Prepared by:
Matthew Lyon
Prepared for:
Office of Human Services Policy
Office of the Assistant Secretary for Planning and Evaluation
U.S. Department of Health and Human Services
Introduction
The purpose of this analysis is to identify the characteristics of the custodial parent families who used the services of the child support enforcement system authorized under Title IV-D of the Social Security Act in 1997. The analysis uses survey data from the Census Bureau to describe this population by various family characteristics, including income, poverty, and participation in government programs, including cash assistance. It also provides information on demographic characteristics such as the gender and marital status of the custodial parent and the residence of the noncustodial parent. Finally, it addresses the question of whether the family has an child support agreement in place or has received child support payments.
The report examines these outcomes for the entire population of child support-eligible families in 1997, as well as for those both receiving and not receiving IV-D services. In addition, subgroup analysis has been done for families reporting different racial and ethnicity status, and for those families with incomes both below or near the federal poverty line.
Participation in the IV-D system was determined by a number of variables that are detailed below and in the technical appendix. Respondents are further broken down into three categories:
- Families receiving cash benefits through the Temporary Assistance for Needy Families (TANF) program;
- Families using other means-tested non-cash public assistance programs (i.e. Medicaid, food stamps, housing assistance, and Supplemental Security Income (SSI)), and
- Families that reported no public assistance usage during 1997.
Data Source and Methodology
The source for this analysis is the data file matching the March 1998 Current Population Survey (CPS) and the April 1998 Child Support Supplement (CSS). Every March, the Census Bureau administers an expanded version of the monthly CPS that includes key demographic and income variables. Every other April, with funding from the Office of Child Support Enforcement (OCSE), 75 percent of the March CPS sample is given a supplemental survey that includes questions on children with absent parents and the child support program. The March and April data are then merged into one single file. The CPS-CSS match file from March/April 1998 covers the status of families in calendar year 1997, and is the most recent major national child support survey for which data are available.
This analysis is a follow up to a May 1999 report that used the 1996 CPS/CSS match file to look at the circumstances of families receiving IV-D services in 1995 (see Characteristics of Families Using Title IV-D Services in 1995, http://aspe.hhs.gov/hsp/CSE-Char99/CSE-Char99.htm). However, due to revisions made to the 1998 file by the Census Bureau, the 1995 and 1997 data are not directly comparable.
The population of interest for this analysis, custodial families that use the IV-D system, was drawn from the identified population of child support-eligible parents using variables from both the March and April surveys. These variables included: parents who reported that they had contacted the child support program for help or had been contacted by the child support program; parents who reported receiving their child support payment through the child support or welfare agency, and; parents who said they did not know the amount of child support due because the child support agency had filed the paperwork. Parents who received cash benefits from the TANF program or were enrolled in Medicaid were also considered to be in the IV-D program. This assumption was made because of child support enforcement cooperation requirements that are part of the regulations guiding those programs. There is currently no way of comparing the numbers in this analysis to the IV-D caseload data reported by the states to the federal Office of Child Support Enforcement (OCSE). State OCSE caseload reporting requirements are not based on counts of custodial parent families like the estimates from the household-based sample of the CPS-CSS survey.
The technical appendix, found at the end of this document, includes a complete explanation of each of the variables used to make up the various categories in the analysis. Since many of the characteristics included in this piece could not be captured by only one or two variables on the CPS-CSS match file, a variety of assumptions needed to be made. These assumptions, and the effects that they may have had on the findings, are also detailed in the technical appendix.
Findings
Overall Child Support-Eligible Population
There were 13,949,218 child support-eligible families in the United States in 1997. A child support-eligible family is defined as a custodial parent with an own child under age twenty-one living in the household whose other parent is living but absent from the household. A majority of the child support-eligible population receives services through the IV-D program. This analysis found that 8.4 million families, or 60 percent of the 13.9 million child support-eligible parents, participated in the IV-D system. Table 1 provides an overview of all of the family variables analyzed in this report, shown for the entire child support-eligible population. Beginning with Table 2, all subsequent tables identify these characteristics within the IV-D and non IV-D populations.
