CLASP conducted a state-by-state analysis of key program trends between 1991 and 1995 in an effort to find out whether the child support (IV-D) program improved over the five year period. Our analysis looked at key performance indicators, caseload trends, program expenditures, and staffing ratios. Because of the wide variations in state demographics and data quality, we compared the performance of each state to itself over time.[1] By focusing on time line trends, we could look at improvements or declines in each state's performance during the five year period.

This paper summarizes nationwide trends in child support program performance between 1991 and 1995. Preliminary 1996 data also is included where available.[2] State-by-state tables are attached. A companion paper summarizes nationwide trends in child support caseloads, while a subsequent paper will highlight those states that appear to have made the most progress during the five year period.

Performance Indicators

In our study, we calculated eleven performance indicators from annual data reported by the states to the federal Office of Child Support Enforcement.[3] These eleven indicators can be broken into three groups. The first group is output measures. Output measures are a tally of successful actions performed by the program over the year (such as the number of support orders established during the year). They can tell you if the volume of work performed by the program has increased or decreased over time. To determine whether an increase in the program's volume of work actually represents increased effectiveness, we compared output measures to (1) caseload growth, (2) increase in expenditures, and (3) increase in staffing ratios. We included the following output measures in our study:

number of paternity orders established during the yearnumber of support orders established during the yeardollar amount of support collected during the year

The second group of performance indicators is caseload ratios. Caseload ratios represent a "snapshot" of the caseload at a point in time. They compare the number of cases with a certain status (such as cases with support orders) to the broader caseload. In part, caseload ratios reflect the demographic makeup of the caseload. For example, a state with a lower non-marital birthrate will probably have higher proportion of cases with paternity and legal support obligations established when they enter the caseload. However, they also can tell you how deeply the program is reaching into its caseload, and over time, whether the program is working a larger proportion of its cases.

We included the following caseload ratios in the study:

• Paternity Establishment Percentage (PEP) percentage of cases with a support orderpercentage of cases with a collection percentage of cases with a support order and collectioncurrent support received as a percentage of current support dueamount of collections in cases with a support order

The third group of performance indicators is cost-effectiveness ratios. Cost-effectiveness ratios compare the amount of collections made by the program to the amount of money spent by the program. In part, they reflect program efficiency, and in part they reflect the program's spending priorities. For example, a state that targets its spending on higher-yield cases or activities (such as collections) will look better on cost-efficiency ratios than one that works more difficult cases or puts more money into activities that do not result in an immediate pay-off (such as paternity establishment). Automated systems development substantially increased state expenditures during the period, decreasing overall cost-effectiveness. We included the following cost-effectiveness ratios in the study:

1. cost-effectiveness ratio
2. adjusted cost-effectiveness ratio, excluding statewide system expenditures
[4]

The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 directed HHS to propose changes in the current incentive payment funding formula. Current incentive payments are paid to states based on their cost-effectiveness ratios. The performance-based incentive payment proposal, expected to be reintroduced in Congress this year, would base state incentive payments on five performance indicators. Four of the performance indicators included in the CLASP's analysis are included in the proposal.[5] The four performance indicators are:

PEP Percentage of cases with a support orderCurrent support received as a percentage of current support dueCost-effectiveness

Caseload Ratios

As Table 1 reflects, most child support caseload ratios barely budged between 1991 and 1996 nationwide.[6] This means that while the nationwide child support program maintained performance levels, it did not make significant gains overall. The proportion of cases that had child support orders was 59 percent in 1991 and 59 percent in 1996.[7] The proportion of cases with any collection during the year was 19 percent in 1991 and 20 percent in 1996.[8] For cases with support orders in place, the collection rate was 33 percent in 1991 and 35 percent in 1996.[9]

However, in 1995 and 1996, some caseload ratios improved slightly. The proportion of cases with a support order increased 3 points between 1994 and 1996. The proportion of cases with a collection increased 2 points. The proportion of cases with an order and collection increased 2 points. The paternity establishment percentage (PEP) increased 9 points in 1995, to 50 percent from 41 percent in 1994 (1996 data is unavailable).[10]

The amount of current support received as a percentage of current support due jumped between 1991 and 1992, but has declined since then.[11] The proportion of current support received increased from 48 percent in 1991 to 55 percent in 1992. However, by 1996, it had dropped to 52 percent. This suggests that the program has not made further inroads in securing timely support through income withholding or voluntary compliance.

