< Back to Search

Child Support Cost Avoidance in 1999, Final Report

Published: June 6, 2003
Information About:
State/Local Child Support Agencies
Topics:
Funding
Types:
Guides/Publications/Reports, Research & Data, HHS/ACF/OCSE Research

IDIQ Contract No. 105-00-8303
Task Order No. 5

Prepared for:
Department of Health and Human Services
Administration for Children and Families
Office of Child Support Enforcement

By Laura Wheaton
The Urban Institute
2100 M Street N.W.
Washington, D.C. 20037

June 6, 2003

Acknowledgements

This report was funded by the Federal Office of Child Support Enforcement (OCSE) through a subcontract with The Lewin Group. The author thanks Gaile Maller and James Rich (OCSE), Burt Barnow (Johns Hopkins University), Tim Dall (The Lewin Group), John Tapogna (ECONorthwest), Linda Mellgren (HHS/ASPE) and Elaine Sorensen and Linda Giannarelli (The Urban Institute) for their conceptual input to the project.

Several people at the Urban Institute contributed to the work presented here. Jessica Kelly performed the programming necessary for the cost avoidance estimates, with additional help from An-Lon Chen and Paul Johnson. Elaine Maag assisted with the development of the SSI estimates, Sandi Nelson assisted in validating the results, and Brenda Brown provided secretarial assistance.

Thanks are also extended to Amanda Barlow and Mike Bratt of the Administration for Children and Families at the U.S. Department of Health and Human for guidance in interpreting the OCSE financial data, and to Vicki Turetsky of the Center for Law and Social Policy for information regarding rules pertaining to the assignment of child support.

This work would not have been possible without use of the TRIM3 microsimulation model, which is developed and maintained by the Urban Institute under contract with the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services.

The views expressed in this document are those of the author and do not necessarily reflect those of the Urban Institute or its funders.

Executive Summary

The nation's Child Support Enforcement (CSE) Program works to promote family self-sufficiency and child well-being through the collection of child support. The CSE Program (also referred to as IV-D) is a federal/state/local partnership that locates non-custodial parents, establishes paternity when necessary, establishes orders for support, and collects child support payments for families.

The CSE Program was established in 1975 in an effort by Congress to reduce public expenditures on welfare. Through a process referred to as "cost recovery," child support collected on behalf of welfare recipients is retained by the government in order to help offset welfare benefits. During the 1980s and 1990s, welfare was transformed from a cash assistance program that eligible families could rely on indefinitely, to a time-limited benefit aimed at moving families to work and self-sufficiency. Given the low wages available to many former welfare recipients, child support was increasingly recognized as an important additional source of income for helping families develop and maintain self-sufficiency. As a result, the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA) included "family first" distribution rules that enable former welfare recipients to receive more of the child support collected on their behalf. The President's budget for FY 2003 proposes legislation that would allow current and former welfare recipients to retain even more of the child support paid on their behalf (OMB 2002).

As more child support is targeted toward helping former welfare recipients achieve self-sufficiency, less is available for cost recovery. Yet even the goal of promoting self-sufficiency suggests some financial benefits for government, since increased levels of financial stability and independence mean that families require less assistance from the government in other areas, such as food, housing, and medical assistance. These financial benefits - referred to here as "cost avoidance" - accrue not just from former welfare recipients, but also from families whose child support (when combined with other income) has enabled them to avoid welfare entirely and to reduce or eliminate their need for assistance from other government programs.

Several studies have attempted to measure cost avoidance at the state and national levels. A recent review of the cost avoidance literature funded by OCSE found two basic approaches to estimating cost avoidance (Barnow et al 2000). The first approach, employed by researchers in several states, involves combining administrative data from various sources in order to measure the reduction in transfer program benefits resulting from payment of child support. The second approach entails using microsimulation techniques to develop nationally representative cost avoidance estimates using household survey data.

This report updates and expands upon the earlier microsimulation approach used by Wheaton and Sorensen (1998). The cost avoidance estimates presented here are for 1999 and cover the Supplemental Security Income (SSI) program, Temporary Assistance for Needy Families (TANF), federal housing assistance, the Food Stamp Program, and Medicaid. Separate estimates are presented for cost avoidance arising from IV-D collections and those arising from private child support transactions. While research suggests that child support provides additional benefits to families and society (i.e., improved child well-being and child educational outcomes, greater contact between fathers and their children, and possibly a reduced level of divorce and out-of-wedlock childbearing), measuring these benefits is beyond the scope of this report.

Methodology

Cost avoidance estimates are generated using the TRIM3 microsimulation model (Transfer Income Model, Version 3). TRIM3 is a comprehensive microsimulation model, developed and maintained by the Urban Institute under contract with the Office of the Assistant Secretary for Planning and Evaluation of the U.S. Department of Health and Human Services. For more than twenty years, the TRIM family of models has been used to estimate the effects of proposed changes to means-tested transfer programs and the tax system. TRIM3 captures state variation in program rules, including the detailed rules of state TANF programs. The cost avoidance estimates are produced using data from the March 2000 Current Population Survey (CPS), which provides detailed household and person-level income and demographic data for calendar year 1999.

Several steps are involved in estimating cost avoidance. First, TRIM3 must calculate the reduction in transfer program benefits that results when child support reduces the amount of benefits a participating family receives, renders an otherwise eligible family ineligible for assistance, or is sufficient to cause an eligible family to decide not to participate. The reduction in transfer program benefits is then disaggregated into that arising from IV-D collections and that arising from private child support payments. Since child support is underreported on the CPS, the initial TRIM3 estimate is likely to understate the total reduction in transfer program benefits. A revised estimate of cost avoidance attributable to IV-D collections is made by multiplying the TRIM3 "cost avoidance ratio" - the average reduction in transfer program benefits per dollar of child support - by the amount of child support distributed according to OCSE financial data.

Summary of Findings

The estimates developed in this report show that cost avoidance from IV-D child support collections exceeds the total amount of cost recovery in FY 1999. Of the total combined $4.9 billion in cost avoidance and cost recovery produced by IV-D child support collections, less than half (48 percent) is from cost recovery (figure 1).

Cost avoidance is highest in the Food Stamp Program, accounting for 37 percent ($936 million) of the total $2.6 billion in cost avoidance from IV-D child support collections (figure 2). Cost avoidance to the Food Stamp Program occurs when child support lowers the amount of benefit a participating household receives, renders an otherwise eligible household ineligible, or is sufficient to cause an eligible household to decide not to participate. Cost avoidance to the Food Stamp Program would be even greater if it were not for the fact that households that pay child support are able to deduct the child support payment from income when calculating the benefit amount. Since the Food Stamp Program is a federally funded entitlement program, the $936 million in cost avoidance reflects direct savings to the federal government.

Figure 1. Government Benefits from IV-D Collections

Government Benefits from IV-D Collections

Source: TRIM3 Microsimulation Model and FY 1999 OCSE Financial Data.

Figure 2. Sources of IV-D Cost Avoidance

Sources of IV-D Cost Avoidance

Source: TRIM3 Microsimulation Model and FY 1999 OCSE Financial Data.

TANF accounts for 29 percent ($735 million) of total cost avoidance from IV-D child support collections. Cost avoidance to the TANF program occurs when child support income renders a family ineligible for assistance or makes an eligible family decide not to participate. Given the TANF block grant structure, a reduction in TANF assistance does not directly decrease federal or state spending on TANF-related programs but rather frees up resources for training programs, job search, transportation, child care, and other support services.

Medicaid accounts for 23 percent ($588 million) of total cost avoidance from IV-D child support collections. This estimate captures cost avoidance that arises when child support renders a person ineligible for Medicaid. In addition, if child support is sufficient to keep a person from receiving SSI or TANF, and as a result, the person chooses not to enroll in Medicaid, then the resulting Medicaid savings are counted as cost avoidance. The cost avoidance estimate does not capture the savings to the Medicaid program that arise when health insurance coverage is obtained from the noncustodial parent through a medical support order. In addition, no cost avoidance is attributed to children who would be eligible for the State Children's Health Insurance Program (SCHIP) in the absence of Medicaid. Since Medicaid is a federal-state matching entitlement program, the $588 million in cost avoidance reflects shared savings to federal and state governments.

SSI accounts for 6 percent ($154 million) of total cost avoidance from IV-D child support collections. Cost avoidance to the SSI program occurs when child support income lowers the benefit amount for a participating child, renders an otherwise eligible child ineligible, or is sufficient to cause an eligible child to decide not to participate. Due to data limitations, the cost avoidance estimate only captures the first source of cost avoidance - a lower benefit resulting from receipt of child support income. Furthermore, cost avoidance arising from reductions in state supplementary SSI payments is not captured. SSI is a federally funded entitlement program, so the $154 million in cost avoidance reflects direct savings to the federal government.

