Table of Contents | Previous | Next |
2.0 Existing Findings Related to Served and Unserved Families
2.1 Overview
The purpose of this chapter is to explore existing data sources, both Head Start and national datasets in order to learn more about the characteristics of Head Start-eligible families. The datasets examined include the Head Start Program Information Report (PIR), the Survey on Income and Program Participation (SIPP), the Panel Survey on Income Dynamics (PSID), the National Longitudinal Survey of Youth (NLSY79), while Head Start FACES was used as a reference dataset.
2.1.1 Background
The purpose of this section is to review some recent papers that may provide information on unserved families. Perspectives from these other studies offer insights into the characteristics of low-income families who are not receiving needed services, including those families who may be eligible for, but not enrolled in, Head Start.
The underlying issue related to the interest in unserved families is that of addressing “unmet need,” which Queralt and Witte (1999) described as the gap or disparity between optimal levels of service (services that meet all needs) and actual levels of service provision (Queralt & Witte, 1999). The causes of this gap are varied, attributable to financial, language, or accessibility barriers that prevent families from using existing services, as well as to the failure of available services to meet the actual need for services. A less apparent cause, which actually does not fit this definition of unmet need, is the preference of some unserved families not to participate in government-sponsored programs. In order to learn more about unmet need, this project had an underlying goal to investigate the characteristics of the unserved families, to assess why they may not be enrolled in Head Start, and to identify methods that may be used to recruit and enroll unserved (unenrolled) families.
Community services providers have always been challenged to meet the changing needs of the families they serve, but recent shifts in the cultural and political landscape have created a moving target. Among the notable challenges facing families are the needs for adequate, accessible health services and for quality, affordable child care. In terms of the former, recent evidence shows that publicly-funded health services (Medicaid and Children’s Health Insurance Program - CHIP) are making concerted outreach efforts to reduce the unmet need for health care for children. This has included an inter-agency effort within the Federal government and by national service organizations to inform families of the opportunities CHIP provides them. While both Medicaid and CHIP have seen enrollment increased because of these efforts, these programs still have not reached the levels of service originally anticipated across the country (Edmunds, Teitelbaum, & Gleason, 2000; Kenney, Haley, & Ullman, 1999).
With regards to child care services, several studies conducted on the national and local levels have focused on how families were managing their need for care while either remaining on waiting lists for available slots or while waiting for child care subsidies to pay for care. These studies found that families on waiting lists reflected an unmet need, and often made compromises by using other family members and neighbors as care providers (Casper, 1996; Coonerty & Levy, 1998), sometimes accepting a lower quality of child care than they would have liked (Coonerty & Levy, 1998). In turn, many of the families that were having to pay for care while waiting for subsidies were putting themselves in a risky financial position by going into serious debt (Armson, 1999; Coonerty & Levy, 1998).
Some recent studies have taken a more focused look at Head Start-eligible families. Examining a sample of families taken from the National Child Care Survey of 1990, Hofferth (1994) investigated the characteristics of families who enrolled in Head Start and those who did not. Hofferth noted that a number of factors were associated with a child’s Head Start enrollment, including the education of the mother, race of the child, geographic region, enrollment of the household head in a training program, and standard of living, as well as the supply of child care available in the family’s county of residence. She also noted that the children most likely to be enrolled in Head Start were those living with unemployed parents. Of the 3- to 5-year-old children not enrolled in Head Start, only about one quarter were enrolled in a center-based child care program, with parent or relative care being the most likely options for their children. Among the eligible families, children from the families with the lowest incomes were about as likely to be enrolled in Head Start as children from families with the highest incomes. Parents with children in Head Start also were more likely to have used multiple services than parents who did not enroll their children in Head Start. The implication of this is that the unmet need for many of these non-enrolled families may extend beyond the help Head Start can provide by itself.
Economic disadvantages of the types encountered by Head Start-eligible families have been associated with problems in children’s cognitive and emotional development, as well as with problem behaviors (Huston, McLoyd, & Garcia, 1997; McLoyd, 1998; Vandivere, Moore, & Brown, 2000). Using the Panel Study of Income Dynamics (PSID), Foster (in press) recently examined data collected in the early 1990's, focusing on Head Start-eligible families and the presence of four key disadvantages for those families: poverty, welfare receipt, female head of household, and parental joblessness. He reported that the level of disadvantage was high among enrolled families, particularly among African-American families. In a comparison of enrolled and eligible, non-enrolled families, Foster found that in spite of similarities in neighborhood characteristics, the children from non-enrolled families encountered an even greater number of disadvantages. He noted that this difference has been in gradual decline over time, suggesting that Head Start may be reaching more of the disadvantaged families it is targeting.
In a recent, unpublished paper, Nord (1999) compared data describing Head Start children and families collected for the Head Start FACES study, with data gathered during the March 1997 Current Population Survey (CPS). The CPS describes the characteristics of and their families. When comparing Head Start children with poor children from the general population, the Head Start children were more likely to live in two-parent families findings noted earlier. Head Start children were more likely than poor 3- to 5- year olds to have had a mother with at least some college experience, to have a mother who worked full families that received welfare and food stamps. The author concluded, with appropriate caution, that Head Start children may be somewhat better off than other poor children, and being reached by Head Start, particularly citing the children who lived in single-parent families with parents who had a low education.
Finally, in a study of Head Start recruitment and enrollment efforts, Love and Grover (1987) examined how Head Start programs from a range of communities attempted to meet the challenge of providing services to the most needy children within the targeted low-income population in their service area. In addressing the issues, the following key findings were reported:
-
While there was wide variation among programs, there appeared to be considerable evidence that many programs used intensive outreach and recruitment efforts to recruit the most needy families, and that large proportions of the eligible population were being served.
-
While many programs structured their services to meet what Head Start perceived as the needs of the families in their communities, some families did not enroll or dropped out because they did not see the program meeting their specific needs.
-
When defining “need,” programs’ criteria generally focused on family income; however, programs also considered other factors such as family problems, disabilities and health problems.
-
While half of all drop-outs resulted from family mobility, programs believed that more families could be retained if additional services or longer program days could be provided.
-
If the proportion of enrollees in the “most needy” category increased, programs would likely need additional staff, additional resources to assist multiple-problem families, reconstruction of handicapped accessible facilities, and additional recruitment costs.
Love and Grover (1987) also concluded that “need” is not a straight-forward concept. The diversity of needs staff reported for low-income populations suggested that the criteria for establishing need went well beyond family income. Head Start programs often addressed this by preparing to meet the variety of economic, social, and emotional needs among the families that were enrolled, but this had implications for recruitment by putting the focus on families Head Start could help.
2.2 Head Start PIR
The Head Start Program Information Report (PIR) contains self-report data from every Head Start program. Collected annually, these reports are the only single source of information from all Head Start programs. Data are reported at the program-level, with very limited information on families. For this study, information on proportions of children 1) leaving the program before attending classes, 2) leaving the program at any point after classes began, and 3) remaining in the program for less than 45 days after enrollment was examined. Information is also presented on the percentage of drop-outs replaced during the year and those replaced within 30 days.
The analysis of the PIR was completed using data from the 1993-94, 1995-96, and 1997-98 program years. A total of 1,565 programs, serving children aged 3 years and older, were in operation and reported data for all three of these years. During this period, the average number of children served (aged 3 years and older only) increased 13% per program. The largest percentage increases were for programs in the Western region (19.7%), while the smallest growth was observed in the Midwest (8.3%). The numbers presented in the following exhibits represent the unweighted mean percentages (and standard deviations) as reported across programs.