Program Outputs

As Table 2 shows, program outputs were mixed during the period. Collections fared the best, increasing at a substantially faster pace than the caseload. While the caseload increased by 44 percent between 1991 and 1996, collections increased by 75 percent. The number of paternities established increased at a somewhat faster rate than the caseload, although not as quickly as collections. During the six-year period, paternities increased by 52 percent.[12]

The weakest area was in obtaining child support orders. The number of child support orders established by the program increased by 32 percent, while the caseload increased by 44 percent. The absolute number of support orders actually dropped in 1994. There is some suggestion in the data that the program may have made internal trade-offs between increased paternities and decreased support orders. While trade-offs of this sort may have accommodated changing program priorities, they resulted in smaller net program gains.

1995 and 1996 were the only years that the child support program held its own in key output measures. This is due in large part to "denominator shrinkage:" the fact that the caseload stopped growing. In other words, paternities, support orders, and collections continued to increase as the caseload stabilized.[13]

Table 3 indicates that expenditures increased substantially more than cases but slightly less than collections between 1991 and 1996. While collections increased 75 percent and the caseload increased 44 percent, expenditures increased 69 percent. Overall expenditures increased much faster than the caseload, primarily because states sharply increased their computer systems expenditures during the period. In turn, increases in collections may be attributable to automation. In 1996, expenditures slowed dramatically. Both cases and expenditures increased only 1 percent in 1996.

Cost-effectiveness Ratios

The cost-effectiveness ratios in Table 3 reflects at least three trends.[14] First, there was a jump in cost-effectiveness in 1992, indicated both in the overall cost-effectiveness ratio and the ratio excluding automated statewide systems costs. Second, there was a sharp decline in the overall cost-effectiveness ratio after 1993, attributable to increased statewide systems expenditures. According to HHS, the increase in cost-effectiveness in 1996 is largely due to decreases in total computer systems costs. Third, if statewide systems costs are set aside, no significant improvements were made in program cost-effectiveness between 1992 and 1995 (1996 systems cost data is not available). In other words, while statewide systems costs drove the overall cost-effectiveness ratio down, they did not mask an improvement in the underlying program cost-effectiveness ratio.

Table 3: Cases, Collections, and Expenditures; Cost-Effectiveness Ratios

Staffing Increases

Constrained staffing resources may have contributed to the program's generally flat performance over the period. Over the five-year period between 1991 and 1995, increases in staff did not keep up with caseload growth, as Table 4 reflects. Although the IV-D caseload increased 43 percent between 1991 and 1995, total staff increased only 30 percent and in-house staff increased 28 percent. It is difficult to draw further conclusions from the data about IV-D staff workloads. In particular, any impact that automated case management might have had on line worker productivity during the period is unclear, since the data do not distinguish employees assigned to different activities.[15] It also is difficult to assess the extent to which staffing resources were diverted to systems activities during the period.

Recent Gains in Paternities and Collections

There are two areas where nationwide performance indicators show recent gains. Small increases in paternity establishment and collections are reflected in both output measures and the caseload ratios for 1995 and 1996. However, it is unclear whether these increases represent real progress, or instead reflect various demographic and data measurement factors.

The first area of recent gains is paternity establishment. A modest improvement is reflected in the number of paternity orders established by the program. The number of paternities increased at a slightly faster rate in 1995 and 1996 than in prior years. In addition, the number of paternities grew at a faster rate than the IV-D caseload in 1995 and 1996. The PEP increased markedly in 1995 (1996 data is not available).