Federal housing assistance programs account for the remaining 6 percent ($142 million) of total cost avoidance from IV-D child support collections. This includes households receiving either project-based assistance (public housing and Section 8 project-based) or household-based assistance (Section 8 certificates and vouchers). Cost avoidance arises when child support income increases the household's rental payments - which are typically set at the larger of 10 percent of gross income or 30 percent of net income. Although the federal housing programs are discretionary programs, cost avoidance may indirectly reduce federal spending on housing assistanceand help localhousing authorities cover the operating costsassociated withpublic housing.

It is important to note that the IV-D estimates presented here show the costs avoided as a result of IV-D collections, and cannot be entirely attributed to IV-D enforcement activities. Undoubtedly, some percentage of IV-D child support collections would be paid even in the absence of the IV-D system. It is beyond the scope of this report to estimate the amount of child support collected and costs avoided solely as a result of IV-D enforcement efforts. However, the IV-D cost avoidance estimates are consistent with estimates that attribute all cost recovery to the IV-D system (Committee on Ways and Means 2000).

The numbers presented here should be viewed as a lower-bound estimate of cost avoidance. In cases where data limitations resulted in a range of estimates, the methodological approach yielding the lowest cost avoidance figure was selected. Furthermore, the estimate does not capture all sources of cost avoidance. Additional cost avoidance arises from means-tested transfer programs not included in the estimates. Notable examples include the Child Care Development Fund (CCDF) and the State Children's Health Insurance Program (SCHIP). Given the limited research concerning the effect of child support enforcement on work, marriage, and childbearing decisions (Barnow et al. 2000), behavioral effects are not included in the estimates presented here.

Conclusion

To the extent that the number of families receiving TANF continues to decline, cost avoidance can be expected to increase in size relative to cost recovery. The President's budget for FY 2003 proposes legislation that would allow current and former welfare recipients to retain more of the child support paid on their behalf (OMB 2002). If implemented, these provisions can be expected to reduce cost recovery, while increasing cost avoidance.

Unlike cost recovery, cost avoidance can only be estimated. It cannot be directly measured. But when even a lower-bound estimate of cost avoidance exceeds the total amount of cost recovery, it is clear that cost avoidance is an important part of the picture. As the child support enforcement community calls attention to child support's ability to improve families' financial stability and independence, it is worth recognizing that this increased independence also implies financial benefits to government through cost avoidance.

Introduction

The nation's Child Support Enforcement (CSE) Program works to promote family self-sufficiency and child well-being through the collection of child support. The CSE Program (also referred to as IV-D) is a federal/state/local partnership that locates non-custodial parents, establishes paternity when necessary, establishes orders for support, and collects child support payments for families.

The CSE Program was established in 1975 in an effort by Congress to reduce public expenditures on welfare under the Aid to Families with Dependent Children (AFDC) program. Through a process referred to as "cost recovery," child support collected on behalf of welfare recipients is retained by the government in order to help offset welfare benefits. During the 1980s and 1990s, welfare was transformed from a cash assistance program that eligible families could rely on indefinitely, to a time-limited benefit aimed at moving families to work and self-sufficiency. Given the low wages available to many former welfare recipients, child support was increasingly recognized as an important additional source of income for helping families develop and maintain self-sufficiency. As a result, the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA), which replaced the cash entitlement AFDC program with the Temporary Assistance for Needy Families (TANF) block grant, also included "family first" distribution rules that enable former welfare recipients to receive more of the child support collected on their behalf.[1] The President's budget for FY 2003 proposes legislation that would allow current and former welfare recipients to retain even more of the child support paid on their behalf (OMB 2002).

Cost recovery represents a relatively small proportion of total child support collections, and that proportion has diminished in recent years as the proportion of the CSE caseload receiving TANF has declined and as family first distribution rules have caused more money to be passed on to families. In 1996, total IV-D collections amounted to $12.0 billion dollars, approximately 19 percent ($2.3 billion) of which represented cost recovery (OCSE 1998). By 2000, total CSE collections had risen to $17.9 billion, of which only 16 percent ($2.4 billion) represented cost recovery (OCSE 2001). Had it not been for the elimination of the "$50 pass-through" policy, under which AFDC recipients were paid the first $50 of child support collected in each month on their behalf, cost recovery would have declined in absolute as well as percentage terms.[2]

Given the goal of promoting self-sufficiency, it is not surprising that cost recovery represents a declining share of CSE collections. Yet even the promotion of self-sufficiency suggests some financial benefits for government, since increased levels of financial stability and independence mean that families require less assistance from the government in other areas, such as food, housing, and medical assistance. These financial benefits - referred to here as "cost avoidance" - accrue not just from former welfare recipients, but also from families whose child support (when combined with other income) has enabled them to avoid welfare entirely and to reduce or eliminate their need for assistance from other government programs.

Several studies have attempted to measure cost avoidance at the state and national levels. A recent review of the cost avoidance literature funded by the Office of Child Support Enforcement (OCSE) found two basic approaches to estimating cost avoidance (Barnow et al 2000). The first approach, employed by researchers in several states, involves combining administrative data from various sources in order to measure the reduction in transfer program benefits resulting from payment of child support.[3] The second approach entails using microsimulation techniques to develop nationally representative cost avoidance estimates using household survey data from the March Current Population Survey (Wheaton and Sorensen 1998). Wheaton and Sorensen estimate that child support generated between half a billion and a billion dollars in cost avoidance to AFDC, the Food Stamp Program, and Medicaid in 1989.[4]

This report updates and expands upon the earlier microsimulation approach used by Wheaton and Sorensen (1998). The cost avoidance estimates presented here are for 1999 and cover the Supplemental Security Income (SSI) program, Temporary Assistance for Needy Families (TANF), federal housing assistance, the Food Stamp Program, and Medicaid. Cost avoidance in the entitlement programs (SSI, the Food Stamp Program, and Medicaid) represents direct budgetary savings for government, whereas cost avoidance in TANF and the federal housing assistance programs frees funds for use in serving additional families and/or providing additional services, and may eventually reduce federal expenditures. Separate estimates are presented for cost avoidance arising from IV-D collections and those arising from private child support transactions.

It is important to note that the IV-D estimates presented here show the costs avoided as a result of IV-D collections, and cannot be entirely attributed to IV-D enforcement activities. Undoubtedly, some percentage of IV-D child support collections would be paid even in the absence of the IV-D system. It is beyond the scope of this report to estimate the amount of child support collected and costs avoided solely as a result of IV-D enforcement efforts. However, the IV-D cost avoidance estimates are consistent with estimates that attribute all cost recovery to the IV-D system (Committee on Ways and Means 2000).

The numbers presented here should be viewed as a lower-bound estimate of cost avoidance. In cases where data limitations resulted in a range of estimates, the methodological approach yielding the lowest cost avoidance figure was selected. Furthermore, the estimate does not capture all sources of cost avoidance. Additional cost avoidance arises from means-tested transfer programs not included in the estimates. Notable examples include the Child Care Development Fund (CCDF) and the State Children's Health Insurance Program (SCHIP). In addition, the estimates do not capture the savings to the Medicaid program that arise when health insurance coverage is obtained from the noncustodial parent through a medical support order. While research suggests that child support provides important additional benefits to families and society (i.e., improved child well-being and child educational outcomes, greater contact between fathers and their children, and possibly a reduced level of divorce and out-of-wedlock childbearing), measuring these benefits is beyond the scope of this report.[5]

Methodology

Cost avoidance estimates are generated using the TRIM3 microsimulation model (Transfer Income Model, Version 3). TRIM3 is a comprehensive microsimulation model, developed and maintained by the Urban Institute under contract with the Office of the Assistant Secretary for Planning and Evaluation of the U.S. Department of Health and Human Services.[6] For more than twenty years, the TRIM family of models has been used to estimate the effects of proposed changes to means-tested transfer programs and the tax system. TRIM3 captures state variation in program rules, including the detailed rules of state TANF programs. The cost avoidance estimates are produced using data from the March 2000 Current Population Survey (CPS), which provides detailed household and person-level income and demographic data for calendar year 1999.[7]

Several steps are involved in estimating cost avoidance. First, TRIM3 must calculate the reduction in transfer program benefits that results from payment of child support. The reduction in transfer program benefits is then disaggregated into that arising from IV-D collections and that arising from private child support payments. Since child support is underreported on the CPS, the initial TRIM3 estimate is likely to understate the total reduction in transfer program benefits. A revised estimate of cost avoidance attributable to IV-D collections is made by multiplying the TRIM3 "cost avoidance ratio" - the average reduction in transfer program benefits per dollar of child support - by the amount of child support distributed according to OCSE financial data.

Calculating the Reduction in Transfer Program Benefits

TRIM3 estimates the total reduction in transfer program benefits attributable to child support by subtracting total program benefits under a "baseline" simulation, in which child support is paid, from total program benefits under an "alternative" simulation, in which no child support is paid. Eligibility and benefits are calculated using the state and federal program rules that were in effect for the majority of calendar year 1999.