Participation in Government Programs, Including Cash Assistance
The March CPS can be used to determine the participation of families in a variety of public assistance programs. Approximately 2.4 million (17 percent) of child support-eligible parents lived in families in which at least one member received cash assistance through the TANF program. Participation in the Medicaid program was reported by 34 percent of these families, and 25 percent were enrolled in the Food Stamp program. Approximately one in one eight of the child support-eligible families reported receiving housing subsidies (in the form of public housing or housing assistance), and about 6 percent collected a portion of their family income from the SSI program.
Participation in the four non-cash public assistance programs (Medicaid, food stamps, housing subsidies, and SSI) was used to identify the three categories of families that are shown in each table: families receiving cash assistance, families receiving assistance from other government programs (but not cash assistance), and families not receiving any public assistance. As reported above, about 2.4 million child support-eligible families were receiving cash assistance, while an additional 3.2 million (23 percent) were receiving other public assistance but not cash assistance. Of the 13.9 child support-eligible families, three in five were not receiving any public assistance.
Significant variations in receipt of public assistance existed between the IV-D and non IV-D populations (Table 2). Among those families in the IV-D program, the percentages of those receiving cash assistance (28 percent), those receiving other public assistance only (34 percent), and those receiving no public assistance (38 percent) were split relatively evenly. However, an overwhelming majority (92 percent) of the 5.6 million child support-eligible families not in the IV-D program received no public assistance at all in 1997. This last statistic is affected by the assumption, stated above, that all families in which either income was received from TANF cash assistance or in which either the parent or child was covered by Medicaid, therefore participated in the IV-D program.
Family Income and Poverty
Income
About 4.4 million (52 percent) of all IV-D families had a family income under $20,000 in 1997, while 5.7 million (68 percent) had an income under $30,000 (Tables 3A and 3B). Families receiving TANF or other public assistance had, on average, lower incomes than those not receiving assistance. More than four-fifths of child support-eligible families receiving cash assistance, and three-fifths of those reporting other public assistance, had incomes of $20,000 or below. Conversely, only one in five of IV-D families receiving no public assistance had incomes below $20,000.
Families participating in the IV-D program generally had lower incomes than non IV-D families. While the annual incomes of a majority of IV-D families fell below $20,000, only 22 percent of non IV-D families fell into that same income range. In both categories, families receiving no public assistance had higher incomes than those reporting some reliance on government programs.
Ratio of Income to Poverty Level
In addition to family income, the ratio of the family's income to the poverty level is an important measure of economic well-being. Each year, the Census Bureau estimates poverty thresholds that are adjusted for the size of the family unit. In 1997, the weighted average poverty threshold was $12,802 for a family of three and $16,400 for a family of four (Table 4A).
Over 3.3 million, or 40 percent of the IV-D families were below this threshold, and over two-thirds had incomes that fell below 200 percent of the poverty level. In addition, nearly 14 percent of IV-D families were in "deep poverty", or had incomes below 50 percent of the poverty level (Tables 4B and 4C). A large majority (74 percent) of IV-D families receiving cash assistance were poor, as were 44 percent of those families receiving only other government assistance. IV-D families with no public assistance were less poor, with only one in ten having incomes below poverty and 61 percent having incomes at or above 200 percent of the poverty level.
In contrast to those families receiving IV-D services, only 12 percent of non-IV-D families were poor, while more than two-thirds of these families had incomes above 200 percent of the poverty level. Additional outcomes for families both in poverty and near poverty (incomes below 200 percent of the federal poverty level) can be found later in this analysis.
Demographics
Gender of Custodial Parent
The vast majority of families in the IV-D system are headed by custodial mothers. There were over 7.5 million IV-D families headed by females in 1997, making up 90 percent of all families receiving IV-D services (Table 5). Only about 840,000 families headed by fathers were in the IV-D caseload. However, while families headed by custodial fathers made up only 10 percent of IV-D families, custodial fathers headed nearly a quarter (22 percent) of child support-eligible families not receiving IV-D services. In addition, custodial fathers were less likely than custodial mothers to depend on public assistance; 58 percent of IV-D families headed by men were receiving no public assistance in 1997, compared to 36 percent of female-headed IV-D families.