The second area of recent gains is in collections. However, the collections picture is mixed. On the one hand, the dollar amount of support collected by the program increased steadily over the period, even as the caseload stopped growing in 1995 and 1996. On the other hand, collection rates did not improve substantially over the period. Only a third of cases with support orders had any collection in 1991. The same was true in 1996. In addition, the proportion of current support collected that is due has dropped recently. The data suggest that while the IV-D program may be collecting more money for families, it is not necessarily helping a larger share of families in the caseload or collecting support more timely.[18]

A Baseline for Future Performance

Nationwide, the child support program did not make significant headway between 1991 and 1995. However, the data reflects small performance gains in paternity establishment and collections in 1995. Preliminary 1996 data also reflects small gains. While the recent data is encouraging, it is too early to tell whether the gains represent improved performance, or simply improved data.

Computerization of the state child support programs has been widely viewed as a potential "magic bullet" that will boost state future performance. Anecdotally, state administrators say that computerization is beginning to have an impact on program effectiveness. Automation has reportedly improved the program's capacity to locate noncustodial parents, to track cases, and to respond faster, and has reportedly improved staff productivity by automating routine paperwork and calculations. According to the GAO, "it is too early to judge the potential of fully developed automated systems, yet bringing the benefits of automation to bear on child support enforcement appears to have played a major role in locating more noncustodial parents and increasing collections."[19]

However, states have had considerable difficulty automating their programs. The Family Support Act of 1988 required states to develop single statewide automated systems by October 1, 1995. In 1995, this deadline was extended to October 1, 1997. Although states are required to have federally-certified child support computer systems in place, less than half of states have obtained certification.[20] Most states are expected to be certified in 1998, but some of the largest states, including California and Michigan, are not close to completion.

The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA) builds on the 1988 computer requirements, and requires the states to significantly expand the capacity and functions of their systems. The new law requires states to implement a new-hire registry, to help states locate noncustodial parents through computer matching with employer reports. The new law requires states to account for and distribute collections through an automated disbursement unit. The new law also requires states to track support orders through a centralized data base, establish new automated interfaces with other public and private organizations, and exchange data with expanded federal child support data bases.

In addition, PRWORA directed HHS to propose changes in the current incentive payment funding structure. If the proposal is adopted by Congress, federal incentive payments to states would be based on the state's performance indicators. Performance-based incentive payments are widely expected to focus state resources and improve performance in critical areas.

The child support program faces enormous challenges ahead. Fundamental changes in national and state welfare policies, including time limits and more stringent eligibility rules, are placing new pressures on state child support programs to obtain child support for the lowest income families. At the same time, reduction in welfare caseloads will dramatically impact child support caseloads, service delivery, and program funding streams.[21]

Performance indicators for the period 1991 through 1996 establish a baseline against which the efficacy of increased computerization and the impact of recent legislative changes can be measured. The next few years will be telling. If the program does not take a turn for the better, policy makers and administrators must begin addressing critical questions about the underlying structure of the child support program and the ability of states to develop and maintain a program environment that can make a substantial difference for families.

By Vicki Turetsky and Andrea Watson (01/98)

Click here to order copies of this report - $7.00 each

---

NOTES

---

1. However, demographic variations among states may be less pronounced in the public assistance portion of the caseload.

Return to Text

---

2. After the attached state-by-state tables were completed, the U.S. Department of Health and Human Services released a preliminary 1996 data report. See Child Support Enforcement: FY 1996 Preliminary Data Report (August 1997). The preliminary 1996 data is included in this paper summary, but not in the state-by-state tables.

Return to Text

---

3. Published in HHS, Child Support Enforcement: 20th Annual Report to Congress for the Period Ending September 30, 1995, and preceding annual reports.

Return to Text

---

4. Statewide system expenditures that are reimbursed by the federal government at the enhanced 90 percent rate.

Return to Text

---

5. See HHS, Report to the House of Representatives Committee on Ways and Means and the Senate Committee on Finance: Child Support Enforcement Incentive Funding (February 1997). The fifth indicator (cases paying toward arrears as a percentage of cases with arrears due) was not included because the data is not currently required to be federally reported.

Return to Text

---

6. In fact, performance of the nationwide child support program has improved very little since the program's inception. In 1982, the collection rate was 15 percent. In 1996, the collection rate was 20 percent. The largest jumps in collection rates were in 1985, 1991, 1995, and 1996.