The child support income used in the baseline simulation is the child support income reported on the CPS, except that here it is treated as the income of the child rather than of the adult who reports it.[8] This modification is necessary in order to correctly simulate the treatment of child support in the SSI and TANF programs.[9] Child support income is set to zero for a small number of households where child support is reported, but there are no children.[10]

For each simulation, TRIM3 performs the same basic steps a caseworker would perform in determining eligibility and establishing the benefit level, using the rules in effect in the household's state of residence. The model first divides the household into relevant "filing units" - the group or groups of people in the household who would file together as a unit when applying for assistance. TRIM3 then determines whether members of the filing unit meet any relevant eligibility requirements, such as those related to age, disability, immigrant and citizenship status, and work status. Income and assets tests are conducted in order to determine whether the unit meets financial eligibility requirements. If so, a benefit amount is calculated.

Not all units eligible for assistance choose to participate, so TRIM3 must select participants from among those found eligible. Eligible units that report receiving assistance from a particular program on the CPS are automatically simulated to participate in the baseline simulation. Additional participants are selected, as needed, in such a way that the size and composition of the simulated baseline caseload matches that of the actual caseload as closely as possible.[11]

For the alternative simulation, child support is set to zero, and eligibility and benefits are recalculated. Units are allowed to change their participation decision in response to a change in their eligible benefit level. For example, a unit may have been eligible for a very low food stamp benefit in the baseline, and so have chosen not to participate. Without child support income, the unit becomes eligible for a higher food stamp benefit, and decides to participate. If a unit starts participating in one program, then that may increase the likelihood of participating in other programs. For example, most SSI and TANF recipients participate in the Food Stamp Program. So most units that start participating in SSI or TANF in the alternative simulation are simulated to participate in the Food Stamp Program as well.

Identifying Benefit Reductions Attributable to IV-D Collections

The CPS does not indicate whether a child support recipient receives IV-D services. Therefore, this information must be imputed onto the file using data from the 1998 CPS April/March Matched File (CPS-CSS).[12] Total child support payments and benefit reductions can then be calculated separately for IV-D and non-IV-D families.

The CPS-CSS collects detailed child support information from custodial mothers and custodial fathers. There is no specific question about whether a custodial parent receives IV-D services. However, IV-D status can be inferred using an approach developed by Matthew Lyon (1999). A custodial parent is counted as receiving IV-D services if anyone in the family receives TANF or Medicaid. Those without TANF or Medicaid are counted as IV-D if they respond that they have had contact with the government regarding child support, do not know whether they are supposed to receive child support because "a child support agency filled out the paper work", or were supposed to receive child support through a child support, welfare, or other public agency.

TRIM3 assumes that all families receiving TANF or Medicaid in the baseline simulation receive IV-D services. A logit model is used to impute the IV-D status of mothers not in families receiving TANF or Medicaid. The model controls for the mother's demographic, income, and child support characteristics. Fathers who are not in families receiving TANF or Medicaid are flagged as receiving IV-D services based on probabilities that vary by poverty level. Because the CPS-CSS does not contain information on nonparent caretakers, the only nonparent caretakers flagged as receiving IV-D services are those in families receiving TANF or Medicaid. See Appendix A for additional information concerning the imputation of IV-D status.

Adjusting for Underreporting of Child Support Income

Because there is no external target for total child support income that captures private payments as well as IV-D collections, it is impossible to know the full extent of underreporting of child support income on the CPS. However, there is little doubt that child support is underreported. With the exception of income from wages and salaries, income amounts are typically underreported in household survey data. The CPS captured an estimated 71 percent of property income, 88 percent of transfer income, and 77 percent of pension income in 1996 (Roemer 2000). Wheaton (1997) estimates that child support on the March 1992 CPS is underreported by as much as 21 to 27 percent relative to child support reported on the 1992 CPS-CSS.

Child support is particularly likely to be underreported for TANF recipients, who are unlikely to report child support payments retained by the government. Therefore, unadjusted TRIM3 estimates will fall far short of the actual level of TANF cost recovery. To the extent that persons who do not receive TANF underreport child support, unadjusted TRIM3 estimates of cost avoidance will be understated as well.

The results presented in this report correct for the problem of underreporting by adjusting the TRIM3 estimates using OCSE financial data obtained from table 1 of the Child Support Enforcement FY 1999 Preliminary Data Report (OCSE 2000).[13] OCSE data on TANF cost recovery are substituted for the unadjusted TRIM3 estimates. Revised IV-D cost avoidance estimates are calculated by multiplying the TRIM3 cost avoidance ratio (average reduction in transfer program benefits per dollar of child support) by the total amount of payments to families as reported in the OCSE financial data. No revised estimate is possible for persons not receiving IV-D services, since there are no comparable data for the total amount of child support paid privately.

Results

Results from each step in the cost avoidance estimation process are displayed below. First, cost avoidance ratios are estimated for non-TANF families. Next, a special simulation is performed to estimate the cost avoidance ratios that result when child support payments are made to TANF families. Finally, the cost avoidance ratios for non-TANF and TANF families are applied to OCSE financial data in order to correct for the underreporting of child support on the CPS.

Cost Avoidance Ratios for Non-TANF Families

Using data from the CPS, TRIM3 estimates that $19.9 billion dollars in child support were paid to non-TANF families in 1999, of which $10.3 billion were paid through the IV-D system (table 1).[14] The remaining $9.6 billion were paid through private transactions. The TRIM3 estimate of child support paid to non-TANF IV-D families is lower than the actual amount reported by OCSE ($13.4 billion) due to the underreporting of child support income on the CPS.[15]

Child support collected by the IV-D system results in an estimated $1,950 million in reduced means-tested benefits for families receiving support, and private child support transactions result in an additional $400 million benefit reduction. These results do not correct for the underreporting of child support on the CPS. The overall cost avoidance ratio - average reduction in transfer program benefits per dollar of child support - is .190 for IV-D non-TANF families. That is, each $1 received by a IV-D non-TANF family results in an average benefit reduction of $0.19 over the five programs evaluated. The cost avoidance ratio is .042 for families receiving child support through private transactions. It is not surprising that the cost avoidance ratio is lower for families receiving child support through private transactions, since they tend to have higher incomes and less interaction with social assistance programs. The overall cost avoidance ratio for all non-TANF families is .118.

Table 1: TRIM3 Cost Avoidance Estimates for Non-TANF Families, 1999[a]
  Child Support Collection Method
IV-D Private Transaction Total
Dollars (Millions) CA Ratio[b] Dollars (Millions) CA Ratio[b] Dollars (Millions) A Ratio[b]
Child Support Benefit Reductions[c] 10,265   9,599   19,864  
TANF 561 .055 56 .006 616 .031
SSI 101 .010 20 .002 121 .006
Housing 98 .010 22 .002 120 .006
Food Stamps 738 .072 95 .010 833 .042
Medicaid 452 .044 209 .022 661 .033
Total 1950 .190 401 .042 2,351 .118
Source: TRIM3 Microsimulation Model using data from the March 2000 CPS.
[a] Note: Dollars reflect underreporting of child support on the CPS).
[b] The cost avoidance ratio is equal to transfer program benefit reductions attributable to child support, divided by the total amount of child support income.
[c] Benefit reductions attributable to child support income.

TANF

Cost avoidance to the TANF program occurs when child support income renders a family ineligible for assistance or makes an eligible family decide not to participate. TRIM3 estimates that, on average, a dollar in child support paid to a non-TANF IV-D family results in 5.5 cents in cost avoidance to the TANF program. A dollar of privately paid child support results in an average 0.6 cents in cost avoidance to the TANF program. For all families, a dollar of child support results in an average 3.1 cents in TANF cost avoidance. Given the TANF block grant structure, a reduction in TANF assistance does not directly decrease federal and state spending on TANF-related programs but rather frees up resources for training programs, tax credits, and other support services.

SSI

The SSI program treats child support income as the income of the child, and so child support income affects the eligibility and benefits of disabled children, but not of disabled or aged adults. Federal rules disregard 33 percent of the child's child support income in determining eligibility and the benefit amount.

Cost avoidance to the SSI program occurs when child support income lowers the benefit amount for a participating child, renders an otherwise eligible child ineligible, or is sufficient to cause an eligible child to decide not to participate. Due to limited disability information for children on the CPS, TRIM3 only captures the first source of SSI cost avoidance - a lower benefit resulting from receipt of child support income.[16] Furthermore, cost avoidance arising from reductions in state supplementary SSI payments is not captured. Each dollar of child support collected by the IV-D system on behalf of non-TANF families results in one cent, on average, in SSI benefit reductions. Privately paid child support results in an average of 0.2 cents in SSI benefit reductions per dollar of child support. For all families, a dollar of child support results in an average 0.6 cents in SSI benefit reductions. Since SSI is an entitlement, cost avoidance in the SSI program produces direct program savings.

