Eligibility for CCDF-Funded Child Care Subsidies under the October 1999 Program Rules: Results from the TRIM3 Microsimulation Model

Eligibility for CCDF-Funded Child Care Subsidies Under the October 1999 Program Rules: Results from the TRIM3 Microsimulation Model

Contents

• Acknowledgements
• Introduction
• Methodology
  • Overview of TRIM3
  • TRIM3's Child Care Eligibility Modeling
    • Units of Analysis
    • Time period
    • Eligibility criteria for the family head and spouse
    • Eligibility criteria for children
    • Income eligibility criteria
  • Input Data for the 1999 Eligibility Estimates
  • Interpreting the Estimates
• The Eligibility Estimates
• State Level Estimates
• Estimates by Demographic Characteristics
  • Description of the Tables
  • Demographic Characteristics of Eligible Children
  • Demographic Characteristics of Eligible Families
• Cross Tabulation of Eligibility by Family Marital Status, Parental Work Status, and Age
  • Description of the tables
  • Results of the cross-tabulation
• Eligibility and Need

• References
• Endnotes
• Appendix Tables

Introduction

This paper reports estimates of the numbers of families and children who may be eligible for child care subsidies under the Child Care and Development Fund (CCDF). The estimates are produced by the Urban Institutes TRIM3 model (the Transfer Income Model, version 3). By eligible we mean that a family would technically qualify for subsidies, if that family chose to apply for subsidies. An eligible family may or may not want or need subsidies. Included are estimates of families and children eligible under state rules in effect as of October 1999. This paper also shows the number of families and children who are under the maximum federal standard of 85 percent of state median income. The paper explores the estimation methodology and also presents detailed national and state-level estimates.

The Child Care and Development Fund (CCDF), authorized by the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, PL 104-193, assists low-income families, families receiving temporary public assistance, and those transitioning from public assistance in obtaining child care so they can work or attend training/education.⁽¹⁾ Under the CCDF, states receive grants from the federal government to operate child care subsidy programs.⁽²⁾ While other child care subsidies are paid for by federal TANF funds and by state, local, and private monies, the CCDF is the single largest source of funding for child care subsidies in many states.⁽³⁾

Each state determines how its CCDF program will operate, within the overall guidelines in the federal regulations. These guidelines stipulate three basic criteria for eligibility: childs age, family income, and parent work status. Typically, a child must be under 13 years old to be eligible. However, children with certain special needs may be eligible up to 19 years old. Family income must not exceed 85 percent of the states median for a family of the same size. Parents must be working or attending a job training or educational program. Within these parameters, each state establishes specific income thresholds and other rules that define who is legally eligible in the state. The states are required to report these rules in their biennial CCDF operating plans, submitted to the federal government by October of each odd-numbered year.

To help understand the potential reach of the CCDF program, as well as to assess possible changes in the system and the impact of variations in state rules, the Urban Institute developed a model under funding from the Department of Health and Human Services to estimate the number of children and families eligible for CCDF subsidies. While this model can estimate eligibility for the program under different sets of rules, it does not attempt to measure need for subsidies. Whether or not a family needs a subsidy, or any type of non-parental child care for that matter, depends on the alternatives for care available to the family.

This paper reports CCDF estimates from the Urban Institutes TRIM3 model (the Transfer Income Model, version 3). Included are two sets of estimates as of October 1999.⁽⁴⁾ One set uses the financial and demographic eligibility criteria that states delineate in their state plans. This set of state rules estimates provides a measure of the number of children and families eligible for CCDF subsidies as the program was implemented by the states in October 1999. The other set is based on the federal guidelines, described above. This set of federal maximum estimates provides a benchmark of the number of children and families who are below 85 percent of state median income, the maximum income threshold allowed under federal law.

TRIM3 estimates that 9.7 million children and 5.8 million families are eligible for CCDF subsidies under October 1999 state rules. The model also shows that 15.7 million children and 9.5 million families are estimated to be below the federal maximum income standard. The difference between these two sets of estimates reflects the fact that many states establish more restrictive eligibility rules than those allowed by the federal law. Most notably, many states set income eligibility thresholds below 85 percent of state median income.

The estimates can be examined for each state and, at the national level, for families and children with different demographic characteristics. For example, 4.4 million children under age six and 5.2 million children ages six through twelve are eligible for CCDF-funded child care subsidies under state rules. Forty-two percent of all children eligible for subsidies under state rules have annual family incomes below poverty. Furthermore, roughly 0.6 million eligible families and 1 million eligible children are current TANF recipients. The number of estimated eligible children and families varies widely by state.

The remainder of this paper explores the estimation methodology and the subsequent results in more detail. The next section discusses the fundamentals of the TRIM3 model and defines the parameters for the CCDF subsidy estimation. The third section describes the national and state level estimates that are presented in the attached tables. The paper concludes with a final discussion of the distinction between estimate of eligibility and estimates of child care need.