Marital Status of Custodial Parent
Among all custodial parents of IV-D families in 1997, 38 percent had never been married, while 41 percent were divorced or separated, 19 percent were currently married, and a small number had been widowed. By contrast, a much lower percentage (18 percent) of non IV-D parents had never been married (Table 6). Of all families headed by a never-married parent, about three-fourths were participating in the IV-D program. Never-married parents also made up a high percentage of the IV-D family heads who also received TANF cash assistance in 1997 (54 percent). Of those families not receiving IV-D services, a large majority 81 percent were headed by parents who were divorced, separated, or currently married.
Residence of Noncustodial Parent
The data indicate that the custodial parent and noncustodial parent did not live in the same state in nearly 2.3 million IV-D families, comprising 27 percent of the IV-D caseload (Table 7). The percentage of interstate cases was not substantially different within the IV-D caseload than out of it (27 percent versus 24 percent). Within the IV-D caseload, custodial mothers receiving TANF or other public assistance were no more likely than those not receiving public assistance to report that the noncustodial parent lived in a different state. However, among those not receiving IV-D services, custodial parents who reported receiving non-cash government assistance were slightly less likely than those not receiving assistance to live in a different state than the noncustodial parent (68 percent as compared to 77 percent).
Agreement/Receipt Status
Nearly two-thirds (63 percent) of all IV-D families had child support agreements in place in 1997, while 45 percent of families reported the receipt of some amount of child support payment (Table 8). Families receiving no child support payments made up a majority of both the IV-D and non IV-D populations, those families participating in the IV-D program were more likely to at least have a child support agreement (63 percent as compared to 50 percent). IV-D families receiving no public assistance were more likely to have an both an agreement and receipt of payment (54 percent) than IV-D families reporting receipt of TANF (29 percent) or other government assistance (40 percent). In both the IV-D and non IV-D populations, families with both no agreement and no receipt were more likely to receive some public assistance than were other families.
Subgroup Analysis
Race/Ethnicity
One aspect missing from the earlier report on the status of child support-eligible families in 1995 was subgroup analysis using the wide breadth of data available in the CPS and CSS. One such subgroup analysis is differences in outcomes across individuals and families of different races and ethnicities.
Tables 9A and 9B show outcomes for all heads of child support-eligible families, as well as for those who are non-Hispanic white, non-Hispanic black, and of Hispanic origin. The majority of child support-eligible family heads (60 percent) are non-Hispanic white. About one-fourth (26 percent) of families eligible for child support in 1997 were non-Hispanic black, and 14 percent were of Hispanic origin.
However, among families receiving IV-D services in 1997, a slightly smaller percentage (52 percent) were non-Hispanic white, while 31 percent were non-Hispanic black, and 16 percent were of Hispanic origin. Within the IV-D caseload, the tables show that only 37 percent of those child support-eligible families receiving TANF cash assistance in 1997 were white. Non-Hispanic whites also appeared, on average, to be doing better economically than non-Hispanic blacks and Hispanics. Families headed by non-Hispanic whites made up 41 percent of child support-eligible families in poverty, and 35 percent of those in deep poverty (incomes below 50 percent of the poverty level). In contrast, these families accounted for 80 percent of all child support-eligible families with incomes above 300 percent of the poverty level.
Families at or Near Poverty
Another subgroup analysis focuses on the outcomes for low-income child support-eligible families. Tables 10A and 10B show the characteristics for families with incomes below the federal poverty line, both for all families and by race and ethnicity. Tables 11A and 11B show the same characteristics, but include families who are near poverty (with incomes below 200 percent of the federal poverty level).