Return to Text

---

7. When compared over time, this indicator measures the program's increasing or decreasing effectiveness in establishing paternity and obtaining support obligations for children in the caseload that lack orders, as well as demographic changes in the caseload.

Return to Text

---

8. This is a bottom line indicator of program effectiveness, since it includes all IV-D cases with and without support orders established in the denominator of the ratio. A support order is a legal prerequisite to collecting support. Therefore, a low collection rate may be attributable to a high proportion of cases without a legal obligation, a small amount of collections made on cases with support obligations in place, or both.

Return to Text

---

9. This indicator specifically measures program effectiveness in collecting on existing support obligations.

Return to Text

---

10. The PEP is a IV-D performance standard first adopted by Congress in 1988, and subsequently amended in 1993, 1994, and 1996. PL. 100-485, Sec. 111(a); PL. 103-66, Sec. 103721; PL. 103-432, Sec. 213. If a state's PEP

fails to improve every year, the state is subject to a fiscal penalty. There are a number of problems with using the PEP as a measure of program performance. One is that the definition changed during the period. Another is that states may elect to use alternative definitions and to switch from year to year. Before 1995, the PEP was supposed to be calculated based on non-marital children needing paternity established in either AFDC or IV-D cases who then have their paternity established. Beginning in 1995, the PEP was supposed to be based on newborn non-marital children in the AFDC/FC or IV-D caseload, who have their paternity established or acknowledged during the fiscal year. For a fuller discussion of the PEP definition, see Paula Roberts, A Guide to Establishing Paternity for Non-Marital Children (CLASP, 1996).

Return to Text

---

11. This indicator specifically measures program effectiveness in collecting timely and regular support.

Return to Text

---

12. Unlike the PEP, these numbers only reflect paternities established by the IV-D program and do not include paternities acknowledged at birth through hospital paternity programs before the children are added to the IV-D caseload.

Return to Text

---

13. The recent declines in caseload growth rates are unprecedented in the program's history. The declines are probably due to a combination of fewer new cases and more case closures. Welfare reform changes have led to fewer public assistance cases. In addition, a number of states have pursued more aggressive case closure policies as they converted their case data to automated files and closed duplicate or unworked cases. It also is likely that some states are focusing more on their caseload "denominator" in anticipation of proposed performance-based funding changes. By the same token, the unprecedented caseload upsurges in 1992 and 1993 probably were due not only to increases in new cases, but the failure to close duplicate cases. See Little Hoover Commission, State of California, Enforcing Child Support: Parental Duty, Public Priority (May 1997).

Return to Text

---

14. The cost-effectiveness ratio is a measure of the program's collections amount compared to its expenditures. Either increased collections or decreased expenditures will increase the ratio.

Return to Text

---

15. However, the PEP probably reflects the effect of in-hospital voluntary paternity establishment, which did not rely primarily on IV-D staff.

Return to Text

---

16. Including staff under cooperative agreements and purchase of service staff. A large proportion of contractual staff were probably related to systems development and implementation.

Return to Text

---

17. According to HHS preliminary data, the total number of FTE staff in 1996 was 52,459. This represents a 2 percent increase over 1995. However, more detailed 1996 information has not been published.

Return to Text

---

18. The proportion of collections made through income withholding has increased steadily since 1991. Currently, 57 percent of collections is made through income withholding. However, the amount of current support due increased faster than the amount of current support collected in 1995 and 1996.

Return to Text

---

19. GAO, Child Support: State Progress in Developing Automated Enforcement Systems, GAO/HRD-89-10FS (1989).

Return to Text

---

20. The Family Support Act of 1988 required states to develop single statewide automated systems by October 1, 1995. In 1995, this deadline was extended to October 1, 1997. As of December 8, 1997, twenty-one states had certified systems, and four more were awaiting the results of certification reviews.

Return to Text

---

21. The child support program is largely funded through collections made on behalf of families receiving welfare benefits and assigned to the state as a condition of welfare eligibility. PRWORA changed a number of assignment and support distribution rules affecting current and former welfare families.

Return to Text

CLASP is a member of the Electronic Policy Network,
a project of The American Prospect.