[ Go to Contents ]

Methodology

TRIM3 is a comprehensive model of tax, transfer, and health programs affecting the U.S. population.⁽⁵⁾ This section provides an overview of TRIM3 followed by a detailed discussion of how TRIM3 models eligibility for child care subsidies, a description of the input data used for these simulations, and information on how to interpret the estimates.

Overview of TRIM3

As a microsimulation model, TRIM3 uses survey data on individuals, families, and households to simulate the effects of major government programs. The TRIM3 computer program applies the rules of the particular social program being simulated to each household in the microdata input file. The model follows program rules as closely as possible given the limitations of the available data and policy information.

TRIM3s input file is a slightly modified version of each years March Current Population Survey (CPS). The March CPS samples about 55,000 households, and is nationally representative of the civilian non-institutionalized population of the United States. The survey is used extensively for analysis of the demographic and economic characteristics of the population. For example, it is the source of the official statistics on the countrys poverty rate.

From a substantive perspective, the main modification made to the CPS prior to its use in TRIM3 simulations is the allocation of certain annual variables across the months of the year.⁽⁶⁾ The CPS only asks about annual income amounts. However, to simulate programs that operate

on a monthly basis, TRIM3 needs monthly income amounts and monthly information on labor force activity. TRIM3 allocates reported annual earnings and reported labor force activity in a manner consistent with each individuals reported annual labor force participation and consistent with actual trends in employment and unemployment across the months of the year. Because TRIM3 models government programs on a monthly basis, it is not possible to reproduce its results using unadjusted annual CPS data.

TRIM3 simulation results can be examined for the entire population, for subgroups of the population, and for states. When state-specific estimates are needed, TRIM3 projects may use three consecutive years of CPS data and average the resulting estimates in order to reduce the margin of error associated with each estimate. That strategy was used for this research, and is discussed further below.

TRIM3S Child Care Eligibility Modeling

One of TRIM3s modules estimates eligibility for CCDF-funded child care subsidies. The eligibility estimates are intended to assist researchers and policymakers in understanding the potential reach of the CCDF program. In addition, the eligibility estimates will serve as a basis for identifying the recipients of subsidies in the TRIM-CPS data, facilitating other TRIM analyses.

To develop the CCDF eligibility estimates, TRIM3 examines each household in the input data to determine whether the family(ies) and child(ren) in that household are eligible for CCDF-funded subsidies. A family or child is deemed eligible if it meets all the necessary criteria based on (1) the activities of the family head (and spouse); (2) the ages of the children; and, (3) the familys income.

TRIM3 can model CCDF eligibility under any specified set of program rules. As noted in the introduction, this project models eligibility under two sets of program rules. To simulate state rules eligibility, the model uses the actual income thresholds and other rules defined in each states October 1999 CCDF plan.⁽⁷⁾ To derive the federal maximum estimates, the model sets the income eligibility thresholds at 85 percent of state median income (the maximum allowed by federal law), and uses the same requirements for head/spouse activity status and childrens ages across all states. Below, we describe the sequential process by which TRIM3 calculates eligibility, and we detail the program rules that define each simulation.

Units of Analysis

CCDF eligibility estimates are provided for both children and families, but the model determines eligibility based on resources measured at the family level.⁽⁸⁾ Within a family, some children may be eligible for subsidies while others are not, depending on their age and disability status. A family with at least one eligible child is considered an eligible family, even if all of the children in the family are not eligible for subsidies.

Family size is an important variable used to determine whether a family is financially eligible for CCDF subsidies for both the state rules and the federal maximum simulations. All members of the family are included for purposes of calculating family size, including older children and relatives other than the head and spouse. TRIM3 considers children who live with caretakers who are not their parents (for example foster children or children living with grandparents) as members of their non-parent caretakers family.⁽⁹⁾

Time period

The state rules and the federal maximum simulations both determine CCDF eligibility on a monthly basis.⁽¹⁰⁾ For each month of the year a child or family is eligible only if it meets all of the eligibility criteria in that month. As a result, TRIM3 may designate a family with relatively high annual income as eligible for child care subsidies in some months, if monthly income is low during one or more months of the year. Conversely, TRIM3 may designate a family with low annual income as ineligible for subsidies in months when income is relatively high.

Eligibility criteria for the family head and spouse

To be eligible for CCDF-funded subsidies, TRIM3 requires that both the head of the family and the spouse (if there is a spouse present) are either employed or in school.⁽¹¹⁾ The state rules simulation and the federal maximum simulation use different definitions of working. In the simulation modeling the states actual program rules, the definition of working is established by definitions in the state CCDF plans. In some states, a person must work at least a minimum number of hours per week (e.g., 20 hours) to be eligible for CCDF-funded subsidies. Other states do not require families to meet an hours requirement to be eligible for subsidies (or do not state such a requirement in their state plan). The minimum hours requirement for each state according to the October 1999 CCDF plans is listed in Appendix Table 1. In the federal maximum simulation, TRIM3 submits all family heads and spouses to the same definition of work: the presence of any earnings during the month.