Housing

TRIM3 simulates the value of rental subsidies for CPS-reported public housing tenants and subsidized renters. This includes households receiving either project-based assistance (public housing and Section 8 project-based) or household-based assistance (Section 8 certificates and vouchers). Cost avoidance arises when child support income increases the household's rental payments - which are typically set at the larger of 10 percent of gross income or 30 percent of net income. In addition, child support might render some households ineligible for assistance. However, since housing assistance is not an entitlement and is often of limited availability, this is not a key source of cost avoidance. TRIM3 captures the subsidy reduction that occurs when child support income increases a household's rental payments, but not when child support renders a household ineligible for assistance.

Child support collected by the IV-D system on behalf of non-TANF families results in an average of one cent in increased rental payments per dollar of child support collected. Privately paid child support results in an average of 0.2 cents in increased rental payments per dollar of child support. For all families, a dollar of child support results in an average of 0.6 cents in increased rental payments per dollar of child support collected. Although the federal housing programs are discretionary programs, cost avoidance may indirectly reduce federal spending on housing assistanceand help localhousing authorities cover the operating costsassociated withpublic housing.[17]

Food Stamp Program

Cost avoidance to the Food Stamp Program occurs when child support lowers the amount of benefit a participating household receives, renders an otherwise eligible household ineligible, or is sufficient to cause an eligible household to decide not to participate. TRIM3 captures each of these sources of cost avoidance. Cost avoidance is offset to a certain extent because households that pay child support are able to deduct the child support payment from income when calculating the benefit amount. Because TRIM3 does not model the child support deduction, the effect of the deduction is not captured in the estimates presented in table 1. However, an externally derived estimate of the value of the child support deduction is incorporated into the final results presented in table 4.

Child support collected by the IV-D system on behalf of non-TANF families results in an average of 7.2 cents of Food Stamp Program benefit reductions per dollar of child support collected. Privately paid child support results in an average of one cent in benefit reduction per dollar of child support. For all families, a dollar of child support results in an average of 4.2 cents in Food Stamp Program benefit reductions. Since food stamps are an entitlement, cost avoidance in the Food Stamp Program produces direct program savings.

Medicaid

The child support system provides two sources of cost avoidance to the Medicaid program. Health insurance and/or medical support payments provided by the noncustodial parent may eliminate the need for Medicaid coverage or reduce the costs of providing coverage. Child support income may render a person financially ineligible for Medicaid assistance, or may be sufficient to cause the person to decide not to enroll.

TRIM3 captures cost avoidance that arises when child support renders a person ineligible for assistance. However, no cost avoidance is counted for children who are eligible for SCHIP in the baseline but are simulated to receive Medicaid in the alternative simulation.[18] Cost avoidance from enrollment decisions is captured to a certain extent. If child support is sufficient to keep a person from receiving SSI or TANF, and as a result, the person chooses not to enroll in Medicaid, then the resulting Medicaid savings are counted as cost avoidance.

Savings to the Medicaid program are estimated by multiplying the average monthly Medicaid insurance value by the number of months during the year in which child support income keeps a person from enrolling in Medicaid. The average monthly insurance value varies by state and by whether the person is disabled, elderly, a non-disabled child, or a non-disabled and non-elderly adult.[19]

Child support collected by the IV-D system on behalf of non-TANF families results in an average of 4.4 cents of savings to the Medicaid program per dollar of child support collected. Privately paid child support results in an average of 2.2 cents in savings to the Medicaid program per dollar of child support. For all families, a dollar of child support results in an average of 3.3 cents in savings to the Medicaid program. Since Medicaid is an entitlement, cost avoidance produces direct program savings.

Variation in Cost Avoidance by Income Level and IV-D Status

As shown above, cost avoidance arising from IV-D child support collections is much greater than that arising from private child support payments. This is to be expected, since families receiving IV-D services tend to have lower incomes than those who do not (Lyon 1999) and cost avoidance ratios tend to be higher for lower-income families (Wheaton and Sorensen 1998).

Table 2 shows the total cost avoidance ratio (summed over all programs), by poverty level, for non-TANF families.[20]Cost avoidance ratios fall dramatically as family income relative to the poverty threshold rises. The ratio drops from a high of .68 for poor families, to a low of .011 for families with incomes greater than 300 percent of the poverty threshold. This is to be expected, since eligibility and benefits for means-tested transfer programs are targeted to lower-income families.

Table 2: Cost Avoidance Ratios for Non-TANF Families by Poverty Level, 1999
Poverty Level Child Support Collection Method
IV-D Private Transaction Total
< 100% poverty .644 NA[a] .680
100 - < 150% .433 .211 .411
150 - < 200% .170 .093 .149
200 - < 300% .076 .046 .061
300% + .024 .003 .011
Total .190 .042 .118
Source: TRIM3 Microsimulation Model using data from the March 2000 CPS.
[a] Sample size is not sufficient for a reliable estimate.

The cost avoidance ratio for privately paid child support is lower than the IV-D cost avoidance ratio for families of the same income level. A lower cost avoidance ratio is expected for families who receive support privately, since the fact that a family does not receive IV-D services suggests that it has not received AFDC or TANF in the past and may be less likely than a similar IV-D family to seek government assistance.

Because cost avoidance ratios are higher at lower income levels, the overall cost avoidance ratio will be higher to the extent that a greater proportion of the child support is collected on behalf of low-income families. As can be seen from table 3, a major explanation for the higher overall cost avoidance ratio experienced by the IV-D system is that 41 percent of total child support dollars collected by the IV-D system for non-TANF families is paid to families with incomes under 200 percent of the poverty threshold. In contrast, only 10 percent of privately paid child support goes to families with incomes that low.

Table 3: Child Support for Non-TANF Families, by Income Level, 1999[a]
Poverty Level Child Support Collection Method
IV-D Private Transaction Total
Dollars (Millions) % Dollars (Millions) % Dollars (Millions) %
< 100% poverty 1,180 11% 210 2% 1,390 7%
100 - < 150% 1,534 15% 173 2% 1,707 9%
150 - < 200% 1,539 15% 562 6% 2,101 11%
200 - < 300% 2,296 22% 2,268 24% 4,564 23%
300% + 3,716 36% 6,385 67% 10,101 51%
Total 10,265 100% 9,599 100% 19,864 100%
Source: TRIM3 Microsimulation Model using data from the March 2000 CPS.
[a]Dollars are not corrected for underreporting of child support on the CPS.

Cost Avoidance for TANF Families

The vast majority of child support income collected on behalf of families currently receiving TANF is retained by the government to help offset TANF expenditures. However, a small amount is paid to families, and this in turn produces cost avoidance to other programs.

Under AFDC rules in effect until 1997, states were required to "pass through" to the family the first 50 dollars in child support collected in each month. This was discontinued under TANF, although states were permitted to continue the pass-through practice so long as they financed it themselves. Federal law allowed two states that had pass-through related waivers under the AFDC program to continue the practice under TANF. Under these waivers, Connecticut passes through up to $100 per month of child support, and Wisconsin passes through the full amount of the child support collection.[21]

The OCSE financial data count child support pass-through amounts in the two waiver states as "payments to families," but not the optional pass-through amounts financed by the states. Other types of child support payments counted as "payments to families" include collection overages, "fill the gap" payments in states that use child support to fill the gap between the state's maximum benefit and the income eligibility standard, and collections made for TANF families on behalf of children outside the TANF assistance unit (such as an SSI child).

In order to maintain consistency with the OCSE financial data, only those child support collections that meet the OCSE definition of "payments to families" are assumed to generate cost avoidance. The optional pass-through amounts financed by the states are counted as "cost recovery" because this is consistent with how the dollars are classified in the OCSE financial data.

Unfortunately, the CPS sample size is too small to generate reliable cost avoidance ratios for the waiver and fill-the-gap states, and there is no way to directly simulate the cost avoidance that arises from payments to families for collection overages. In order to provide enough data for a rough cost avoidance estimate for all TANF families, a special simulation is performed in which all states are simulated to have at least a fifty-dollar pass-through. Cost avoidance ratios are calculated by comparing the results from this simulation to the alternative simulation in which no child support is paid. The cost avoidance ratio is set equal to the change in transfer program benefits (for TANF families) divided by the amount of child support paid to TANF families (net of what is retained by the government). While this is not a real-world policy scenario, it provides enough data to generate a rough estimate of the cost avoidance that arises when child support is paid directly to TANF families.

The simulation produces a total cost avoidance ratio of .402 for child support payments to TANF families. The ratio for SSI is .177 (reflecting the child support collected on behalf of children outside the TANF assistance unit), and the ratios for housing assistance and food stamps are .074 and .151, respectively. The simulation does not yield any cost avoidance for Medicaid, since it is assumed that a person receiving TANF will receive Medicaid regardless of the amount of child support paid to the family.

IV-D Cost Avoidance

The total cost avoidance dollar estimates presented in table 1 understate the true extent of cost avoidance because they are derived from CPS data in which child support income is underreported. To the extent that child support income is underreported on the CPS, benefit reductions resulting from child support will also be understated. To correct for this problem, revised IV-D cost avoidance estimates are calculated by multiplying the TRIM3 cost avoidance ratio (average reduction in transfer program benefits per dollar of child support) by the total amount of payments to families as reported in the OCSE financial data. No revised estimate is possible for persons not receiving IV-D services, since there are no comparable data for the total amount of child support paid privately.