Exhibit 2-1 contains information regarding the mean percentages across programs of children who left their program before classes or home visits began. This information was not included in the 1997-98 PIR. Each year, about 4% of children left their program after receiving some services (such as health screenings, family needs assessments, etc.), but before the child actually attended classes or received a home visit. The percentages were slightly higher in urban programs (about 5%) than in rural programs (about 3%) and were generally higher in the Northeast (about 5%) and Midwest (just over 6%) than in the South (less than 3%) and West (just over 3%).
| Number of Programs* |
1994 | 1996 | 1998 | |
|---|---|---|---|---|
| Overall | 1565 | 4.0 (7.5) | 4.1 (7.6) | - - - |
| Rural | 688 | 2.6 (6.0) | 3.2 (6.2) | - - - |
| Urban | 877 | 5.0 (8.7) | 4.8 (8.7) | - - - |
| Northeast | 412 | 4.8 (8.7) | 5.5 (9.2) | - - - |
| South | 559 | 2.0 (4.5) | 2.6 (6.6) | - - - |
| Midwest | 319 | 6.6 (11.0) | 6.0 (7.2) | - - - |
| West | 275 | 3.7 (7.5) | 2.8 (7.5) | - - - |
| * Only programs in existence in all years are included in the analysis (that is, programs initiated or defunded after 1994 were removed from the database) |
Exhibit 2-2 contains information regarding the mean percentages of children who left Head Start anytime during the program year. Across all programs, these percentages increased slightly over the period examined, from 13.7% to 14.6%. The percentages for each year were similar across urban programs and rural programs. Again, mean program dropout rates were generally higher in the Northeast and West regions (typically above 15%) than in the South and Midwest regions (14% or less).
| Number of Programs* |
1994 | 1996 | 1998 | |
|---|---|---|---|---|
| Overall | 1565 | 13.7 (6.7) | 14.5 (6.4) | 14.6 (6.5) |
| Rural | 688 | 13.3 (6.5) | 14.2 (5.9) | 14.5 (6.2) |
| Urban | 877 | 13.9 (6.8) | 14.8 (6.8) | 14.8 (6.7) |
| Northeast | 412 | 15.0 (7.5) | 16.1 (6.6) | 16.4 (6.4) |
| South | 559 | 11.9 (5.0) | 12.7 (6.2) | 12.9 (6.6) |
| Midwest | 319 | 13.3 (7.8) | 14.0 (6.8) | 13.7 (6.5) |
| West | 275 | 15.6 (6.3) | 16.8 (6.1) | 16.6 (6.4) |
| * Only programs in existence in all years are included in the analysis (that is, programs initiated or defunded after 1994 were removed from the database) |
Exhibit 2-3 indicates the mean proportions of children who left the program before 45 program-days elapsed, considering only those children who attended at least some classes. This 45 program-day period may vary for some children, as it begins with a child’s first day, not necessarily the first day of the program year. Overall, these percentages declined from 6.3% in 1993-94 to 4.7% in both 1995-96 and 1997-98, a pattern that was reflected in both the urban and rural subsamples. In 1993-94, only the Southern region lost less than 6.0% of enrolled children in less than 45 program-days, while the rates were generally higher in the West (7.4%) and Northeast (6.9%). By 1997-98, all regions were losing less than 6.0% of their children, while programs in the South and West reported losing less than 4.5% of the children during the first 45 days.
| Number of Programs* |
1994 | 1996 | 1998 | |
|---|---|---|---|---|
| Overall | 1565 | 6.3 (4.5) | 4.7 (4.3) | 4.7 (3.5) |
| Rural | 688 | 6.2 (5.1) | 4.6 (4.7) | 4.6 (3.3) |
| Urban | 877 | 6.4 (4.0) | 4.8 (4.1) | 4.7 (3.6) |
| Northeast | 412 | 6.9 (5.1) | 5.5 (6.3) | 5.2 (3.6) |
| South | 559 | 5.4 (4.1) | 4.0 (3.3) | 4.3 (3.4) |
| Midwest | 319 | 6.1 (4.8) | 4.1 (3.7) | 4.2 (3.3) |
| West | 275 | 7.4 (3.8) | 5.7 (4.1) | 5.5 (3.5) |
| * Only programs in existence in all years are included in the analysis (that is, programs initiated or defunded after 1994 were removed from the database) |
As noted earlier, the PIR data do not include family-level data and, therefore, provide no insights into why these families left the program (either before or after the start of classes). Overall, the means do not suggest any strong patterns of differences based on urbanicity or geographic region.
The final two PIR exhibits contain information about the replacement of program drop-outs. Exhibit 2-4 indicates the mean percentages of drop-out children who were replaced during the program year (ranging from 85.6% to 88.0% over the years examined), while Exhibit 2-5 presents the mean percentages of those replacements that were completed within 30 days or less (approximately 94% in each year). In both 1995-96 and 1997-98, programs in urban areas seemed to have slightly greater success at replacing drop-outs than rural programs (88.0% versus 84.9%). During the time frame, programs in the Northeast became increasingly less successful in replacement (declining from 85.4% to 80.9%), while programs in the South became increasingly more successful (80.2% in 1993-94 to 86.8% in 1997-98). Programs in all regions, and in both rural and urban areas, consistently completed between 95.3% and 91.1% of their replacements within 30 days.
| Number of Programs* |
1994 | 1996 | 1998 | |
|---|---|---|---|---|
| Overall | 1565 | 85.6 (6.4) | 88.0 (6.0) | 86.6 (6.0) |
| Rural | 688 | 85.3 (5.6) | 86.0 (5.2) | 84.9 (5.4) |
| Urban | 877 | 85.8 (6.8) | 89.6 (6.6) | 88.0 (6.5) |
| Northeast | 412 | 85.4 (6.2) | 83.0 (6.0) | 80.9 (6.3) |
| South | 559 | 80.2 (5.4) | 87.9 (5.2) | 86.8 (5.6) |
| Midwest | 319 | 91.9 (6.8) | 92.0 (6.3) | 89.4 (6.1) |
| West | 275 | 89.7 (6.9) | 91.1 (6.1) | 91.5 (6.0) |
| * Only programs in existence in all years are included in the analysis (that is, programs initiated or defunded after 1994 were removed from the database) |
| Number of Programs* |
1994 | 1996 | 1998 | |
|---|---|---|---|---|
| Overall | 1565 | 94.6 (4.6) | 93.5 (4.8) | 93.5 (4.5) |
| Rural | 688 | 94.5 (4.5) | 93.5 (4.2) | 93.7 (4.2) |
| Urban | 877 | 94.6 (4.6) | 93.4 (5.2) | 93.3 (4.7) |
| Northeast | 412 | 93.9 (4.5) | 92.2 (4.7) | 93.2 (4.4) |
| South | 559 | 94.7 (4.1) | 94.9 (4.8) | 94.1 (4.6) |
| Midwest | 319 | 94.9 (4.8) | 91.3 (4.9) | 91.1 (4.5) |
| West | 275 | 94.9 (4.2) | 95.0 (4.3) | 95.3 (4.4) |
| * Only programs in existence in all years are included in the analysis (that is, programs initiated or defunded after 1994 were removed from the database) |
Overall, while the universe of Head Start programs is generally successful in quickly replacing children who leave the program, the mean percentages presented in Exhibit 2.4 suggest that additional insights are needed regarding retention and the use of waiting lists by Head Start. The perspectives of program staff on these particular issues are reported in Chapter 3.
2.3 Head Start FACES Program Drop Reports and Parent Exit Interviews
During the periods of the fall of 1997 through the spring of 1999, data were collected on a sample children who dropped out from the 40 Head Start programs participating in FACES. Program Drop Reports, completed by staff at each of the FACES sites, were received for a total of 611 children. Of the 611 sample children who dropped out from the program, 47% were male and 53% were female. Approximately 34% attended Head Start in the morning only, 24% in the afternoon only, and 32% attended Head Start for six hours or more daily. The most frequent reason staff reported for a child dropping out of the program was that the family had moved (42.1%). When sample children were reported to have dropped from the program, FACES interviewers attempted to conduct an exit interview with their primary caregivers. Only a small proportion of the families that had moved could be located and interviewed. Among the families who were contacted, a total of 156 exit interviews were completed. Therefore, exit interviews were conducted with 25.5% of parents of all sample children who dropped out of the programs, and 44.1% of the children who according to staff left the program early for a reason other than a family relocation.
2.3.1 Reasons for Leaving Head Start.
While a family move was the most frequent reason cited by program staff for a family leaving Head Start (42.1%), the other frequently noted reason was that the child moved to another program or kindergarten (23.3%). The following reasons were cited by less than 8% of the respondents: poor attendance, family or personal problems, transportation problems, and scheduling conflicts (Exhibit 2-6). Programs very rarely reported that a child was withdrawn due to parental dissatisfaction with one or more elements of the program.
These results were largely confirmed by primary caregivers participating in exit interviews. Among the 156 families for whom exit interviews were completed and no relocation had taken place, frequently noted reasons for leaving Head Start were: 1) transferred to another program (38.2%), 2) transportation problems (29.5%), 3) parent problems with the program (15.2%), usually difficulties with a staff member, and 4) a child’s refusal to attend (6.8%). Poor attendance, family and personal problems, and scheduling conflicts were rarely mentioned (each less than 8%).