The state rules and federal maximum simulations use the same definition of being in school, based on the limited information reported in the CPS interview. If a person does not work the entire calendar year, the CPS asks why not; and if a person works for less than the entire year, the CPS asks what s/he was doing the remainder of the year. If the person responds that she was in school, then TRIM3 allocates student status to that person for some portion of the calendar year.⁽¹²⁾

Eligibility criteria for children

In both the state rules simulation and the federal maximum simulation, the maximum age at which a non-special needs child is eligible for CCDF-funded subsidies is 12. However, the two sets of program rules vary on the maximum age at which children with special needs are eligible for subsidies. In the state rules simulation, TRIM3 uses information from the state CCDF plans to determine the maximum age at which a special needs child may be eligible for subsidies.⁽¹³⁾ Some states limit eligibility for these children to those age 12 or younger, while others extend eligibility through the teen years up to age 19. Appendix Table 2 lists the maximum ages recorded in the October 1999 state plans. Under the federal maximum simulation, regardless of the familys state of residence, a special needs child is assumed to be eligible for CCDF subsidies through the age of 18.

Due to data limitations, the only type of special needs child that TRIM3s child care module captures is one who is disabled, and only if the child receives Supplemental Security Income (SSI).⁽¹⁴⁾ To the extent that there are children who are considered special needs for CCDF purposes who do not receive SSI, TRIM3 will undercount the number of eligible special needs children in both the federal maximum and the state rules simulations. For the remainder of this paper, we refer to this subgroup of SSI-receiving special needs children as disabled.

Income eligibility criteria

To be eligible for CCDF-funded subsidies in a particular month, a family must have monthly income below the threshold for that familys size and state of residence. TRIM3 compares the familys monthly income to the appropriate income threshold. As discussed above, the family size is the total number of people in the family, with subfamilies treated as separate families.

Monthly family income is measured differently in the state rules and federal maximum simulations. For the federal maximum simulation, family income comprises the total cash income of the head and spouse, plus the SSI income of eligible children.⁽¹⁵⁾ For the state rules simulation, TRIM3 captures some of the variation in definitions of income used by the states, as indicated in their state plans. For example, some state plans specify that the state does not count certain types of income  such as SSI or general assistance income  in their determination of CCDF eligibility. Furthermore, some state plans explicitly disregard a portion of earned income. These variations in income definition (according to the October 1999 CCDF plans) are listed in Appendix Table 3.

In addition to using different definitions of family income, the state rules and federal maximum simulations use different income thresholds. The state rules simulation uses the income thresholds that are contained in each state plan.⁽¹⁶⁾ In some states, the thresholds vary by whether a family is newly applying for eligibility or already eligible for subsidies. If a states thresholds vary by initial vs. continuing eligibility, and the family is simulated as eligible for CCDF in the prior month, then the continuing eligibility limits are used for the current month; otherwise, the initial eligibility limits are used. A handful of states allow families with disabled children to have higher initial or continuing eligibility limits. Each states initial eligibility limits for non-disabled children as of October 1999 are listed in Appendix Table 4. Appendix Table 5 lists each states continuing eligibility limit, or disabled eligibility limit where applicable. Note that the majority of states have the same limits for both initial and continuing eligibility for all children.

The federal maximum income thresholds are set at 85 percent of state median income (SMI) for family sizes ranging from one person to 10 people.⁽¹⁷⁾Appendix Table 6 lists the 85 percent SMI income thresholds used in the federal maximum simulation for all states and all family sizes. Comparing Appendix Tables 4 and 5 with Appendix Table 6, note that the state rules thresholds are often substantially lower than 85 percent of SMI.

Input Data for the 1999 Eligibility Estimates

To improve the precision of the state-level eligibility estimates and estimates for demographic sub-groups of the national population, TRIM3 uses three years of CPS data to estimate CCDF eligibility. For the results presented in this paper, the three input files are based on the March 1998, March 1999, and March 2000 CPS files, with income data from calendar years 1997, 1998, and 1999. The same rules are applied to each year of data, except the income limits used in the 1999 run are deflated from 1999 to either 1998 or 1997 dollars using the Consumer Price Index (CPI). The eligibility estimates in the following tables are calculated by taking a simple average of the results from the three individual runs.⁽¹⁸⁾ Note that by averaging eligibility estimates over the three years of data, we may misestimate the number of families and children who were eligible for CCDF in October 1999, due to changes in demographic and economic circumstances over the period.

Interpreting the Estimates

The eligibility counts presented in the tables are our best estimates of CCDF eligibility. However, we also report the margin of error  or confidence interval  associated with the estimates in tables 1 through 4. The confidence interval is the range inside which the true number is likely to fall. In these tables, we show the 95-percent confidence interval for each estimate.⁽¹⁹⁾ There is a 95 percent chance that the true number falls in the range defined by the 95-percent confidence interval. For example, in the state rules simulation, our best estimate of the number of eligible children aged 0 through 2 is 2.10 million (Table 3); but we can be 95 percent certain only that the number is between 1.96 and 2.24 million.