Table 4 shows the total estimated cost avoidance produced when TRIM3 cost avoidance ratios are multiplied by the actual level of distributed child support as reported in the OCSE financial data (OCSE 2000). Cost avoidance from IV-D child support distributions amounts to an estimated $2,555 million for FY 1999, almost entirely from families not currently receiving TANF.

The child support payments in table 4 are obtained from table 1 of the Child Support Enforcement FY 1999 Preliminary Data Report (OCSE 2000). The $113 million reported for TANF families is the total amount distributed to TANF families (IV-A) and children receiving foster care maintenance services under title IV-E of the Social Security Act. TRIM3 does not simulate the IV-E program and so makes the assumption that the cost avoidance ratio for payments to IV-E children is the same as for payments to TANF families.

The $13,362 million reported for non-TANF families represents the sum of "medical support" and "payments to families" for "former assistance" and "never assistance" families. TRIM3 does not distinguish between families that formerly received AFDC or TANF, and those that never received assistance. Therefore, separate cost avoidance estimates for former assistance and never assistance families are not possible.

"Medical support" payments represent cash payments made by the noncustodial parent for the purpose of helping defray the child's medical expenses. The $21 million in medical support paid to TANF and IV-E families is assumed to go primarily to the Medicaid program to help offset Medicaid expenditures, and so is not included in the cost avoidance estimate shown in table 4. However, the $13,362 million in payments to non-TANF families includes about $75 million in medical support payments, which are assumed to be treated the same as child support income in calculating eligibility and benefits under the various transfer programs.[22]

Table 4: IV-D Cost Avoidance, FY 1999
  TANF Families Non-TANF IV-D  
CA Ratio[a] Dollars (Millions) CA Ratio[b] Dollars (Millions) Total IV-D (Millions)
Child Support Payments Distributed to Families[c] na 113 na 13,362 13,475
Cost Avoidance[d]
TANF na na .055 735 735
SSI .177 20 .01 134 154
Housing .074 8 .01 134 142
Food Stamp Program .151 17 .072 962 979
Net CS Deduction[e] na na na na 936
Medicaid 0 0 .044 588 588
Total na na na na 2,555
Source: Cost avoidance ratios are estimated by the TRIM3 microsimulation model using data from the March 2000 CPS. Child support payments are obtained from Child Support Enforcement FY 1999 Preliminary Data Report. U.S. Department of Health and Human Services. The effect of the Food Stamp Program's child support deduction is estimated using data from Characteristics of Food Stamp Households, Fiscal Year 1999, by Randy Rosso and Lisa Fowler.
[a] Cost avoidance ratio (dollars of benefit reductions per dollar of child support) estimated for calendar year 1999 using TRIM3.
[b] Cost avoidance ratio (dollars of benefit reductions per dollar of child support) estimated for calendar year 1999 using TRIM3.
[c] From the FY 1999 Preliminary Data Report: the value for TANF families is obtained from total "payments to families" as reported for current assistance IV-A and IV-E families; the value for non-TANF families is the sum of "medical support" and "payments to families" for former assistance and never assistance families.
[d] TRIM3 cost avoidance ratio, by program, and final estimated cost avoidance amount calculated by multiplying the TRIM3 cost avoidance ratio by total child support payments to families.
[e] Noncustodial parents who pay child support are allowed to deduct child support from their net income in determining their food stamp benefit, offsetting benefit reductions from cost avoidance. Table 4 assumes that all noncustodial parents receiving the deduction pay child support through the IV-D system, and that the deduction increases their benefit by 30 percent of the amount of child support paid.

The greatest source of IV-D cost avoidance is to the Food Stamp Program, with an estimated $936 million dollars of benefit reductions in FY 1999. Cost avoidance would be even higher ($979 million) if it were not for the child support deduction available to noncustodial parents who pay child support. TRIM3 does not simulate the food stamp child support deduction, so a rough approximation is made for the purpose of table 4. All persons claiming the child support deduction are assumed to pay child support through the IV-D system.[23] The aggregate amount of the deduction is estimated using data from the FY 1999 Food Stamp Program Quality Control Sample (Rosso and Fowler 2000), and the deduction is assumed to increase the noncustodial parent's food stamp benefit by 30 percent of the deducted amount.[24] Since the Food Stamp Program is a federally funded entitlement program, the $936 million in cost avoidance reflects direct savings to the federal government.

IV-D child support collections reduce TANF benefits by an estimated $735 million by reducing the number of families eligible for or participating in TANF. Given the TANF block grant structure, a reduction in TANF assistance does not directly decrease federal or state spending on TANF-related programs but rather frees up resources for training programs, job search, transportation, child care, and other support services.

Medicaid cost avoidance arising from changes in eligibility and enrollment amounts to an estimated $588 million. As mentioned previously, no cost avoidance is attributed to children who are eligible for SCHIP but would receive Medicaid in the absence of child support. The Medicaid estimate does not include savings from enforcement of medical support orders. Since Medicaid is a federal-state matching entitlement program, the $588 million in cost avoidance reflects shared savings to federal and state governments.

IV-D child support collections reduce the SSI benefits of disabled children by an estimated $154 million. As mentioned previously, this estimate only captures the benefit reduction for participating children. It does not capture the savings that occur when child support renders a child ineligible for SSI or is sufficient to cause an eligible family to decide not to participate. Furthermore, cost avoidance arising from reductions in state supplementary SSI payments is not captured. SSI is a federally funded entitlement program, so the $154 million in cost avoidance reflects direct savings to the federal government.

Federal housing assistance programs account for the remaining $142 million of cost avoidance from IV-D child support collections. Although the federal housing programs are discretionary programs, cost avoidance may indirectly reduce federal spending on housing assistanceand help localhousing authorities cover the operating costsassociated withpublic housing.

IV-D Cost Avoidance And Cost Recovery

The financial benefits to government of the CSE Program have traditionally been measured by the extent of cost recovery - the total amount of child support retained by the government to offset current and past TANF and AFDC expenditures. In FY 1999, cost recovery equaled $2,347 million dollars (OCSE 2000). Of this amount, $1,346 million were from collections made on behalf of current TANF and IV-E recipients and the remaining $1,002 million were from collections made on behalf of former recipients. In addition, most of the $21 million in medical support payments made on behalf of current TANF and IV-E recipients can be assumed to be transmitted to the Medicaid program for the purpose of offsetting the medical expenses of Medicaid enrollees. Therefore, this $21 million is included in the "Cost Recovery" row shown in table 5, bringing the total amount up to $2,368 million.

Table 5 shows the percentage break down of cost recovery and cost avoidance resulting from IV-D child support collections in FY 1999. Of the total estimated $4,923 million in cost recovery and cost avoidance, less than half (48 percent) results from TANF cost recovery and direct reimbursement of the Medicaid program. The remaining $2,555 million results from cost avoidance, of which 37 percent is for the Food Stamp Program, 29 percent is for TANF, 23 percent is for Medicaid, and the remaining 12 percent is split between SSI and the federal housing assistance programs.

Table 5: IV-D Cost Recovery and Cost Avoidance, FY 1999
  Dollars (Millions) Percent of Total Cost Recovery + Cost Avoidance Percent of Cost Avoidance Dollars
Cost Recovery[a] 2,368 48% na
Cost Avoidance 2,555 52% na
TANF 735 15% 29%
SSI 154 3% 6%
Housing 142 3% 6%
Food Stamp Program 936 19% 37%
Medicaid 588 12% 23%
Source: TRIM3 Microsimulation Model using data from March 2000 CPS. FY 1999 OCSE Financial data.
[a] Includes $1.3 billion for current assistance IV-A and IV-E, $1.0 billion for former assistance families, and $21 million for medical support to current assistance IV-A and IV-E.

Conclusion

Cost avoidance reflects an increased level of family self-sufficiency. By receiving child support from noncustodial parents, families are less reliant on assistance from means-tested transfer programs. For some families, child support, when combined with other family income, is sufficient to eliminate the need for assistance entirely.

Cost avoidance generated by IV-D child support collections and private child support transactions represents direct budgetary reductions in federal and state outlays for entitlement programs such as SSI, the Food Stamp Program, and Medicaid. For block grant programs, such as TANF, and discretionary programs, such as federal housing assistance, cost avoidance frees funds for use in serving additional families and/or providing additional services, and may eventually result in lower federal expenditures.

Cost avoidance amounted to at least $2.6 billion in FY 1999, exceeding the total amount of cost recovery in that year. To the extent that the number of families receiving TANF continues to decline, cost avoidance can be expected to increase in size relative to cost recovery. The President's budget for FY 2003 proposes legislation that would allow current and former welfare recipients to retain more of the child support paid on their behalf (OMB 2002). If implemented, these provisions can be expected to reduce cost recovery, while increasing cost avoidance.