Although the exit interview sample was not large enough to support strong conclusions, the differences between staff drop-out reports and parent exit interviews for families leaving but not relocating indicated that neither respondent group reported many issues that reflected negatively on Head Start. It seems possible that Head Start staff may have underestimated both transportation difficulties and parent dissatisfaction with some element of the program as reasons parents had for withdrawing their children from the program. Parents, on the other hand, were less likely than staff to report the withdrawal of their child due to family or personal problems or because of poor attendance. In this case, program staff seemed more likely to put the responsibility on parents, while parents were more likely to cite program deficiencies.
| Reason for Withdrawal from Head Start |
All children leaving the program | Percentage of children leaving for a reason other than family relocation | |
|---|---|---|---|
| Reported by Head Start
staff (N = 611) % |
Reported by Head Start staff (N = 354) % |
Reported by parents (N = 156) % |
|
| Family relocation | 42.1 | - | _ _ _ |
| Changed program or entered kindergarten | 23.3 | 40.3 | 38.2 |
| Poor attendance | 7.3 | 12.7 | 1.7 |
| Family/personal problems | 6.3 | 10.8 | 0.9 |
| Transportation problems | 5.6 | 9.6 | 29.5 |
| Parent scheduling conflicts with work or school | 4.5 | 7.8 | 3.5 |
| Dissatisfaction with program | 4.2 | 7.2 | 15.2 |
| Child maturity/refusal to attend | 0.0 | 0.0 | 6.8 |
| Other/not specified | 6.7 | 11.5 | 4.2 |
2.3.2 Satisfaction with Head Start.
Satisfaction with the program has been repeatedly reported to be very high among parents with children attending Head Start (ASCI, 1999), and interviews conducted with parents of children in the FACES sample in the spring of 1997, 1998, and 1999 confirmed these findings. However, the exit interviews were conducted, in part, to explore the possibility that parents of children who left the program may have less favorable views of Head Start. As shown in Exhibit 2-7, the parents of children who withdrew from the program were generally satisfied with Head Start. Nevertheless, the percentages of parents reporting dissatisfaction was higher for each issue addressed in the interview than the percentages reported by the parents of children who remained in the program (Exhibit 2-8).
| How satisfied are you with
how well Head Start is. |
Very dissatisfied % |
Somewhat dissatisfied % |
Somewhat satisfied % |
Very Satisfied % |
Don't Know % |
|---|---|---|---|---|---|
| Helping child to grow and develop | 9.3 | 4.7 | 20.0 | 64.0 | 2.0 |
| Open to your ideas and participation | 4.7 | 2.7 | 18.0 | 70.7 | 4.0 |
| Supporting and
respecting your family's culture and background |
5.3 | 0.7 | 8.7 | 83.3 | 2.0 |
| Identifying and
providing services for child - for example, health screening, help with speech and language development |
2.7 | 1.3 | 12.7 | 76.0 | 7.3 |
| Identifying and
helping to provide services that help your family-for example, public assistance, transportation, or job training |
8.7 | 9.4 | 10.7 | 49.0 | 22.1 |
| Maintaining a
safe program - for example, secure playgrounds, clean and tidy classrooms |
2.7 | 2.7 | 8.7 | 84.7 | 1.3 |
| Preparing child to enter kindergarten | 13.3 | 5.3 | 11.3 | 63.3 | 6.7 |
| Helping you become more involved in groups that are active in your community | 6.0 | 2.0 | 16.7 | 49.3 | 26.0 |
| How satisfied are you with how well Head Start is. | Parents of children withdrawn from the program (N = 156) % |
Parents of children remaining in the program (FACES Spring, 1998 N = 2,688) % |
|---|---|---|
| Helping child to grow and develop | 14.0 | 1.9 |
| Open to your ideas and participation | 7.4 | 2.1 |
| Supporting and respecting your family's culture and background | 6.0 | 1.3 |
| Identifying and providing services for child- for example, health screening, help with speech and language development | 4.0 | 1.6 |
| Identifying and helping to provide services that help your family-for example, public assistance, transportation, or job training | 18.1 | 3.3 |
| Maintaining a safe program-for example, secure playgrounds, clean and tidy classrooms | 5.4 | 2.0 |
| Preparing child to enter kindergarten | 18.6 | 2.6 |
| Helping you become more involved in groups that are active in your community | 8.0 | 3.1 |
The areas of greatest dissatisfaction among primary caregivers of children withdrawn from Head Start involved child development services (helping child to grow and develop, preparing child for kindergarten) and identifying and facilitating family services from community agencies. In each of these areas, nearly one in six parents of children withdrawn from the program indicated dissatisfaction with Head Start. Comments from the parents who expressed dissatisfaction generally indicated that the program did not meet their expectations. While this is not a large sample, the sample participating in the exit interviews do reflect perceptions among a minority of families that Head Start services could be improved.
2.4 Head Start FACES Staff: Social Service Coordinators and Center Directors
As part of the FACES data collection, several staff members were interviewed at each program. In particular, Social Service Coordinators (N = 40) and Center Directors (N = 123) were asked several questions about enrollment and waiting list activities. Their responses are summarized below.
2.4.1 Social Service Coordinators.
All of the 40 Social Service Coordinators (SSCs) who were interviewed indicated that their program had a list of families waiting to enter Head Start, and 38 of these programs maintained these waiting lists at individual centers (Exhibit 2-9). The SSCs reported information from a program-wide perspective. For the 33 SSCs who provided an actual waiting list size for their program, the average was 374 children; however, this figure was greatly impacted by four large programs. The median size of the program waiting lists was just over 100 children per program. For the 29 SSCs reporting the percentage of children who were enrolled from their waiting list, the average was 146 children, or just under 40% of the reported waiting lists. The median percentage enrolled was about 55%, with higher proportions generally reported for smaller programs.
| Prog | Funded Enrollment (1996) |
Wait List Size |
Number Enrolled from List |
Percent Enrolled by Program |
|---|---|---|---|---|
| 1 | 218 | 45 | 25 | 55.6 |
| 2 | 1163 | 300 | 200 | 66.7 |
| 3 | 607 | 126 | - - - | - - - |
| 4 | 511 | 180 | 45 | 25.0 |
| 5 | 4376 | 1200 | 300 | 25.0 |
| 6 | 1124 | 30 | 15 | 50.0 |
| 7 | 2685 | 162 | 162 | 100.0 |
| 8 | 371 | 171 | 78 | 45.6 |
| 9 | 786 | 40 | 10 | 25.0 |
| 10 | 458 | dk | - - - | - - - |
| 11 | 243 | 34 | 19 | 55.9 |
| 12 | 2153 | 300 | 150 | 50.0 |
| 13 | 314 | DK | - - - | - - - |
| 14 | 215 | 65 | 35 | 53.8 |
| 15 | 1075 | 50 | 50 | 100.0 |
| 16 | 278 | 45 | 30 | 66.7 |
| 17 | 1994 | 3000 | 1000 | 33.3 |
| 18 | 3968 | DK | - - - | - - - |
| 19 | 356 | 35 | 14 | 40.0 |
| 20 | 189 | 25 | - - - | - - - |
| 21 | 422 | 167 | 2 | 1.2 |
| 22 | 1514 | DK | - - - | - - - |
| 23 | 292 | 130 | 40 | 30.7 |
| 24 | 214 | 30 | 27 | 90.0 |
| 25 | 737 | 150 | 100 | 66.7 |
| 26 | 1195 | 250 | 150 | 60.0 |
| 27 | 240 | 100 | 20 | 20.0 |
| 28 | 578 | 30 | 30 | 100.0 |
| 29 | 542 | 50 | 50 | 100.0 |
| 30 | 1140 | DK | - - - | - - - |
| 31 | 1080 | DK | - - - | - - - |
| 32 | 1023 | DK | - - - | - - - |
| 33 | 246 | 172 | 70 | 5.8 |
| 34 | 460 | 4000 | 1000 | 25.0 |
| 35 | 381 | 25 | 25 | 100.0 |
| 36 | 635 | 66 | 50 | 75.8 |
| 37 | 368 | 45 | 45 | 100.0 |
| 38 | 662 | 240 | - - - | - - - |
| 39 | 3171 | 678 | 384 | 56.6 |
| 40 | 212 | 450 | - - - | - - - |
| Wait List Totals |
N = 33 Mean = 374 S.D. = 827.8 |
Number Enrolled From Wait List |
N = 29 Mean = 146 S.D. = 250.7 |
2.4.2 Center Directors.
Each Center Director reported information for his/her center only, and within programs, only those centers participating in Head Start FACES are represented. Of the 123 Center Directors (CDs) reporting, 83% reported having a center-specific waiting list (Exhibit 2-10). Of those CDs reporting the waiting list size for their centers, the number of children on these lists ranged from 2 to 71 names (mean = 22.4; median = 20).