Using three years of CPS data produces smaller confidence intervals than would be the case with only one year of CPS data, due to the increased sample size. However, the confidence intervals around some of the estimates  particularly the estimates for the states with smaller populations  are sizable relative to the estimates themselves. For instance, while the point estimate for the number of children eligible under Maines own state rules is 56 thousand, the 95 percent confidence interval is 42 thousand to 70 thousand  plus or minus 18 percent of the point estimate. Ranges of this type are common for state-specific estimates generated from CPS data; nevertheless, readers should keep the confidence intervals in mind in interpreting the estimates.^{(20, 21)}

[ Go to Contents ]

The Eligibility Estimates

Tables 1 through 6 present the TRIM3 estimates of CCDF eligibility as of October 1999. The tables disaggregate the nationwide estimates of children and families eligible for CCDF child care subsidies by state of residence and various demographic characteristics. Tables 1 through 4 juxtapose estimates of eligibility under October 1999 state rules and counts of children and families under federal maximum rules, with the odd-numbered tables counting children and the even-numbered counting families. Tables 5 and 6 provide a more detailed description of identified children, first under state rules (Table 5) and then under federal maximum rules (Table 6).

The tables show our estimates of eligibility in the average month of the year. For example, for a particular state and year we may estimate 100,000 children eligible in January, 105,000 in February, 102,000 in March, etc., and an average monthly number over the calendar year of 103,000. The numbers of children or families eligible in the average month will always be lower than estimates of those who are ever eligible during the calendar year.⁽²²⁾

The remainder of this section explains the layout of the tables, defines the dimensions on which eligibility is broken down, and highlights some interesting findings from this analysis.

[ Go to Contents ]

State Level Estimates

Tables 1 and 2 list the estimates of eligible children and families in each state and the District of Columbia. Although there is a margin of error around each number, these are our best point estimates of the numbers of children and families eligible for CCDF-funded subsidies. The tables also include the 95 percent confidence interval associated with each estimate. As discussed above in the methodology section, the 95 percent confidence interval gives the range that we are 95 percent confident contains the actual number of eligible children/families. Note that the confidence intervals can be fairly large relative to the point estimates, especially for smaller states, making cross-state comparisons difficult.

The last row in each table displays the nationwide total estimates that were highlighted in the introduction. TRIM3 estimates that 9.7 million children are eligible for CCDF subsidies under the state rules in effect in October 1999 and 15.7 million children are potentially eligible under federal law. At the family level, 5.8 million families are eligible under October 1999 state rules, while 9.5 million families are potentially eligible.

[ Go to Contents ]

Estimates by Demographic Characteristics

Tables 3 and 4 display estimates of average monthly eligibility on the national level broken down by the demographic characteristics of children and families. Below, we first describe the contents of the tables, and then present the results.

Description of the Tables

The left side of each table presents eligibility estimates under October 1999 state rules, while the right side shows estimates using the maximums allowed under federal law. The 95 percent confidence intervals are also included in these tables, indicating the precision of the estimates for each demographic sub-category. The % of total columns for the state rules and federal maximum estimates can be used to compare, within each demographic category, the relative size of each class of eligible children or families.

Tables 3 and 4 count children and families according to six characteristics:

• Age: This category separates eligibility estimates by the age of the child, or, in the family estimates, the age of the youngest child. Eligible children age 13 or older are only those that TRIM3 identifies as SSI recipients. However, estimates for this age group are not shown due to small sample sizes. The CPS does not capture changes in age across the year (or exact birthdays) so a child or family is in the same row for all months of eligibility.
• Annual poverty status: Family income as a percentage of poverty is calculated by comparing the annual family income to the annual poverty threshold. In contrast to the narrow definition of the family that TRIM uses when determining eligibility, i.e. all subfamilies considered separately, this breakdown uses the broad definition of the family which includes related subfamilies as part of the households primary family. This definition is consistent with standard Census Bureau methodology for determining the poverty characteristics of the population. Note that a familys poverty status in the month(s) of eligibility may differ from the annual poverty status. For example, a family with annual income more than twice the poverty threshold may have much lower income in some months, and might be eligible for CCDF subsidies only in those months, but all of the familys months of eligibility would still be counted in the 200+ row. On the other hand, a family that only works for part of the year may have a low annual income that puts it in the poverty category, but it is only eligible in the months that the head is working and in those months its income is higher.
• Monthly income level: Monthly income includes the total cash income of the family head and spouse, plus the SSI benefits of eligible children. The monthly income breaks in the table are based on annual incomes of $10,000, $15,000, $20,000, and $25,000.⁽²³⁾ Note, however, that family incomes can fluctuate over the months of the year, so having monthly income under $833 is not necessarily equivalent to having annual income under $10,000.
• Monthly TANF receipt: A family or child is classified by whether TANF is received by anyone in the family during the month. The variable used to determine TANF receipt is a TRIM-simulated variable that corrects for the under-reporting of TANF in the CPS data.⁽²⁴⁾ If a family receives TANF for part of the year but not the entire year, the family and children would be counted in the TANF row for CCDF-eligible months when TANF is received, and in the non-TANF row for CCDF-eligible months when TANF is not received.
• Work status: Families and children are categorized by the work status of the head and spouse (if present) of the family during the month: half-time or more, or anything less than half-time. Half time is defined as 20 hours per week. If the family head is married, both the head and spouse must work at least half time to be counted in the >half time category. Adults are considered students only if they did not work at all that month. All students fall into the half time category. If the work status of the head or spouse changes during the year, a family/child might fall into different categories in different months of eligibility.⁽²⁵⁾
• Marital status: Families and children are categorized by the marital status of the family head. The CPS does not capture changes in marital status during the year, so a family or child is in the same category for all months of eligibility.