It is important to note that the estimates presented here show the costs avoided as a result of IV-D collections, and cannot be entirely attributed to IV-D enforcement activities. Undoubtedly, some percentage of IV-D child support collections would be paid even in the absence of the IV-D system. It is beyond the scope of this report to estimate the amount of child support collected and costs avoided solely as a result of IV-D enforcement efforts. However, the IV-D cost avoidance estimates presented here are consistent with estimates that attribute all cost recovery to the IV-D system (Committee on Ways and Means 2000).

Unlike cost recovery, cost avoidance can only be estimated. It cannot be directly measured. But when even a lower-bound estimate of cost avoidance exceeds the total amount of cost recovery, it is clear that cost avoidance is an important part of the picture. As the child support enforcement community calls attention to child support's ability to families' financial stability and independence, it is worth recognizing that this increased independence also implies financial benefits to government through cost avoidance.

References

Barnow, Burt S., Timothy M. Dall, Mark W. Nowak, and Barbara E. Dannhausen (April 2000). The Potential of the Child Support Enforcement Program to Avoid Costs to Public Programs: A Review and Synthesis of the Literature. Final report prepared for the U.S. Department of Health & Human Services. Contract No. HHS-100-97-0007.

Committee on Ways and Means (2000). 2000 Green Book. Washington, D.C.: U.S. House of Representatives.

Formoso, Carl. (February 1999). The Effect of Child Support and Self-Sufficiency Programs on Reducing Direct Support Public Costs. Olympia, WA: Washington State Division of Child Support.

Garasky, Steven, Shao-Hsun Keng, and Helen H. Jensen (1999). Child Support Enforcement Cost Avoidance: Evidence from Iowa. Report prepared for The Lewin Group.

General Accounting Office (December 1998). WelfareReform: Effect on HUD's Housing Subsidies Is Difficult to Estimate (GAO/RCED-99-14). Washington, D.C.

Koball, Heather, Laura Wheaton, and Elaine Sorensen (2002). Final Report. Final report prepared under U.S. Department of Health & Human Services Administration for Children and Families IDIQ Contract No. 105-00-8302, Task Order No. 14. Washington, D.C.: The Urban Institute.

Lerman, Robert I. and Elaine Sorensen (April 2001). Child Support: Interaction Between Private and Public Transfers. NBER Working Paper No. w8199. Cambridge, Massachusetts: National Bureau of Economic Research, Inc.

Lyon, Matthew (May 1999). Characteristics of Families Using Title IV-D Services in 1995. U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation.

Office of Child Support Enforcement (1998). Twenty-Third Annual Report To Congress. U.S. Department of Health and Human Services.

Office of Child Support Enforcement (September 2000). Child Support Enforcement FY 1999 Preliminary Data Report. U.S. Department of Health and Human Services.

Office of Child Support Enforcement (July 2001). FY2000 Preliminary Data Preview Report. U.S. Department of Health and Human Services.

Office of Management and Budget (2002). Budget of the United States Government, Fiscal Year 2003. Washington, D.C.

Roberts, Paula (October 2000). Child Support Distribution and Disbursement. Washington, D.C.: Center for Law and Social Policy.

Roberts, Paula (Spring 2002). The Importance of Child Support Enforcement: What Recent Social Science Research Tells Us. Washington, D.C.: Center for Law and Social Policy.

Roemer, Marc I. (June 2000). Assessing the Quality of the March Current Population Survey and the Survey of Income and Program Participation Income Estimates, 1990-1996. U.S. Census Bureau, Income Surveys Branch, Housing and Household Economic Statistics Division.

Rosso, Randy and Lisa Fowler (December 2000). Characteristics of Food Stamp Households, Fiscal Year 1999. Food Stamp Program Report No. FSP-00-CHAR, Food and Nutrition Service, U.S. Department of Agriculture. Washington, D.C.: Mathematica Policy Research, Inc.

Sard, Barbara and Will Fischer (May 2003). Housing Voucher Block Grant Bills Would Jeopardize An Effective Program And Likely Lead To Cuts In Assistance For Low-Income Families. Washington, D.C.: Center on Budget And Policy Priorities.

Sorensen, Elaine and Kate Pomper (2002). Assessing the Causal Relationship Between Child Support and Visitation. Working Paper. Washington, D.C.: The Urban Institute.

Tapogna, John (January 2002). Issues Related to Cost Avoidance and TRIM. Memorandum prepared under U.S. Department of Health & Human Services Administration for Children and Families IDIQ Contract No. 105-00-8303, Task Order No. 5. Portland, Oregon: ECONorthwest.

Wheaton, Laura (1997). "Child Support Underreporting: A Tale of Two Supplements." American Statistical Association 1997 Proceedings of the Section on Government Statistics and Section on Social Statistics, p. 370-375.

Wheaton, Laura and Elaine Sorensen (1998). "Reducing Welfare Costs and Dependency: How Much Bang for the Child Support Buck?" The Georgetown Public Policy Review. 4(1), 23-27.

Appendix A.

Imputing IV-D Status

The CPS does not indicate whether a child support recipient receives IV-D services. Therefore, this information is imputed onto the file. The 1998 CPS April/March Matched file provides the data used to develop the imputation. The CPS-CSS collects detailed child support information from custodial mothers and custodial fathers. There is no specific question about whether a custodial parent receives CSE services. However, CSE status can be inferred using an approach developed by Matthew Lyon (1999).

Lyon counts a family as receiving CSE services if anyone in the family receives TANF or Medicaid. Families without TANF or Medicaid are counted as IV-D if they respond they have had contact with the government regarding child support, do not know whether they are supposed to receive child support because the child support agency had filled out the paper work, or were supposed to receive child support through a child support, welfare, or other public agency.

Two modifications were made to the Lyon methodology for the purpose of developing the imputation of IV-D status. First, TRIM3-simulated TANF and Medicaid were merged onto the CPS-CSS data and used in place of reported TANF and Medicaid in identifying IV-D recipients. This was done for consistency with the TRIM3 simulation, where families simulated to have TANF or Medicaid are automatically assumed to receive IV-D services. The second modification was the addition of a variable indicating whether child support was actually received through a child support, welfare, or other public agency. If so, then the parent was counted as IV-D.

Once IV-D status had been determined, an imputation methodology was developed to predict, based on variables available in TRIM3, whether or not a parent received IV-D services. Due to data limitations, different imputation methodologies are used for custodial mothers, custodial fathers, and non-parent caretakers.

Custodial Mothers

If anyone in a custodial mother's family is simulated to receive TANF or Medicaid in the baseline simulation, then the custodial mother is assumed to receive

IV-D services. Otherwise, the mother's IV-D status is imputed using a logit model developed using unallocated observations from the CPS-CSS. The imputation is performed for child support recipients only, since TRIM3 does not need to know the IV-D status of nonrecipients in order to simulate cost avoidance.

The logit model imputes the log of the odds ratio that a custodial mother (who receives child support and is not in a family with TANF or Medicaid) receives IV-D services. The log of the odds ratio is converted to a probability, which is then compared to a uniform random number. If the mother's random number is less than or equal to the imputed probability, then she is flagged as receiving IV-D services.

The logit model controls for a variety of demographic and child support characteristics. Demographic characteristics include the mother's age, marital status, race and ethnicity, region of residence, educational attainment, urban/rural status, poverty level, number of children with an absent parent, age of oldest child with an absent parent, and household receipt of food stamps or housing assistance. Child support characteristics include whether or not child support was due under a formal arrangement, whether or not the full amount due was received, and the level of child support income relative to other family income.

Most of the explanatory variables are available on the raw CPS. However, three variables are imputed by the TRIM3 child support module using data from the CPS-CSS. The imputed variables include the number of children with an absent parent, whether child support was due under a formal arrangement, and whether the full amount due was received.

Table 6 displays the logit model results. The fit of the model is very good. The model accurately predicts 78 percent of the observed responses. The model Chi-Square has a significance of .0001. All of the explanatory variables related to child support are statistically significant at the ten percent level, as well as some of the explanatory variables related to age, race, region of residence, and poverty level.

Custodial mothers who receive child support despite not being due support under a formal arrangement are 27 percent as likely to be receiving IV-D services as are those who receive child support under a formal arrangement. This is not a surprising result, since one of the first things that a CSE agency would attempt to do is to establish a formal child support order.

The direction of the coefficients on the remaining child support variables is consistent with a hypothesis that more difficult child support cases gravitate toward the IV-D system. Mothers who receive less than the full amount of support due are 4 times more likely to be receiving IV-D services than are those who receive the full amount. Those whose child support income is less than 30 percent of their other family income (excluding TANF and SSI) are 1.8 times more likely to be receiving IV-D services than are those whose child support income exceeds 30 percent of their other family income.

Women with family incomes between 100 and 200 percent of the poverty threshold are about two and a half times more likely to be receiving IV-D services than are those with family incomes above 500 percent of the poverty threshold. The dummy variable for poor women is not statistically significant. Since the model excludes TANF and Medicaid recipients, the lack of a significant effect for poor women is consistent with a hypothesis that poor women who do not receive TANF or Medicaid tend to be averse to government assistance and so are not any more likely to seek government help in obtaining child support than are higher income women.