| Program | Centers without wait list |
Centers with wait list |
Wait List Size |
|---|---|---|---|
| 1 | 0 | 4 | 5,5,15,-- |
| 2 | 3 | 1 | |
| 3 | 0 | 3 | 28,-.-- |
| 4 | 0 | 4 | 3,5,-,-- |
| 5 | 0 | 3 | 8,25,54 |
| 6 | 0 | 1 | |
| 7 | 2 | 1 | |
| 8 | 2 | 2 | 6,-- |
| 9 | 3 | 1 | |
| 10 | 0 | 4 | 4,4,5,-- |
| 11 | 1 | 3 | 6,20,-- |
| 12 | 1 | 0 | |
| 13 | 0 | 4 | 14,30,40,50 |
| 14 | 0 | 3 | 20,30,-- |
| 15 | 0 | 3 | 24,71,-- |
| 16 | 1 | 3 | 3,4,7 |
| 17 | 0 | 4 | 6,10,25,30 |
| 18 | 0 | 4 | 30,-,-,-- |
| 19 | 0 | 1 | 35 |
| 20 | 0 | 1 | 22 |
| 21 | 0 | 2 | 20,40 |
| 22 | 0 | 3 | 3,-,-- |
| 23 | 0 | 3 | -,-,-- |
| 24 | 0 | 0 | |
| 25 | 1 | 2 | 6,10 |
| 26 | 1 | 3 | 30,-,-- |
| 27 | 0 | 4 | -,-,-,-- |
| 28 | 2 | 2 | 5,10 |
| 29 | 1 | 2 | 30,-- |
| 30 | 0 | 4 | 8,19,20,50 |
| 31 | 0 | 4 | 2,30,43,65 |
| 32 | 1 | 0 | |
| 33 | 0 | 3 | -,-,-- |
| 34 | 1 | 3 | 7,50,60 |
| 35 | 1 | 3 | 30,-,-- |
| 36 | 0 | 4 | 50,-,-,-- |
| 37 | 0 | 4 | 5,5,12,60 |
| 38 | 0 | 2 | 15,-- |
| 39 | 0 | 4 | 39,40,50,-- |
| 40 | 0 | 0 |
| Centers | Wait List Sizes |
|
|
2.5 Family/Household Databases Overview
In order to learn what might be known regarding the differences between enrolled and non-enrolled eligible families, project staff reviewed three existing national databases. The project team and expert consultants determined that the following datasets would be targeted: The National Longitudinal Study of Youth 1979 (NLSY79), the Survey of Income and Program Participation (SIPP), and the Panel Survey of Income Dynamics (PSID). These datasets were selected because they offered some comparability along several dimensions with data collected during the FACES study, and, most importantly, each contained a variable identifying Head Start participation by a 3- to 5-year-old child. The focus in this review was on examining information regarding the following: family description, education, employment, family income and program participation, housing, health status and insurance coverage, and child care arrangements.
While each dataset has strengths in describing particular characteristics of the families of interest, collectively, their inherent differences do not allow direct comparisons across datasets, or with the FACES data. For example, these datasets do not all report upon the same time period. A second, and perhaps more important example, is that while FACES collected data on the characteristics of the primary caregiver, the target respondent was not consistent across these datasets. The person whose characteristics were reported upon in each datasets was:
-
FACES - the primary caregiver of the child,
-
NLSY79 – the mother of the child,
-
SIPP – the designated parent or guardian of the child (as reported by the SIPP respondent), and
-
PSID – the head of the household.
Such distinctions are critical, and caution is advised with regards to making direct comparisons across datasets. A detailed description of each of these datasets is provided in Appendix A.
2.5.1 Findings From the Datasets
The presentation of weighted findings from each dataset is accompanied by an exhibit highlighting the findings from that dataset. Each exhibit has six data columns. The first contains data for all families with 3- to 5-year-old children in the sample. The second column features unweighted data from the Head Start FACES baseline data collection, which serves as a point of reference for other Head Start findings. The third column focuses on weighted data for all families in the dataset reporting a Head Start child. The second set of three columns contain weighted data only for families determined to be eligible for Head Start. In cases where the dataset did not have a variable matching Head Start FACES and at least one other dataset, the variable remains in the exhibit with a line indicating no data were available.
Since the purpose of the study is to learn more about eligible non-enrolled families, this discussion focuses on the two right-most columns in the accompanying exhibits, presenting the characteristics of eligible enrolled and non-enrolled families (significant differences between the groups at p < .05 are noted in the exhibits). Head Start eligibility was determined by applying the income and other program participation criteria specified in the Head Start regulations. A discussion of the Head Start criteria for program eligibility, along with a copy of the 1996 Head Start Family Income Guidelines memorandum is presented in Appendix B.
In reviewing the exhibits that follow, note that the first set of columns represents findings on all children, and all children reported to be enrolled in Head Start, without selection based on Head Start eligibility. This group may include over-income families who have enrolled children with disabilities or some other special needs, as well as families who passed the income criteria between the time of Head Start enrollment and their response during the data collection for the survey in which they are participating. This is to say that responses to the surveys may not reflect levels of a family’s need as recorded by Head Start at the time of that family’s enrollment.
For the NLSY79 and the SIPP, sample characteristics were estimated using person-level weights that were provided with each dataset. The weights for the SIPP were constructed using Current Population Survey (CPS) estimates of the non-institutionalized United States population by age, gender, race and Hispanic ethnicity. The CPS estimates were derived from data collected as part of the 1990 decennial Census. The NLSY79 child weights are based on the NLSY79 mother weights, which were derived from the CPS of 1978 (the original sample was drawn in 1979 and included young people living in the United States that were between the ages of 14 and 21 as of December 31, 1978). Child weights were adjusted to account for differences in age, race and gender. Finally, the weights provided with the PSID were based on a combination of factors: 1) an adjustment for differences in sample selection probabilities for sample individuals and families; 2) an adjustment for non-response attrition and death; and 3) an adjustment to post-stratify weighted sample distributions for demographic and household variables to population values measured in the 1990 decennial Census and the CPS.
Unweighted findings from the nationally representative Head Start FACES study are referenced in each of the specific dataset discussions, serving as a context for interpreting the findings presented for that dataset. The shaded column in the tables contains Head Start FACES baseline data collected in the fall of 1997.
NLSY79. Exhibit 2-11 presents findings from all children age 3- to 5-years old and for the Head Start-eligible children in the NLSY79 sample. The family composition of enrolled and non-enrolled families was similar, with each having a mean of 4.5 family members, but with enrolled families reporting slightly more children under 18 years of age (3.3 compared with 3.0). Significantly higher percentages of White (non-Hispanic) and Black (non-Hispanic) children were in the enrolled group, which subsequently had significantly fewer American Indian, Eskimo, Aleut, and Hispanic children than the non-enrolled group. Among FACES respondents, Black (non-Hispanic) children comprised the largest component (36.9%) of the sample.
The original cohort of women included in the NLSY79 was between the ages of 14 and 21 years of age on January 1, 1979, and, consequently, was between the ages of 31 and 38 during the 1996 survey period. The mean age of these women, 34.6 years, was the same for both groups. A significantly lower percentage of women in enrolled families was married, while the non-enrolled families included significantly fewer formerly or never married women. The mean age of the primary caregivers responding to the FACES survey was 29.7 years, and just under half (42.7%) were married.
A significantly lower percentage of mothers from enrolled families (33.6%) had a high school diploma than did mothers from non-enrolled families (35.4%), while significantly more of the mothers in the enrolled group had some college education. The level of educational attainment for Head Start FACES respondents was similar to that of the enrolled and the non-enrolled NLSY79 respondents. With regards to ongoing education, approximately 3.6% of mothers from non-enrolled families were currently in school or training, significantly greater than the 3.0% of mothers from enrolled families who reported that they were receiving ongoing education. By contrast, almost one quarter (22.8%) of the FACES respondents were in school or training at the time of their interview.
While two thirds of the mothers from enrolled families (67.3%) reported being either unemployed or not in the labor force, compared with 61.0% among the mothers from non-enrolled families, significantly more of the mothers from enrolled families (26.7%) reported full-time employment than their non-enrolled counterparts (23.9%). Approximately one third of NLSY79 mothers from enrolled (32.6%) and non-enrolled (35.3%) families reported full- or part-time employment, and over half (51.8%) of FACES respondents were fully- or partially- employed.