Demographic Characteristics of Eligible Children

Age. The distributions of eligible children by age, shown in Table 3, are very similar across the two simulations. In the state rules simulation, 22 percent of the eligible children are under age 3, 23 percent are 3 through 5 years of age, and 54 percent are age 6 or older. The percentages are almost the same in the federal maximum simulation.⁽²⁶⁾ However, the percentages translate into different absolute numbers in the two simulations. For instance, there are an estimated 2.1 million eligible children under age 3 in the state rules simulation compared to 3.2 million in the federal maximum simulation.

Annual Poverty Status. The distribution of children by annual poverty status is quite different for the state rules simulation and the federal maximum simulation. Almost 4.1 million poor children are eligible for CCDF subsidies under state rules, slightly fewer than are counted under the federal maximum simulation. However, poor children represent 42 percent of the identified population under state rules, while they make up only 27 percent of the eligible population under the federal maximum simulation. Under state rules, an additional 4.5 million eligible children (47 percent of the eligible children) live in families with annual incomes between 100 percent and 200 percent of the poverty line. Only 1.1 million eligible children, or 11 percent of the state rules total, have annual family incomes over 200 percent of poverty. In contrast, 32 percent of the children under the federal maximum level have family incomes in excess of 200 percent of poverty. These results reflect the fact that many states use lower income thresholds than allowed by the federal regulations.

Monthly Income. The breakdown of eligible children by monthly family income also indicates that the maximum limits under federal law are more generous than the state rules. Under the federal maximum simulation, 48 percent of children in the average monthlive in families earning over $2084 in that month (or roughly $25,000 a year if earnings were stable over the months). The distribution of children among the other four categories is relatively even. Under state rules, the highest income category constitutes only 24 percent of the total population of eligible children. The lowest two income categories  income up to $833/month and $834 to $1250 per month  each contain about 21 percent of eligibles in the state rules simulation, compared to 13 percent in each category in the federal maximum simulation.

TANF Receipt, Parent Work Status, and Marital Status. The distribution of children by parent work status is very similar across the two simulations, with approximately 85 percent of children in families where the parents work at least half time. However, the distributions of eligible children by TANF receipt and by marital status differ between the two simulations.

In the state rules simulation, approximately 1.0 million of the children who are eligible for subsidies (11 percent of the total) are in families that receive TANF. The number of TANF children is only slightly higher in the federal maximum simulation, but in that simulation they comprise a smaller percentage of the total  about 7 percent. Almost all of the children in the federal maximum simulation but not in the state rules simulation are in families that do not receive TANF.

A larger share of eligible children reside in unmarried-parent families under the state rules (71 percent) than under the federal maximum simulation (57 percent). This is consistent with the generally lower incomes and higher poverty among those eligible under state rules as compared with the federal maximum rules.

Demographic Characteristics of Eligible Families

The demographic characteristics of families with CCDF-eligible children are largely consistent with those of the children themselves, and are shown in Table 4.

Age. The distribution of families by the age of the youngest child is very similar across the two simulations. In the state rules simulation, 32 percent of families have a child under age 3, 26 percent have a youngest child age 3, 4, or 5, and 41 percent have no child younger than 6. The comparable figures from the federal maximum simulation differ by no more than two percentage points from the state rules results.

Annual Poverty Status and Monthly Income. For each simulation, the distribution of families by annual poverty status is nearly identical to the distribution of children by poverty status. In the simulation of state rules, 41 percent of the eligible families are poor, 47 percent have incomes from 100 to 200 percent of poverty, and only 12 percent have income above 200 percent of poverty. In the simulation of the federal maximum regulations, 26 percent of families are poor on an annual basis, 41 percent have incomes between 100 and 200 percent of the poverty threshold, and 33 percent have income over 200 percent of poverty. The estimated number of poor families simulated to be eligible for CCDF subsidies is about the same in both simulations, at 2.4 million.