In 1997, 3,930,765 child support-eligible families, or 29 percent of the total population of families eligible for child support, had incomes below the federal poverty line. An additional 25 percent of families had incomes between 100 percent and 200 percent of poverty, leaving a majority of child support-eligible families either at or near poverty. Poor families were much more likely than the general population of child support-eligible families to receive IV-D services. Over 83 percent of poor child support-eligible families were in the IV-D system in 1997, compared to 60 percent of the general population of child support-eligible families.
Poor families generally fared worse than all child support-eligible families across a number of outcomes. Poor families had lower incomes and custodial parents were more likely to be never married than in other child support-eligible families. In addition, child support-eligible families who were poor were slightly less likely than families with incomes above the poverty line to have a child support award in place or be receiving child support. However, approximately the same percentage (74 percent) of poor families had the non-custodial parent living in the same state as was the case for all families.
Limitations and Sensitivity to Changes in Assumptions
A variety of limitations exist in this study. One of the most crucial lies in the definition of the IV-D population. Since there is no direct question on the April CSS survey regarding receipt of IV-D services, this number had to be estimated through the use of variables from both the March and April surveys. Changing some of the assumptions used in constructing the IV-D population could significantly affect its size and its interaction with the various characteristics. For example, former TANF families who do not report contact with the IV-D program could be erroneously identified as being outside the IV-D service population even if, based on their former welfare status, they are still receiving services. The use of "family" variables from the CPS, which include only the family and related subfamilies, may also affect the data.
Unlike the previous report on the circumstances of this population, this report uses data from a period either during or after the implementation of the TANF program in the vast majority of states. In fact, while the 1995 and 1997 are not directly comparable due to revisions made by the Census Bureau to the CPS/CSS data file in Summer 2001, there is a clear reduction between the two periods in the percentage of child support-eligible families receiving cash assistance. Since, according to the assumptions of this analysis, receipt of cash assistance is one of the variables that places families in the "receiving IV-D category", this reduction led to a subsequent decline in the percentage of families receiving IV-D services between 1995 and 1997.
Technical Appendix
This appendix explains the process used to estimate the child support-eligible population, IV-D population, and various characteristics within this analysis. The choice of variables represents a "best guess" of the IV-D population and their receipt of TANF and other public assistance. If some of the assumptions made in the development of this analysis were changed, it would clearly have an effect the findings cited in this document.
Child Support-eligible Population
The number of child support-eligible families was determined by PRSELIG, a recoded variable on the April CSS. This recode designates that a parent is eligible to be asked the questions on the CSS; that is, they are a custodial parent with an own child under age twenty-one living in the household whose other parent is absent from the household.
IV-D Population
This number had to be estimated through the use of variables from both the March and April surveys. Four questions from the April CSS were used:
- PES400: If the respondent reported ever having contacted a child support enforcement or IV-D office, a department of social services, a welfare office, or any state or local government agency about anything to do with child support, than he/she was included.
- PES401: If the respondent reported ever having been contacted by one of these agencies about anything to do with child support, then he/she was included.
- PES300: Individuals were asked if the noncustodial parent was supposed to make any payments for their child between January 1 and December 31, 1997. If the respondent answered that they didn't know because the Child Support Enforcement Office had filed the paperwork, then he/she was included.
- PES303: Individuals were asked how payments were supposed to be sent to them. If the respondent replied that payments were to be sent by a child support, welfare, or other public agency, then he/she was included.
In addition, all families in which income was received from TANF or in which either the parent or child was covered by Medicaid were included in the estimation of the IV-D caseload. These families were designated by the March CPS variables FINC_PAW (family income received from public assistance or other welfare) and MCAID (the parent or child was covered by Medicaid). This assumption was made because of the child support enforcement cooperation requirements that are part of the regulations guiding those programs.
Public Assistance Variables
With one exception, family variables from the March CPS were used to determine receipt of public assistance of child support-eligible families (see explanation of Census definition of "family", below). Participation in five different public assistance programs was estimated:
- TANF: The variable FINC_PAW was used to determine if the child support-eligible family received any income from a cash assistance or welfare program.