Although the mean annual family income was higher for enrolled families than non-enrolled families ($12,593 vs. $11,296), this difference was not statistically significant. The median income was less for enrolled families ($8,748 vs. $10,368). When family income was compared against the Federal Poverty Level (FPL), substantial proportions of both enrolled (80.2%) and non-enrolled (84.1%) families reported incomes less than the FPL. This was to be expected since the FPL is a major factor in determining Head Start eligibility. A significantly greater proportion of non-enrolled families (41.7%), however, reported family incomes less than 50% of the FPL than did enrolled families (23.4%). While just over one half (53.8%; 16.9% below 50% of the FPL) of the FACES respondents reported incomes below the FPL, the FACES data were based on the incomes of all those living in the household.
In general, significantly larger proportions of enrolled families reported participation in Federal support programs, i.e., Aid to Families with Dependent Children (AFDC), Women, Infants, and Children (WIC) nutrition program, food stamps, and public/subsidized housing programs. The exception to this was that significantly more non-enrolled families reported receiving Supplemental Security Income (SSI). Head Start FACES families generally reported lower rates of participation in these family support programs than enrolled families in the NLYS79.
Significantly more enrolled families reported child health insurance coverage, especially Medicaid (81.7%), than non-enrolled families (57.3%), and although the percentages were very low, a significantly lower percentage of enrolled families reported their Head Start children had a disability. This is particularly surprising, given Head Start’s focus on enrolling disabled children and on screening all enrolled children.
The picture that emerges from the NLSY79 data is that fewer married and formerly married women have children participating in Head Start, a larger proportion of non-enrolled mothers report incomes less than 50% of the FPL, and enrolled families generally report more participation in Federal support programs (e.g., AFDC, WIC) than do non-enrolled families. The NLSY79 data do not answer the question of whether these differences reflect the efforts of Head Start. It must be remembered, however, that the NLSY79 data are not representative of all American women. They are representative of a national sample of women who were between the ages of 31 and 38 on January 1, 1996.
| Dataset: NLSY79 Children: 1996 Survey (Weighted) | ||||||
|---|---|---|---|---|---|---|
| (Numbers in thousands) | All Children Age Age 3- to 5-years |
Head Start Eligible
Children Age 3- to 5-years |
||||
| Characteristic | Total (n = 4,445) |
FACES Fall 1997 (n = 3,156) |
Enrolled
in Head Start (n = 184) |
Total (n = 609) |
Enrolled
in Head Start (n = 108) |
Not enrolled in Head Start (n = 500) |
| Family Description | ||||||
| Number in household 12 | ||||||
| Number in family | Mean = 4.3 | Mean = 4.5 | Mean = 4.4 | Mean = 4.5 | Mean = 4.5 | Mean = 4.5 |
| Number of children under 18 years | Mean = 2.4 | Mean = 2.6 | Mean = 2.9 | Mean = 3.0 | Mean = 3.3 | Mean = 3.0 |
| Child race or ethnicity | ||||||
| White (non-Hispanic) | 75.9% | 27.7% | 49.8% | 44.7% | 48.3% | *43.9% |
| Black (non-Hispanic) | 10.9 | 36.9 | 33.5 | 34.4 | 42.6 | *32.7 |
| American Indian, Eskimo, Aleut | 5.7 | 1.9 | 7.2 | 10.8 | 5.9 | *11.8 |
| Asian or Pacific Islander | 0.4 | 1.1 | 0.0 | 0.0 | 0.0 | 0.0 |
| Hispanic | 6.2 | 24.2 | 9.5 | 10.1 | 3.2 | *11.6 |
| Age of primary caregiver13 | ||||||
| Less than 20 years | 2.4% | |||||
| 20-29 years | 59.3 | |||||
| 30-39 years | 29.0 | |||||
| 40 years and older | 9.3 | |||||
| Mean=34.5 | Mean =29.7 | Mean=34.3 | Mean=34.6 | Mean=34.6 | Mean=34.6 | |
| Median =28.0 | ||||||
| Marital status | ||||||
| Married | 79.9% | 42.2% | 31.8% | 26.2% | 12.5% | * 29.1% |
| Formerly married | 13.0 | 21.3 | 45.4 | 42.4 | 52.5 | *40.2 |
| Never married | 7.9 | 36.0 | 22.8 | 31.4 | 35.0 | *32.7 |
| Education | ||||||
| Primary caregiver's education14 | ||||||
| Less than high school diploma | 11.1% | 27.6% | 32.0% | 33.1% | 30.9% | *33.6% |
| High school diploma/GED | 32.1 | 36.5 | 33.9 | 35.0 | 33.6 | *35.4 |
| Some college | 56.6 | 33.5 | 34.1 | 31.5 | 35.5 | * 30.6 |
| Bachelor's degree or higher | --------- | 2.8 | --------- | --------- | --------- | --------- |
| Currently in school or training | 4.4% | 22.8% | 2.5% | 3.5% | 3.0% | *3.6% |
| Employment | ||||||
| Employment status15 | ||||||
| Employed: Full-time | 46.0% | 34.4% | 41.7% | 24.4% | 26.7% | *23.9% |
| Employed: Part-time | 16.4 | 17.4 | 9.8 | 10.4 | 5.9 | *11.4 |
| Unemployed | 3.6 | 19.7 | 8.9 | 10.0 | 14.1 | *9.1 |
| Not in labor force | 31.9 | 28.0 | 39.6 | 52.2 | 53.2 | *51.9 |
| Multiple jobs concurrently | 1.6% | 3.9% | 0.0% | 0.5% | 0.0% | 0.5% |
| Family Income and Program Participation | ||||||
| Total family annual income16 | ||||||
| Mean | $78,436 | $14,907 | $19,876 | $11,522 | $12,593 | $11,296 |
| Median | $49,000 | $13,200 | $14,400 | $10,000 | $8,748 | $10,368 |
| Federal Poverty Level (FPL)17 | ||||||
| Income < 50% of FPL | 5.3% | 16.9% | 15.4% | 38.5% | 23.4% | *41.7% |
| Income btwn 50-99% of FPL | 6.1 | 36.9 | 37.5 | 44.9 | 56.8 | *42.4 |
| Income btwn 100- 199% of FPL | 15.7 | 36.2 | 25.4 | 12.3 | 12.3 | 12.3 |
| Income 200% or greater than FPL | 72.9 | 10.0 | 21.7 | 7.5 | 7.5 | *3.5 |
| Other sources of support | ||||||
| AFDC | 6.1% | 30.6% | 41.0% | 43.9% | 69.5% | *38.3% |
| Supplementary Security Income | 2.4 | 13.4 | 6.0 | 16.8 | 10.2 | *18.2 |
| WIC | 12.4 | 56.1 | 49.0 | 49.3 | 68.7 | *45.1 |
| Food Stamps | 9.1 | 50.2 | 49.3 | 55.6 | 80.2 | *50.2 |
| Housing | ||||||
| Public or subsidized Housing | 3.8% | 22.1% | 14.5% | 20.2% | 23.7% | *19.4% |
| Number of moves in previous 12 months18 | ||||||
| None | --------- | 64.3% | --------- | --------- | --------- | --------- |
| One or more | 35.2 | |||||
| Health Status/Insurance Coverage | ||||||
| Primary caregiver health status | ||||||
| Excellent | 21.3% | |||||
| Very good | --------- | 28.4 | --------- | --------- | --------- | --------- |
| Good | 33.8 | |||||
| Fair | 14.4 | |||||
| Poor | 2.2 | |||||
| Child health status | ||||||
| Excellent | 44.8% | |||||
| Very good | --------- | 30.5 | --------- | --------- | --------- | --------- |
| Good | 18.0 | |||||
| Fair | 6.2 | |||||
| Poor | 0.4 | |||||
| Child disability | 1.0% | 17.6% | 0.6% | 1.9% | 1.1% | *2.1% |
| Child health insurance coverage 19 | ||||||
| Private | 86.9% | 31.4% | 65.7% | 46.7% | 51.2% | * 45.7% |
| Medicaid | 12.1 | 59.3 | 56.3 | 61.7 | 81.7 | *57.3 |
| Other | ----- | ----- | ----- | ----- | ----- | ----- |
| Child Care Arrangements | ||||||
| Arrangements | ||||||
| Family/friend | 19.2% | |||||
| Family day care | --------- | 2.4 | --------- | --------- | --------- | --------- |
| Center-based day care | 6.1 | |||||
| Hours/week in primary arrangement | --------- | Mean = 19.2 | --------- | --------- | --------- | --------- |
SIPP. Exhibit 2-12 presents findings from the SIPP for children aged 3- to 5-years. The table follows the same format as the NLSY79 table, and includes FACES’ data in the shaded column. As with the NLSY79, the discussion will focus on the two right-most columns, comparing the characteristics of eligible enrolled and non-enrolled families. The respondent for the SIPP is the designated parent or guardian of the Head Start age child(ren).
