are being collected and entered for analysis.

For additional information, see Preliminary Report on Debt and Retroactive Support Intervention, a report of the Colorado Model Office Project, April 24, 1998.