The distribution of monthly income among eligible families is also similar to that of eligible children. In each simulation, however, a higher percentage of families than children are in the two lowest monthly income category. For example, in the state rules simulation, 46 percent of eligible families earn less than $1,250 a month, a larger portion than the 42 percent of eligible children in families earning less than $1,250. These differences suggest that the eligible families with lower monthly incomes have fewer children, on average, than the eligible families with higher incomes.

TANF Receipt, Parent Work Status, and Marital Status. In each simulation, the distribution of eligible families by TANF receipt and by parent work status is very similar to the distribution of eligible children by those characteristics. In the state rules simulation, 11 percent of eligible families (0.6 million) receive TANF, and 84 percent (4.9 million) are headed by parents who work at least half time.

The distributions of eligible families by marital status differs somewhat from the comparable distributions for eligible children. In both simulations, the percentage of eligible families headed by a married couple is lower than the percentage of eligible children in married-couple families. For instance, in the state rules simulation, 25 percent of eligible families are headed by a married couple, whereas 29 percent of eligible children are in married-couple families. The differences suggest that the eligible married-couple families have more eligible children, on average, than the eligible families where the family head is not married.

[ Go to Contents ]

Cross Tabulation of Eligibility by Family Marital Status, Parental Work Status, and Age

Tables 5 and 6 are both child-level tables showing the distribution of eligibility for CCDF-funded subsidies looking at three characteristics at the same time: family marital status, parental work status, and age of the child.⁽²⁷⁾ The tables give a more nuanced picture of the composition of the eligible population than looking at the distribution by one characteristic at a time.

Description of the tables

Table 5 contains state rules estimates, while Table 6 contains estimates from the federal maximum simulation. The top panel in each table provides the estimate and the bottom panel expresses each cell as a percent of the total number of children estimated to be eligible for CCDF subsidies under the given set of rules. The left columns of each panel show the results for children living in families with a married head, while the right columns include the results for children living in families with an unmarried head.

The tables divide children into two age groups  under age 6 and aged 6 to 12  and exclude children over the age of 12. Teenagers are only eligible for CCDF subsidies if they are disabled; the relatively small population of disabled teens that could be identified with the TRIM-CPS data did not allow the kind of subbreaks shown in this table.

As in Tables 3 and 4, the tables define work hours by the lesser of the hours worked by the head of the family and the spouse if present. Here, however, hours of work are broken into three categories: 1-20 hours, 21-34 hours, and 35 hours or more. In addition, the tables separate children whose parents are students from other children. A child is in the student category if his/her single parent is a student or if at least one of his/her married parents is a student.

Results of the cross-tabulation

Under the state rules simulation, full time work is the most common for eligible children of all ages and in both marital status categories (Table 5). Of the 9.6 million children under the age of 13 who are eligible for subsidies, 6.3 million (66 percent) have parents who work full time. This figure includes 0.6 million children under six with married parents (7 percent of the total), and 2.1 million children under six with a single parent (22 percent of the total). A larger share of the eligible children of single parents have parents who work full-time than their counterparts in married families (71 percent versus 54 percent, percents not shown).

As previously seen in Table 3, children estimated to be eligible for subsidies under state rules are much more likely to be children of single parents than children under the federal maximum regulations. A comparison of the results in Tables 5 and 6 highlights this finding. The lower income thresholds implemented by the states reduce the size of the population of children with married parents by 58 percent (compare 2.8 million children in Table 5 to 6.8 million children in Table 6). In contrast, the estimated size of the eligible population of children with single parents under state rules is only 23 percent smaller than the population under federal maximum regulations (compare 6.8 million in Table 5 to 8.8 million in Table 6).

Approximately 33 percent of the children eligible under state rules are preschoolers with single parents, and more than one in five eligible children under state rules are preschoolers of a single parent who works full-time (Table 5). Across all ages and family marital statuses, the distribution of eligible children by parental work status remains fairly constant between the two simulations. The relative stability of this distribution is probably due to two counterbalancing forces. First, work and income are highly correlated, and lower income limits under the state rules simulation would typically create an eligible population with less work effort. However, many states impose minimum hours requirements, thus shifting the eligible population towards more work effort. In addition, working single parent families with only one parental earner are more likely to have lower incomes than working married parent families. Thus, the majority of the children eligible for CCDF subsidies under state rules have single parents, most of whom work full-time.

[ Go to Contents ]

Eligibility and Need

One important issue that must be considered when reviewing these estimates is that they are only estimates of eligibility for subsidies, not who needs subsidies, who receives subsidies, or the value of potential subsidies. Not all eligible families receive subsidies, either because they do not want or need assistance, they do not apply for benefits for some other reason, or they apply for subsidies but do not receive assistance due to limited subsidy availability and/or a limited supply of providers who will accept CCDF subsidies. Among those families who do receive subsidies, many are required to pay a co-payment, and therefore subsidized care does not always equal free care.