The average number of family members for enrolled families (4.3 total; 2.8 under 18 years) was similar to that reported by non-enrolled families. Almost half (49.1%) of the enrolled children were Black (non-Hispanic), while less than a quarter (23.1%) of the non-enrolled children were Black (non-Hispanic), a significant difference. Subsequently, the non-enrolled group included significantly more White, Asian/Pacific Islander, and Hispanic children than the enrolled group. The enrolled group was generally similar to the FACES sample on distribution of race and ethnicity.
Almost two thirds (64.7%) of the primary caregivers from enrolled families were in the 20- to 29-year old age range compared with about half (51.3%) of the parents from the non-enrolled families, another significant difference. The mean age (29.3 years) and median age (27.0 years) for respondents from enrolled families were both lower than but not significantly different from their non-enrolled counterparts (mean = 30.7 years; median = 29.0 years). The ages of the primary caregivers in the FACES study were in the same range (mean = 29.7 years, and median = 28.0 years).
Only 20.7% of the parents from enrolled families reported being married while most were either formerly married (24.9%) or never married (54.4%). Conversely, among the parents from non-enrolled families, 45.9% reported being married, 27.0% reported never being married, and 27.1% were formerly married. These differences between the enrolled and non-enrolled groups for the percentages of parents who were married and who were never married were significant. Data from the FACES primary caregivers tended to resemble the latter pattern, with 42.7% reporting being married, 36.0% reporting never being married, and only 21.3% formerly married.
With regards to the education of primary parents, the patterns found among both enrolled and non-enrolled families were quite similar. About two fifths of both groups reported less than a high school diploma, and about one third reported having a diploma or GED. Significantly more of the parents (25.8%) from non-enrolled families reported attending some college than did parents from enrolled families (20.5%). Only about one quarter (27.6%) of FACES primary caregivers reported less than a high school degree and over one third (36.3%) reporting some college or a college degree or higher.
Similar patterns of employment status emerged across the two target categories, with approximately one quarter of both enrolled (25.8%) and non-enrolled (25.5%) reporting full-or part-time employment. Just over half (51.8%) of FACES respondents reported full- or part-time employment.
Among SIPP respondents, the non-enrolled families generally reported significantly higher incomes (mean = $12,928; median = $10,200) than enrolled families (mean = $10,649; median = $9,930). FACES respondents reported still higher incomes (mean = $14,097; median = $13,200) than the SIPP respondents, but FACES data reflect household income, while SIPP data reflect family income. When income and family size data were measured against the FPL, over four fifths of both enrolled (82.6%) and non-enrolled (85.9%) respondents reported family incomes below the FPL. The distribution of incomes below the FPL differs between the two groups, however. A significantly larger proportion of respondents from enrolled families (42.3%) reported family income below 50% of the FPL than did their non-enrolled counterparts (34.9%). Only 16.9% of the FACES respondents reported household incomes below 50% of the FPL. With respect to other sources of support, the respondents from enrolled SIPP families uniformly reported significantly higher rates of participation across the five support programs (including Medicaid and public/subsidized housing) than did respondents from non-enrolled families.
When asked about the health status of their children, 51.6% of SIPP respondents from enrolled families reported that the health status was excellent or very good, significantly less than the 63.0% of respondents from non-enrolled families, and less than the 75.3% of FACES respondents answering the same question.
In terms of child care arrangements, parents from both enrolled and non-enrolled families reported similar patterns: Slightly more than two fifths used a family member or friend and between one fifth and one quarter used center-based day care. The percentages of Head Start FACES families reporting child care use were lower across all categories, but this finding was expected because in the FACES interview, respondents were asked about the use of child care in addition to Head Start.
In summary, among the designated parents or guardians of the Head Start-age children who constituted the SIPP respondent sample, the respondents from enrolled families tended to be younger and less often married. However, respondents from both enrolled and non-enrolled families reported similar education and employment patterns. While non-enrolled families reported higher mean and median incomes, a greater proportion also reported incomes below 50% of the FPL and lower participation in support programs than did enrolled families. While one may expect that Head Start had an impact on this finding, the available data were not able to show if these differences in program participation were due to the efforts of Head Start.
| Dataset: SIPP 1993 Panel, Wave 9 (Weighted) | ||||||
|---|---|---|---|---|---|---|
| (Numbers in thousands) | All
Children Age Age 3- to 5-years |
Head
Start Eligible Children Age 3- to 5-years |
||||
| Characteristic | Total (n = 11,226) |
FACES
Fall 1997 (n = 3,156) |
Enrolled in
Head Start (n = 678) |
Total (n = 3,075) |
Enrolled in Head Start (n = 418) | Not enrolled
in Head Start (n = 2,656) |
| Family Description | ||||||
| Number in household21 | ||||||
| Number in family | Mean = 4.3 | Mean = 4.5 | Mean = 4.4 | Mean = 4.5 | Mean = 4.3 | Mean = 4.6 |
| Number of children under 18 years | Mean = 2.5 | Mean = 2.6 | Mean = 2.7 | Mean = 2.9 | Mean = 2.8 | Mean = 3.0 |
| Child race or ethnicity | ||||||
| White (non-Hispanic) | 67.2% | 27.7% | 39.3% | 42.4% | 28.9% | *44.5% |
| Black (non-Hispanic) | 13.3 | 36.9 | 39.8 | 26.6 | 49.1 | *23.1 |
| American Indian, Eskimo, Aleut | 0.8 | 1.9 | 1.0 | 1.2 | 1.5 | 1.1 |
| Asian or Pacific Islander | 3.7 | 1.1 | 4.1 | 4.0 | 2.3 | *4.2 |
| Hispanic | 15.1 | 24.2 | 16.0 | 26.0 | 18.2 | *27.1 |
| Age of primary caregiver22 | ||||||
| Less than 20 years | 0.4% | 2.4% | 2.0% | 1.0% | 3.2% | *0.7% |
| 20-29 years | 36.3 | 59.3 | 61.6 | 53.1 | 64.7 | *51.3 |
| 30-39 years | 53.4 | 29.0 | 29.2 | 35.8 | 24.5 | *37.5 |
| 40 years and older | 10.9 | 9.3 | 8.3 | 10.1 | 7.6 | *10.5 |
| Mean | 32.1 | 29.7 | 29.8 | 30.5 | 29.3 | 30.7 |
| Median | 32.0 | 28.0 | 28.0 | 29.0 | 27.0 | 29.0 |
| Marital status | ||||||
| Married | 76.9% | 42.7% | 40.1% | 42.5% | 20.7% | *45.9% |
| Formerly married | 11.9 | 21.3 | 21.15 | 26.7 | 24.9 | 27.0 |
| Never married | 11.2 | 36.0 | 38.78 | 30.8 | 54.4 | *27.1 |
| Education | ||||||
| Primary caregiver's education23c | ||||||
| Less than high school diploma | 18.3% | 27.6% | 28.1% | 39.7% | 40.8% | 39.6% |
| High school diploma | 33.5 | 36.5 | 42.3 | 35.2 | 38.8 | 34.6 |
| GED | ----- | ----- | ----- | ----- | ----- | ----- |
| Some college | 48.3 | 33.5 | 29.6 | 25.1 | 20.5 | *25.8 |
| Bachelor's degree or higher | ----- | 2.8 | ----- | ----- | ----- | ----- |
| Currently in school or training | 6.7% | 22.8% | 14.8% | 12.2% | 17.0% | *11.4% |