- Medicaid: This was the only variable in which full family receipt was not used as the definition of coverage. The person variable MCAID was pulled out of the March questionnaire. A positive answer to this question meant that the mother was covered by Medicaid.
Next, the variable FAMCAID was constructed with data from the related subfamily. The family sequence (FH_SEQ) and family position (FF_POS) variables were used to differentiate the child from the other members of the immediate family. Finally, MCAID was used to gather the child support-eligible children covered by Medicaid whose custodial parents were not covered. The covered parents and covered children were combined into the dummy variable FAMCAID. Note that while the parents of child-only Medicaid cases are not technically required by federal statutes to cooperate with child support enforcement, many states have policies (or had policies in 1997) that encourage cooperation with the IV-D agency if any member of the family is receiving Medicaid.
- Food Stamps: The family variable F_MV_FS, denoting the total value of the family's food stamps, was pulled from the March CPS. A dummy variable, FSFLG, was then constructed, with a value of one indicating that the amount of food stamps the family received was greater than zero.
- Housing Subsidies: The family variable HOUSSUB shows that the family is receiving some value of housing subsidy, including housing assistance or public housing. This variable was used to construct the dummy HOUSFLG, with a value of one denoting that the family received some amount of housing subsidy in 1997.
- Supplemental Security Income (SSI): The family variable FINC_SSI was used to determine if the child support-eligible family received any income from the SSI program in 1997.
After these variables were estimated and run against the various characteristics, they were condensed into three categories: families receiving cash benefits under the TANF program, families using other means-tested non-cash public assistance programs, and families that reported no public assistance usage during 1997. Families receiving other public assistance were defined as those families with positive responses to any of the non-cash public assistance categories but that received no income from the TANF program. Families receiving no public assistance were defined as those with negative responses to each of the five public assistance categories.
Income Variables
Family income and poverty values were drawn from the March CPS. The Census definition of "family" includes the immediate family of the respondent as well as any related subfamilies in the same household. It excludes unrelated subfamilies living within the same household. For example, if a custodial mother lived with her two parents in 1997, the incomes of both the mother and her parents would be included when determining the family income. If the custodial mother instead lived with a friend of the family, her income alone would determine the family income.
The March variables FTOTVAL and POVLL were used to produce the family income and the ratio of family income to the poverty level, respectively. Two new variables, FAMINC and POVRATIO, were constructed to develop the ranges of income and ratios of income to poverty that are seen in Tables 3A, 3B, 4A, and 4B.
Demographic Variables
The person-level variables A_SEX and A_MARITL were drawn from the March survey to determine the gender and marital status of the custodial parent. A_MARITL was then slightly modified, condensing different classifications of currently married custodial parents. This new variable, MARITAL, produced the five categories of marital status used in this piece. These two demographic characteristics of the child support-eligible, IV-D, and non IV-D populations are shown in Tables 5 and 6.
Another demographic variable used was the residence of the noncustodial parent relative to the custodial parent. In question PES601 of the April supplement, the custodial parent is asked if the noncustodial parent lives in a different state. The respondent's answer to this question was used to determine his/her classification in one of the two categories in Table 7.
Agreement/Receipt Status
Finally, several variables from the April supplemental survey were used to determine the child support status of the child support-eligible population. The first were PRCSREC and PRTYPAWD, recodes denoting the amount of child support received and the type of child support agreement. If the family received a positive amount of child support in 1997, then they were given a value of one in the new dummy variable RECEIPT. If the family had an award in the form of a legal agreement, pending legal agreement, or informal agreement, then were given a value of one in the new dummy variable AWARD. AWARD was then run in a simple crosstab against RECEIPT to produce the four categories in Table 8.
Subgroup Analysis
The subgroup analysis looked at the child support-eligible families by race and ethnicity. This information was gathered from the March CPS person-level variables A_RACE and A_REORGN.
List of Tables
All the tables are contained in a single Microsoft Excel spreadsheet that may be downloaded and viewed by clicking on the Tables link (Excel file, 150KB). [To view this file, you will need the Excel software or other software that reads Excel files.]