The relationship between eligibility for subsidies and need for subsidies is particularly complex. Some eligible families do not apply for benefits because they prefer other methods of managing child care expenses. Other parents who are currently not employed (and therefore not technically eligible for subsidies according to these estimates) might choose to work if they were aware of and could obtain a subsidy. However, the TRIM3 eligibility estimates do provide an important first step for analyzing the population targeted by the CCDF program.

[ Go to Contents ]

References

Bureau of the Census, 2000. Source and Accuracy of the Data for the March 2000 Current Population Survey Microdata File, http://www.bls.census.gov/cps/ads/2000/ssrcacc.htm (and analogous publications for earlier years)

Collins, Ann M., Jean I. Layzer, J. Lee Kreader, Alan Werner, and Fred B. Glantz, 2000. National Study of Child Care for Low-Income Families, State and Community Substudy Report. Project report to the U.S. Department of Health and Human Services, Administration for Children and Families, from Abt Associates.

Department of Health and Human Services, 1999. Access to Child Care for Low-Income Working Families, http://www.acf.dhhs.gov/programs/ccb.old/research/ccreport/ccreport.htm.

Greenberg, Mark, Joan Lombardi, and Rachel Schumacher, 2000. The Child Care And Development Fund: An Overview. Washington DC: Center for Law and Social Policy.

Tenny, Daniel, and Bob Zahradnik. 2001. The Poverty Despite Work Handbook, 2001. Washington, DC: The Center on Budget and Policy Priorities.

Wheaton, Laura, and Linda Giannarelli, 2000. Underreporting of Means-Tested Transfer Programs in the March CPS, American Statistical Association 2000 Proceedings of the Section on Government Statistics and Section on Social Statistics.

[ Go to Contents ]

Endnotes

1. The CCDF replaced the former Child Care Development Block Grant (CCDBG) program.

2. Each state qualifies each year for a share of funds available without matching requirements. In addition, a state can elect to receive additional matching federal funds if the state meets certain maintenance of effort requirements and commits additional state funds to draw down the matching funds. Finally, a state can elect to transfer up to 30 percent of its Temporary Assistance for Needy Families (TANF) block grant funds to the CCDF program. See Greenberg, Lombardi, and Schumacher, 2000, for a more complete description of CCDF provisions.

3. See Collins, Layzer, Kreader, Werner, and Glantz, 2000, Appendix Table 2.7a.

4. The TRIM3 methodology was previously used to produce estimates of eligibility based on the October 1997 federal and state rules (Department of Health and Human Services, 1999). Estimates based on the October 2001 rules are forthcoming. The sets of estimates may vary due to fluctuations in demographic and economic conditions, changes to state program rules, ongoing efforts to refine the model, and sampling variability.

5. Full documentation of the TRIM3 system is available on the projects website, trim.urban.org.

6. For instance, if a person says she worked for 26 weeks and was unemployed for 26 weeks, with one spell of unemployment, the allocation process places all the persons earnings in either the first half or last half of the year, and assigns the person as a worker in 6 months and as unemployed for the other 6 months. There is a random component in the assignment of weeks of work, but the model controls the assignment so that the trends in employment and unemployment across the year match the trends according to the BLS. For more details, see the document titled Monthly Allocation on the TRIM3 projects website, trim.urban.org.

7. The model is limited to program requirements that are clearly defined in the state plans. If a state has requirements that are not in its state plan, or if certain categories of requirements are not consistently included in the majority of state plans, the model does not capture these requirements.

8. CPS subfamilies, both related and unrelated, are treated as separate families.

9. Evidence from state plans suggests that children with non-parent caretakers are treated differently in some states. Further, some states may not include older children or non-parent adults in the family unit for purposes of family size and/or family income. However, the plans were not explicit on these matters. Therefore, we have chosen to use the same family definition for all states rather than to incorrectly infer a states rules.

10. Re-determining eligibility each month is a simplification of the process. States establish a minimum certification period, often six months, so that they can assess each familys child care needs on a regular basis. However, states can require families to report income changes that occur during this period, so eligibility status may change in any month.

11. In the TRIM3 model, the family head and his/her spouse must meet the activity requirements regardless of their relationship to the children in the family. This provision is consistent with the broad definition of parent in the CCDF regulations  a parent by blood, marriage or adoption and also ... a legal guardian, or other person standing in loco parentis (Federal Register, July 24, 1998, Volume 63, Number 142; available from the Child Care Bureau at http://www.acf.dhhs.gov/programs/ccb/policy1/current/finalrul/index.htm). Some state plans indicate that the state may provide child care to foster children regardless of the activity of the foster parent(s), but TRIM3 does not model this detail. For simplicity, the tables and the discussion of the results in this paper use parent to describe the head/spouse of a childs family.