| Employment | ||||||
| Employment status24 | ||||||
| Employed: full-time | 42.0% | 34 .4% | 32 .3% | 17.1% | 16.6% | 17.2% |
| Employed: part-time | 13.4 | 17.4 | 13.2 | 8.4 | 9.2 | 8.3 |
| Unemployed | 3.4 | 19.7 | 11.2 | 8.5 | 16.4 | *7.3 |
| Not in labor force | 41.3 | 28.0 | 43.4 | 66.0 | 57.8 | *67.3 |
| Multiple jobs concurrently | 1.6% | 3.9% | 2.2% | 1.2% | 2.4% | *1.0% |
| Family Income and Program Participation | ||||||
| Total family annual income25 | ||||||
| Mean | $40,472 | $14,907 | $21,317 | $12,618 | $10,649 | *$12,928 |
| Median | $34,548 | $13,200 | $15,366 | $10,116 | $9,930 | $10,200 |
| Federal Poverty Level (FPL)26 | ||||||
| Income < 50% of FPL | 9.8% | 16.9% | 26.3% | 35.8% | 42.3% | *34.9% |
| Income btwn 50-99% of FPL | 13.5 | 36.9 | 25.0 | 49.6 | 40.3 | *51.0 |
| Income btwn 100- 199% of FPL | 22.5 | 36.2 | 29.6 | 11.0 | 15.0 | *10.4 |
| Income 200% or > FPL | 54.2 | 10.0 | 19.1 | 3.7 | 2.4 | 3.7 |
| Other sources of support | ||||||
| AFDC | 12.9% | 30.6% | 38.4% | 46.9% | 62.3% | *44.5% |
| Supplementary Security Income | 2.8 | 13.4 | 8.6 | 10.4 | 14.0 | *9.8 |
| WIC | 11.5 | 56.1 | 32.2 | 30.2 | 42.2 | *28.4 |
| Food Stamps | 18.0 | 50.2 | 49.3 | 59.0 | 70.0 | *57.2 |
| Housing | ||||||
| Public or subsidized housing | 8.9% | 22.1% | 25.4% | 26.4% | 41.1% | *23.8% |
| Number of moves in previous 12 months27 | ||||||
| None | ----- | 64.3% | ----- | ----- | ----- | ----- |
| One or more | 35.2 | |||||
| Health Status/Insurance Coverage | ||||||
| Primary caregiver health status | ||||||
| Excellent | 21.3% | |||||
| Very good | ----- | 28.4 | ----- | ----- | ----- | ----- |
| Good | 33.8 | |||||
| Fair | 14.4 | |||||
| Poor | 2.2 | |||||
| Child health status | ||||||
| Excellent | 49.0% | 44.8% | 28.4% | 37.0% | 32.1% | *37.7% |
| Very good | 23.2 | 30.5 | 21.3 | 24.5 | 19.5 | *25.3 |
| Good | 8.1 | 18.0 | 15.8 | 11.2 | 16.8 | *10.3 |
| Fair | 2.2 | 6.2 | 5.1 | 5.8 | 4.5 | 6.0 |
| Poor | 0.3 | 0.4 | 0.0 | 1.1 | 0.0 | *1.3 |
| Child disability | ----- | 17.6% | ----- | ----- | ----- | ----- |
| Child health insurance coverage28 | ||||||
| Private | 62.3% | 31.4 | 29.1% | 17.7% | 10.9% | *18.8% |
| Medicaid | 24.3 | 59.3% | 55.5 | 71.3 | 76.3 | *70.5 |
| Other | ----- | ----- | ----- | ----- | ----- | ----- |
| Child Care Arrangements | ||||||
| Arrangements29 | ||||||
| Family/friend | 40.5% | 19.2% | 46.1% | 42.6% | 44.6% | 42.3% |
| Family day care | 9.2 | 2.4 | 5.7 | 6.3 | 6.3 | 6.3 |
| Center-based day care | 27.9 | 6.1 | 24.4 | 24.6 | 22.6 | 25.0 |
| Hours/week in primary arrangement | Mean = 32.7 | Mean =19.2 | Mean = 28.8 | Mean = 29.4 | Mean =28.2 | Mean = |
PSID. Exhibit 2-13 contains findings from the PSID for families with children aged 3- to 5-years. As with the previous sections, the discussion will focus on the characteristics of eligible enrolled and non-enrolled families, with references to the Head Start FACES data. The respondents for the PSID were the heads of the households, who were typically adult males (79.4% of the household heads in this subsample were male).
Enrolled PSID families had an average of 4.6 family members and 3.0 children under 18 years, with the latter being significantly greater than non-enrolled families (2.6) and more than the Head Start FACES families (2.6). In the PSID, race and ethnicity data were only collected for head of household. While the non-enrolled group had a significantly higher proportion of White (non-Hispanic) respondents (60.1%) than the enrolled group (41.2%), the enrolled group was more evenly balanced between Black (non-Hispanic) (48.0%) and White (non-Hispanic) (41.2%) respondents, and contained a significantly greater proportion of Black (non-Hispanic) respondents than the non-enrolled group (24.8%). The racial and ethnic distribution of the respondents in both PSID groups was more weighted towards Whites and Blacks (non-Hispanic; 89.2% combined) than FACES, which had a larger proportion of non-White and non-Black children (27.2%).
The age distribution of the household heads was virtually identical for both the enrolled and non-enrolled respondents, with each group reporting that almost two fifths of the primary caregivers were in their twenties (38.9%, 37.8% respectively) or in their thirties (38.6% for both groups). Among the FACES primary caregivers, three fifths (59.3%) were in their twenties. Approximately one quarter (26.1%) of the PSID respondents from enrolled families reported being married (including ‘separated’ respondents), while a significantly larger proportion (37.9%) of their counterparts from non-enrolled families reported being married. The remaining respondents in both groups were evenly split between formerly married and never married. Over half (51.3%) of the FACES respondents reported that they were married.
While just under half of the PSID household heads from enrolled (49.0%) and non-enrolled (45.4%) families reported less than a high school diploma, one tenth (9.9%) from enrolled families reported some college, compared with one quarter (24.6%) from non-enrolled families reporting some college or higher. By contrast, among the primary caregivers responding in the FACES study, a lower proportion reported less than a high school diploma (27.6%), and a higher proportion reported some college or higher (36.3%).
Just over one half of household heads from enrolled and non-enrolled families reported full- or part-time employment, which, in turn, was similar to the percentage of FACES primary caregivers reporting full- or part-time employment (51.8%). While mean and median incomes appeared to be somewhat similar for household heads from both enrolled and non-enrolled families, a higher proportion of the respondents from enrolled families (89.8%) reported family incomes less than the FPL than did their counterparts from non-enrolled families (77.7%). A significantly greater proportion of the enrolled families than the non-enrolled families fell between 50% and 99% of the FPL. FACES respondents generally reported higher (household) incomes and lower poverty experience. Significantly higher proportions of PSID respondents from enrolled families reported participation in social support programs than did non-enrolled respondents, including the use of housing subsidies. With respect to Medicaid, the percentage of respondents from enrolled families (55.8%) reporting participation was still higher than the percentage reported by the respondents from non-enrolled families (51.4%), but this difference was not significant.
While there were no data collected for the PSID regarding child health status, a higher percentage of household heads from non-enrolled families (48.9%) reported that their own health status was either excellent or very good than did household heads from enrolled families (41.0%). Among FACES respondents, 49.7% reported being in excellent or very good health.
In summary, while the age distributions for the household heads from both enrolled and non-enrolled families and the employment experiences reported by both groups were quite similar, the respondents from non-enrolled families were more often married and were more likely to have reported some college education or higher. While both enrolled and non-enrolled respondents reported poverty-level incomes and use of social support program participation by more than one half of the families, in both cases, the proportions were consistently higher for families in the enrolled group.