Finally, women who are under age 25 and are receiving child support are 3.2 times more likely to be receiving IV-D services than are women aged 45 or above. African-American women receiving child support are nearly twice as likely to be receiving IV-D services than are other non-Hispanic women. Women in the Midwest and West who receive child support are more likely to be receiving IV-D services than are women in the Northeast. The dummy variables for divorced and separated status are nearly significant, with each decreasing the likelihood of receiving IV-D services relative to married women.

Table 6: Logit Model: Does Custodial Mother Receive IV-D Services? (Child Support Recipient Mothers in Families Without TANF or Medicaid)
  Parameter Estimate Standard Error Wald Chi-Square Pr > Chi-Square Odds Ratio
Intercept -1.5278 .6262 5.9527 .0147  
Child Support Characteristics
Not Due Child Support -1.3225 .3541 13.9494 .0002 .266
Receive Less Than Due 1.4022 .2270 38.1430 .0001 4.064
Child Support / Other Family Income (excluding SSI and TANF)
< .05 .6409 .3712 2.9813 .0842 1.898
.05 - < .15 .5888 .3385 3.0261 .0819 1.802
.15 - < .30 .5880 .3562 2.7256 .0987 1.800
Percent of Poverty
< 100% .2398 .6358 .1423 .7060 1.271
100% - < 150% 1.0248 .4762 4.6309 .0314 2.786
150% - < 200% .9137 .4060 5.0650 .0244 2.493
200% - < 300% .2040 .3124 .4261 .5139 1.226
300% - < 500% .2923 .2838 1.0604 .3031 1.339
Age of Mother
15 - 24 1.1618 .6477 3.2175 .0729 3.196
25 - 29 .5232 .4790 1.1934 .2746 1.688
30 - 34 .5656 .3394 2.7776 .0956 1.760
35 - 39 .2027 .2975 .4644 .4956 1.225
40 - 44 -.2271 .2804 .6559 .4180 .797
Race and Ethnicity
Hispanic .2327 .3791 .3768 .5393 1.262
African-American .6665 .3495 3.6363 .0565 1.947
Region
Midwest 1.2035 .2779 18.7559 .0001 3.332
South .2247 .2592 .7519 .3859 1.252
West .4559 .2741 2.7666 .0963 1.578
Marital Status
Divorced -.3455 .2246 2.3669 .1239 .708
Separated -.6028 .3754 2.5792 .1083 .547
Never Married -.0529 .4036 .0172 .8957 .948
Educational Attainment
Less than High School .2280 .4499 .2568 .6123 1.256
Some College .2523 .2117 1.4210 .2332 1.287
Bachelor Degree + -.1255 .2644 .2253 .6350 .882
Urban/Rural Status
SMSA .0552 .2281 .0585 .8088 1.057
Not Identifiable .0636 .2759 .0531 .8177 1.066
Number of Custodial Children
One -.1987 .3103 .4101 .5219 .820
Two or More -.3034 .3029 1.0033 .3165 .738
Age of Oldest Custodial Child
Under 6 -.5261 .4651 1.2796 .2580 .591
6 to 10 -.1250 .2988 .1750 .6757 .883
11 to 15 -.0985 .2208 .1991 .6555 .906
Household Food Stamps or Housing Assistance .6195 .5705 1.1793 .2775 1.858
N = 705
-2 Log Likelihood: intercept only 977.223
intercept and covariates 793.325
chi-square for covariates 183.897 with 34 DF (p=0.0001)

In order for TRIM3 to produce accurate estimates of the IV-D cost avoidance ratio, it is important that the distribution of recipient custodial mothers imputed as receiving IV-D services, by TANF and Medicaid recipiency and poverty level, is as close as possible to the distribution from the CPS-CSS. The actual number of TRIM3 IV-D child support recipients can be expected to fall short of the number in the CPS-CSS, given the underreporting of child support in the CPS relative to the CPS-CSS (Wheaton 1997). However, this does not present a problem for this analysis, since the problem of underreporting is addressed by applying the TRIM3 cost avoidance ratio to OCSE data concerning the actual level of child support distributions.

As table 7 indicates, the distribution of recipient custodial mothers imputed IV-D status in TRIM3 matches closely the distribution of custodial mother IV-D recipients in the CPS-CSS. The percentage of IV-D recipient mothers in families receiving TANF or Medicaid is somewhat lower in the TRIM3 imputed results, however this may be expected given the decline in welfare roles between 1997 (the year of the income data on the 1998 CPS-CSS) and 1999 (the year of the imputed TRIM3 results). The total number of custodial mother IV-D recipients in the imputed TRIM3 results is only about 76 percent of the number in the 1998 CPS-CSS. This is consistent with prior estimates of the level of child support underreporting on the CPS relative to the CPS-CSS (Wheaton 1997).

Table 7. Distribution of IV-D Child Support Recipient Custodial Mothers, By Poverty Status and Simulated TANF/Medicaid
  1998 CPS-CSS (CY 1997) TRIM3 Imputation (CY 1999)
Total IV-D Child Support Recipients (In Millions) 3,677 2,809
Family Receives TANF or Medicaid[a] 56% 52%
Other Families Below Poverty 3% 2%
Other Families 100% - < 150% Poverty 4% 5%
Other Families 150% - < 200% Poverty 6% 8%
Other Families 200% - < 300% Poverty 11% 11%
Other Families 300% - < 500% Poverty 13% 15%
Other Families 500% Poverty + 6% 7%
Total 100% 100%
Source: 1998 CPS-CSS and TRIM3 microsimulation model using data from the March 1998 and March 2000 CPS.
[a] TANF and Medicaid status is obtained from the TRIM3 1997 baseline simulation (for the 1998 CPS-CSS tabulation) and the 1999 baseline simulation (for the TRIM3 imputed estimates).

Custodial Fathers

If anyone in a custodial father's family is simulated to receive TANF or Medicaid in the baseline simulation, then the custodial father is assumed to receive IV-D services. The sample size for recipient fathers on the CPS-CSS is too small to support an imputation of IV-D status for other recipient fathers, and much of the data for custodial fathers that exists on the CPS-CSS is allocated (Koball et al 2002). The child support topical module in Wave 5 of the 1996 Survey of Income and Program Participation (SIPP) includes more observations for custodial fathers, but fewer of the variables necessary to identify IV-D status.

The results in this report rely on a rough imputation of IV-D status for custodial fathers that was developed using a combination of SIPP and CPS-CSS data. The SIPP was used to calculate the ratio of the percent of fathers receiving IV-D services to the percent of mothers receiving IV-D services, by poverty level (under 200 percent of poverty, 200 to 400 percent of poverty, and 400 percent of poverty and above). The calculation was restricted to recipient mothers and fathers who do not receive TANF or Medicaid, and resulted in ratios of 1.09 (for families under 200 percent of poverty), 0.89 (for families at 200 to 400 percent of poverty), and 0.86 (for families at 400 percent of poverty or more). The ratios were then multiplied by the corresponding percentages of mothers on the CPS-CSS identified as IV-D recipients (57 percent, 49.5 percent, and 44.4 percent respectively) in order to estimate the percentage of fathers who are IV-D recipients (62 percent, 44.3 percent, and 38.3 percent respectively). TRIM3 then assigned IV-D status to any custodial father whose uniform random number was less than or equal to the percent corresponding to his poverty level.

The imputation of IV-D status for custodial fathers is rough, at best. Fathers who receive child support make up a relatively small percentage of all child support recipients. As shown in the sensitivity analysis below, the treatment of fathers makes only a small difference to the overall IV-D cost avoidance ratio.

Nonparent Caretakers

Neither the CPS-CSS nor the SIPP collects detailed child support information for nonparent caretakers. Therefore, TRIM3 makes the simplifying assumption that only those nonparent caretakers in families receiving TANF or Medicaid in the baseline receive IV-D services. Nonparent caretakers make up a very small percentage of all child support recipients, so whether or not they are counted as IV-D has only a small impact on the overall IV-D cost avoidance ratio. This is demonstrated in the sensitivity analysis below.

Sensitivity Analysis for Custodial Fathers and Nonparent Caretakers

The methodology described above produces a cost avoidance ratio estimate of .190 for non-TANF IV-D families. Since nonparent caretakers and fathers make up a relatively small proportion of all child support recipients, even major changes to the assumptions regarding their IV-D status have relatively little effect on the cost avoidance ratio. This is demonstrated through two sensitivity runs - one that treats all nonparent caretakers and fathers as IV-D recipients, and one that excludes them entirely.

If all nonparent caretakers and fathers are simulated as IV-D recipients, then the cost avoidance ratio estimate for non-TANF IV-D families is reduced from .190 to .176. This is the expected direction of change, since the additional caretakers and fathers introduced into the IV-D caseload under this scenario tend to have higher incomes. Higher-income child support recipients are less likely to receive government assistance even in the absence of child support, so including them in the IV-D estimate lowers the cost avoidance ratio.