12. TRIM3 allocates student status across the months in a manner consistent with the other allocations. For instance, if a person worked 26 weeks and said she was in school the other 26 weeks, then TRIM3 distributes her earnings across 6 months and assigns student status for the other 6 months. Note that TRIM3 does not identify cases when a person attends school during weeks that s/he is also working (such as someone who attends school part-time while working either full-time or part-time). Also, note that there is not a perfect relationship between the activities captured as school in the CPS and the education and training activities covered by CCDF programs. Some training may be missed by the CPS questions, and some school activities that are captured by the CPS may not be covered by all CCDF programs.

13. The CCDF regulations allow a special needs child to be defined as one who is physically or mentally incapable for caring for himself or herself, or under court supervision (see footnote 11 for citation). Children under court supervision include those receiving protective services or foster care.

14. The CPS does not directly report SSI receipt by children. A childs SSI income is intended to be captured on the income record of a parent or guardian, so a childs SSI receipt can be inferred if a parent/guardian who is neither elderly nor disabled reports SSI income. However, the number of childrens SSI cases inferred through parents reports is much less than the actual number of children who receive SSI. The TRIM3 model corrects for the under-reporting by assigning SSI receipt to additional eligible children in a way that reproduces the key characteristics of the actual childrens SSI caseload.

15. To correct for the under-reporting of welfare receipt in surveys, the CCDF eligibility model uses TRIM3-simulated SSI and TANF income in place of the reported amounts of SSI and TANF in its calculation of family income.

16. Some state plans indicate that the state may waive, on a case-by-case basis, the income eligibility requirements for children receiving protective services. However, the CPS does not have information on protective services receipt, so TRIM3 cannot make this exception and therefore requires all families to meet the states income requirements for eligibility.

17. TRIM3s 85 percent of SMI threshold for each state is based on the state median incomes for a family of size 4 in CY 1997, published by the Census Bureau (www.census.gov/hhes/income/4person.html). Due to the lag in the publication of these figures, the 1997 medians are the most recent that would have been available as of October 1999. We calculate SMI for other family sizes using multipliers used by the Low Income Home Energy Assistance Program (LIHEAP). Those multipliers are: .52 for a 1-person family, .68 for 2, .84 for 3, 1.16 for 5, 1.32 for 6, 1.35 for 7, 1.38 for 8, 1.41 for 9, and 1.44 for a 10-person family. These multipliers may or may not result in figures equal to the true state median incomes for families of each size. Families with more than 10 members are subjected to the income threshold for a family with 10 members.

18. More information about the input datasets and the methods for combining three years of CPS data is available from the authors.

19. The calculation of confidence intervals for estimates generated from three combined years of CPS data is described in Bureau of the Census, 2000.

20. For a further discussion of estimates, confidence intervals, and interpreting results, see Tenny and Zahradnik, 2001.

21. In particular, a difference between a states eligibility estimate from this report and the analogous estimate from the October 1997-based simulations may or may not represent a true increase or decrease in eligibility. Please contact the authors for more information on comparing these estimates with the prior set of estimates.

22. Annual eligibility estimates are also produced by TRIM3, and can be requested from the authors.

23. Measured in 1999 dollars. For simulations on 1997 and 1998 data, these income breaks are deflated to 1997 and 1998 dollars to determine the numbers of families and children at each income level.

24. The TRIM simulation of TANF, like the TRIM simulation of CCDF eligibility, operates on a monthly basis. The model simulates each familys eligibility for TANF in each month of the year. Eligible families that report TANF receipt are assumed to actually receive benefits. Additional eligible families are tagged as TANF recipients in a way that reproduces the actual size and characteristics of the TANF caseload according to the administrative data. This corrects for the fact that the CPS captures only about 61 percent of the actual average monthly TANF caseload (Wheaton and Giannarelli, 2000).

25. The CPS does not capture changes from part-time to full-time status, or vice versa, during the year. Instead, the survey asks for the usual hours worked per week during weeks that the person was working. However, this variable might still differ for a family or child across the months of eligibility if a parent was working in some months but a student in other months.

26. Although all states have income limits high enough to include all poor families, some states impose minimum work requirements that may disqualify some poor working families from eligibility.

27. We do not show confidence intervals for these tables; contact the authors for that information.

Acknowledgements

The authors gratefully acknowledge our HHS/ASPE task officer, Julia Isaacs, for her substantive guidance in developing these estimates. We also thank Joan Turek, the TRIM3 contract's project officer, for her support in the development of the TRIM3 system as a whole.

Appendix

Contents

Appendix Tables:

1. Minimum Number of Work Hours Required for CCDF Eligibility According to October 1999 State Plans
2. Maximum Age of Disabled Child for CCDF Eligibility According to October 1999 State Plans
3. Special Income Calculations for Determining CCDF Eligibility According to October 1999 State Plans
4. Maximum Monthly Income Limit for Initial Eligibility According to October 1999 State Plans
5. Maximum Monthly Income Limit for Continuing Eligibility According to October 1999 State Plans
6. Federal Maximum Monthly Income Limit 85% CY 1997 Median Income