If no nonparent caretakers or fathers are simulated as IV-D recipients, then the cost avoidance ratio for non-TANF IV-D families increases slightly from .190 to .192. This suggests that the cost avoidance ratio for the nonparent caretakers and fathers who are simulated to be IV-D recipients is lower than the cost avoidance ratio for IV-D recipient mothers, so removing them from the estimate increases the IV-D cost avoidance ratio.

Appendix B. Cost Avoidance Ratios by Program and Poverty Level

Table 8 shows cost avoidance ratios by program and poverty level, for all non-TANF families, IV-D non-TANF families, and families receiving child support through private transactions. As discussed in the main body of the report, only rough estimates of cost avoidance ratios are possible for child support passed through to TANF families: .177 for SSI, .074 for housing assistance, and .151 for the Food Stamp Program. The Food Stamp cost avoidance ratios do not capture the offsetting effects on cost avoidance of the deduction available to households that pay child support.

Table 8: Detailed Cost Avoidance Ratios for Non-TANF Families, 1999
  TANF SSI Housing Food Stamps Medicaid Total
All Non-TANF Families
< 100% Poverty 0.222 0.024 0.041 0.229 0.164 0.680
100% < 150% Poverty 0.094 0.010 0.015 0.199 0.092 0.411
150% < 200% Poverty 0.026 0.017 0.008 0.049 0.050 0.149
200% < 300% Poverty 0.011 0.005 0.004 0.014 0.027 0.061
300%+ Poverty 0.004 0.001 0.000 0.001 0.005 0.011
All Income Levels 0.031 0.006 0.006 0.042 0.033 0.118
IV-D Non-TANF Families
< 100% Poverty 0.238 0.012 0.048 0.242 0.104 0.644
100% < 150% Poverty 0.102 0.012 0.014 0.207 0.098 0.433
150% < 200% Poverty 0.033 0.022 0.009 0.054 0.051 0.170
200% < 300% Poverty 0.018 0.010 0.002 0.020 0.027 0.076
300%+ Poverty 0.008 0.003 0.000 0.001 0.011 0.024
All Income Levels 0.055 0.010 0.010 0.072 0.044 0.190
Private Transaction
< 100% Poverty NA[a] NA NA NA NA NA
100% < 150% Poverty 0.018 0.000 0.029 0.128 0.036 0.211
150% < 200% Poverty 0.007 0.002 0.004 0.033 0.047 0.093
200% < 300% Poverty 0.004 0.000 0.006 0.009 0.027 0.046
300%+ Poverty 0.002 0.000 0.000 0.000 0.001 0.003
All Income Levels 0.006 0.002 0.002 0.010 0.022 0.042
Source: TRIM3 Microsimulation Model using data from the March CPS.
[a] Sample size is not sufficient for a reliable estimate.

 


[1] Rules governing distribution of child support to former welfare recipients are complex, but in general, the government retains child support collected on obligations accrued during the assistance period and on collections obtained through the federal income tax intercept (Roberts 2000).

[2] Under TANF, states were allowed to continue the $50 pass-through, but were required to pay for it from their own funds.

[3] See, for example, Formoso (1999) and Garasky et al (1999).

[4] Wheaton and Sorensen (1998) do not make a clear distinction between cost recovery and cost avoidance. The lower estimate (half a billion dollars) counts all AFDC benefit reductions as cost recovery. The upper estimate (one billion dollars) only counts AFDC benefit reductions to full-year AFDC recipients as cost recovery. The true cost avoidance estimate would fall somewhere between these two extremes.

[5] See Barnow et al (2000), Lerman and Sorensen (2001), and Roberts (2002) for a review of the literature concerning the additional benefits of child support enforcement. Sorensen and Pomper (2002) present recent findings concerning the link between child support and father visitation.

[6] Documentation is available on-line at http://trim3.urban.org/T3Welcome.php.

[7] The March 2000 CPS was the most recent CPS available for TRIM3 simulations at the time of this analysis.

[8] The CPS only asks income questions of persons aged 15 or older. Therefore, child support income received on behalf of children under 15 is reported on the parent's record. Child support received on behalf of older children may be reported on either the child's or the parent's record.

[9] SSI treats child support as the income of the child, and TANF does not count the child support income of children outside the TANF assistance unit (such as a disabled child receiving SSI). For consistency, child support is assumed to be counted as the income of the child when simulating housing assistance, the Food Stamp Program, and Medicaid. Treating child support as the income of the child has the greatest effect on SSI, followed by TANF. The effects on the other programs are negligible.

[10] About 3 percent of persons who report child support on the CPS are in households without children. The vast majority are middle-aged women who are most likely reporting child support income received on behalf of children who were in the household in the prior calendar year, but are no longer present at the time of the survey. Since the household composition (reflecting the survey month) is not consistent with the household income (reflecting the prior calendar year) cost avoidance ratios derived for these households would be unreliable. Setting child support to zero in the baseline and alternative simulations excludes these households from the cost avoidance estimates.

[11] The baseline used in this analysis is based on version two of the 1999 TRIM3 baselines (version one of the TRIM3 Medicaid baseline), except that child support is treated as the income of the child. No additional alignment was performed on the resulting baseline.

[12] The 1998 CPS-CSS was the most recently available CPS-CSS at the time of this analysis.

[13] The FY 1999 OCSE data were selected in order to maintain consistency with the calendar year 1999 TRIM3 results generated here.

[14] The TRIM3 simulation is based on a monthly accounting period. Annual child support amounts reported on the CPS are allocated across the months of the year. A family is counted as a non-TANF family in each month in which the family does not receive TANF, and the family's child support income and benefit reductions in that month are incorporated into this table.

[15] The $13.4 billion is the sum of "medical support" and "payments to families" for "former assistance" and "never assistance" families from table 1 of the Child Support Enforcement FY 1999 Preliminary Data Report (OCSE 2000).

[16] A certain amount of disability information is available for children aged 15 and above, and TRIM3 does capture cost avoidance from changes in eligibility and participation for one or two of these older children. However, this is a negligible part of the overall effect.

[17] Under the Section 8 voucher program, Congress adjusts funding each year so as to cover the costs of all vouchers (Sard and Fischer 2003). Households that receive child support income are able to pay a higher portion of the rent than otherwise be the case, thereby reducing the amount thatmust becovered by the federal government.Under the public housing program, the federal government provides operating subsidies to local housing authorities under a Performance Funding System Formula. However, the federal government does not always provide the full subsidy calculated by the formula, and adjustments for changes in rental revenues may lag by two to three years (GAO 1998). Therefore, theadditional rents from child support income mayinitiallyhelp local housing authorities cover their operating costs and eventually be passed on as savings to the federal government in the form of lower operating subsidies.

[18] TRIM3 simulates SCHIP eligibility but not the SCHIP enrollment decision. The net effect of switching from SCHIP coverage to Medicaid coverage is unclear. To avoid overestimating cost avoidance, no cost avoidance is attributed to children who are eligible for SCHIP in the baseline but receive Medicaid in the absence of child support.

[19] The value is estimated using administrative data that show the total number of non-institutionalized Medicaid enrollees in the average month of the year and the total annual cost of providing Medicaid coverage to non-institutionalized persons. To obtain the average monthly cost per person, the total annual costs are divided by twelve times the average monthly number of Medicaid enrollees. The calculation is performed by state and by person type: disabled, elderly, non-disabled child, and non-disabled and non-elderly adult.

[20] See Appendix B for detailed cost avoidance ratios by transfer program and poverty level.

[21] Information in this section is obtained from John Tapogna based primarily on conversations with Amanda Barlow and Mike Bratt of the Administration for Children and Families at the U.S. Department of Health and Human Services.

[22] An unknown percentage of the medical support payments are sent directly to Medicaid to help offset Medicaid expenditures for persons enrolled in Medicaid. Such payments would actually generate cost avoidance of 100 percent, rather than the lesser amount implied by table 4.

[23] The data used to estimate the effect of the child support deduction do not indicate whether a noncustodial parent pays child support through the IV-D system, or whether the support is paid privately. To the extent that some of these parents pay child support privately, total IV-D food stamp cost avoidance, net the child support deduction, would be higher than that shown in table 4.

[24] According to Rosso and Fowler (2000), 66,000 households claimed the child support deduction in the average month of fiscal year 1999, and the average monthly amount of the deduction for claimant households was $180. The total annual deduction is estimated at $143 million (66,000*$180*12). Under the rules of the Food Stamp Program, a household's food stamp benefit is calculated by subtracting 30 percent of the household's net income from the maximum food stamp allotment. Therefore, the child support deduction is assumed to increase the food stamp benefit for households claiming the deduction by 30 percent of the value of the deduction ($43 million). All noncustodial parents who claim the deduction are assumed to pay child support through the IV-D system, so the entire $43 million is subtracted from the initial estimate of food stamp cost avoidance attributable to IV-D child support enforcement.