| Dataset: PSID1993 Survey (Weighted Data) | ||||||
|---|---|---|---|---|---|---|
| All Children Age Age 3- to 5-years |
Head Start Eligible
Children Age 3- to 5-years |
|||||
| Characteristic | Total (n = 11,122 ) |
FACES Fall 1997 (n = 3,156) |
Enrolled
in Head Start (n = 2,051) |
Total (n = 2,329) |
Enrolled
in Head Start (n = 1,158 ) |
Not
enrolled in Head Start (n = 1,171) |
| Family Description | ||||||
| Number in household | ||||||
| Number in family | Mean = 4.4 | Mean = 4.5 | Mean = 4.6 | Mean = 4.4 | Mean = 4.6 | Mean = 4.3 |
| Number of children under 18 years old | Mean = 2.5 | Mean = 2.6 | Mean = 2.9 | Mean = 2.8 | Mean = 3.0 | *Mean = 2.6 |
| Race or ethnicity30 | ||||||
| White (non-Hispanic) | 76.7% | 27.7% | 41.7% | 51.4% | 41.2% | *60.1% |
| Black (non-Hispanic) | 17.2 | 36.9 | 49.5 | 35.5 | 48.0 | *24.8 |
| American Indian, Eskimo, Aleut | 0.2 | 1.9 | 0.5 | 0.2 | 0.4 | 0.0 |
| Asian or Pacific Islander | 0.0 | 1.1 | 0.0 | 0.0 | 0.0 | 0.0 |
| Hispanic | 2.9 | 24.2 | 5.2 | 6.2 | 7.3 | 5.2 |
| Age of primary caregiver31 | ||||||
| Less than 20 years | 0.0% | 2.4% | 0.0% | 0.0% | 0.0% | 0.0% |
| 20-29 years | 22.3 | 59.3 | 37.5 | 38.3 | 38.9 | 37.8 |
| 30-39 years | 57.3 | 29.0 | 40.4 | 38.6 | 38.6 | 38.6 |
| 40 years and older | 20.5 | 9.3 | 22.1 | 23.1 | 22.5 | 23.6 |
| Mean | 35.0 | 29.7 | 33.8 | 34.2 | 33.4 | 34.9 |
| Median | 33.0 | 28.0 | 31.0 | 32.0 | 30.0 | 33.0 |
| Marital status | ||||||
| Married | 74.5% | 51.3% | 46.6% | 32.5% | 26.1% | *37.9% |
| Formerly married | 13.6 | 12.6 | 25.4 | 33.6 | 36.0 | 31.5 |
| Never married | 12.0 | 36.0 | 28.1 | 34.0 | 37.9 | 30.7 |
| Education | ||||||
| Primary caregiver's education32 | ||||||
| Less than high school diploma | 20.0% | 27.6% | 36.7% | 47.0% | 49.0% | 45.4% |
| High school diploma | 33.6 | 36.5 | 42.5 | 28.5 | 33.5 | 24.3 |
| GED | 4.4 | ----- | 5.9 | 6.7 | 7.7 | 5.8 |
| Some college | 18.9 | 33.5 | 11.2 | 13.7 | 9.9 | 17.1 |
| Bachelor's degree or higher | 23.1 | 2.8 | 3.8 | 4.0 | 0.0 | * 7.5 |
| Currently in school or training | ----- | ----- | ----- | ----- | ----- | ----- |
| Employment | ||||||
| Employment status33 | ||||||
| Employed: full-time | 68.2% | 34 .4% | 39.2% | 24.5% | 21.8% | 26.9% |
| Employed: part-time | 13.4 | 17.4 | 23.5 | 23.2 | 25.1 | 21.6 |
| Unemployed | 6.2 | 19.7 | 9.6 | 11.9 | 11.8 | 11.9 |
| Not in labor force | 11.1 | 28.0 | 27.4 | 36.3 | 40.6 | 32.7 |
| Multiple jobs concurrently34 | 15.7% | 3.9% | 4.9% | 6.4% | 1.8% | *10.4% |
| Family Income and Program Participation | ||||||
| Total family annual income | ||||||
| Mean | $46,106 | $14,907 | $20,273 | $11,738 | $11,276 | $12,133 |
| Median | $32,045 | $13,200 | $15,000 | $8,640 | $8,904 | $8,568 |
| Federal Poverty Level (FPL) | ||||||
| Income < 50% of FPL | 9.9% | 16.9% | 22.6% | 39.5% | 35.9% | 42.6% |
| Income btwn 50-99% of FPL | 11.0 | 36.9 | 33.9 | 43.8 | 53.9 | *35.1 |
| Income btwn 100-199% of FPL | 17.1 | 36.2 | 21.0 | 9.4 | 4.1 | *13.9 |
| Income 200% > FPL | 61.9 | 10.0 | 22.6 | 7.3 | 6.1 | 8.4 |
| Other sources of support | ||||||
| AFDC | 15.8% | 30.6% | 44.5% | 63.0% | 70.8% | *56.2% |
| Supp. Security Income (SSI) | 2.9 | 13.4 | 12.6 | 11.7 | 20.0 | *4.6 |
| WIC | ----- | 56.1 | ----- | ----- | ----- | ----- |
| Food Stamps | 21.1 | 50.2 | 60.6 | 68.1 | 81.8 | *56.4 |
| Housing | ||||||
| Public or Subsidized Housing | 11.1% | 22.1% | 31.5% | 28.6% | 37.4% | *21.1% |
| Number of Moves in Previous 12 Months35 | ||||||
| None | 81.2% | 64.3% | 68.8% | 70.9% | 71.5% | 70.3% |
| One | 18.8 | 35.2 | 31.2 | 29.1 | 28.5 | 29.7 |
| Health Status/Insurance Coverage | ||||||
| Primary caregiver health status36 | ||||||
| Excellent | 24.5% | 21.3% | 11.5% | 14.6% | 7.8% | * 20.4% |
| Very good | 35.7 | 28.4 | 33.8 | 30.7 | 33.2 | 28.5 |
| Good | 29.6 | 33.8 | 38.2 | 32.0 | 40.7 | *24.5 |
| Fair | 7.4 | 14.4 | 9.7 | 17.9 | 11.5 | *23.4 |
| Poor | 2.6 | 2.2 | 5.9 | 4.5 | 6.8 | 2.6 |
| Child health status | ||||||
| Excellent | 44.8% | |||||
| Very good | ----- | 30.5 | ----- | ----- | ----- | ----- |
| Good | 18.0 | |||||
| Fair | 6.2 | |||||
| Poor | 0.4 | |||||
| Child disability | ----- | 17.6% | ----- | ----- | ----- | ----- |
| Child health insurance coverage37 | ||||||
| Private | ----- | 31.4% | ----- | ----- | ----- | ----- |
| Medicaid | 15.9% | 59.3 | 41.7% | 53.4% | 55.8% | 51.4% |
| Other | ----- | ----- | ----- | ----- | ----- | ----- |
| Child Care Arrangements | ||||||
| Arrangements | ||||||
| Family/friend | 19.2% | |||||
| Family day care | ------- | 2.4 | ------- | ------- | ------- | ------- |
| Center-based day care | 6.1 | |||||
| Hours per week in primary arrangement | ------- | Mean=19.2 | ------- | ------- | ------- | ------- |
2.5 Summary of Family/Household Databases
Appendix C is a summary exhibit, containing some of the findings presented previously for each of the specific datasets. For each national dataset, this exhibit presents two columns of findings: one for eligible, enrolled families and one for eligible, non-enrolled families. This represents a different form of presentation for the same findings that were discussed in Chapter 2.
This exhibit allows the reader to review the findings of the various datasets side by side. However, as noted earlier, while some variables are comparable across datasets, others do not readily lend themselves to direct comparison, and to attempt to do so could lead to erroneous conclusions.
2.6 Conclusions Regarding Findings from the Secondary Data Analyses
The following is a summary of the review of existing data described in this chapter:
-
The Head Start PIR indicated that about 4% of recruited children left before actual initiation of services and an additional 15% left after services were initiated. However, over 85% of these children were replaced within 30 days, suggesting that a pool of eligible children was available to most Head Start programs for the replacement of children who dropped out of the program.
-
Head Start FACES interviews with Social Service Coordinators and Center Directors indicated that, at the beginning of the program year, over 80% of programs and centers had a list of children available for replacement of children that left the program. Thus, many Head Start programs appeared to have access to children over and above those currently being served.
-
The FACES Program Drop Reports showed that family moves were the most frequent reason for withdrawal from Head Start (42%). Other common reasons for withdrawal were enrollment in another program (23%), poor attendance (7%), family or personal problems (6%), and transportation difficulties (6%).
-
Three national datasets (The NLSY79, the SIPP, and the PSID) contained extensive information about family characteristics. Ultimately, comparisons across data sets were difficult because information was obtained at different time points using different respondents, different questions, and different methods for ascertaining family income. Nevertheless, each of the data sets indicated that a substantial number of families with incomes below the Head Start income-eligibility criteria and with children eligible for Head Start services were present across the nation. Families with children enrolled in Head Start generally received higher levels of supplemental services, including subsidized housing and Medicaid, but it was not clear if this difference reflected a consequence of Head Start enrollment.
The intent of the review of existing data sources was to investigate what can be learned about the characteristics of Head Start-eligible families, whether enrolled or not enrolled in the program. While the three national data sets provided valuable information regarding the Head Start-eligible population, they generally did not provide consistent, comparable information that might allow building valid conclusions regarding differences and similarities between the Head Start eligible and non-eligible populations.
Each of the national data sets provided a snapshot of particular characteristics associated with the target group of families (i.e., the “Missing FACES”). However, collectively they offered a limited perspective of who Head Start-eligible families really were. Due to their original focus, these national data sets did not provide information about what eligible families knew about Head Start, how families obtained this information, and why families were or were not participating in the program. Although reasonable to question the accuracy of the parents’ ability to correctly report on whether or not their children attended Head Start, the validity of these data are strengthened by the ongoing work of the Early Childhood Longitudinal Study (ECLS-Kindergarten cohort), which supports the notion that parents do respond accurately to this question. In the ECLS-K, parents of kindergarten children were asked whether their children attended Head Start during the previous year, and project staff verified the responses for children reported to be in Head Start. As noted, recall periods do vary for the national datasets, and longer periods, such as those that may have been used for the PSID, could restrict the level of accuracy noted for the ECLS-K.
Similarly, the Head Start PIR data and the Head Start staff interviews provided only limited information because they do not provide background information to explain families’ choices for their children. A strong need for valid and reliable information on eligible but non-enrolled children and their families continues to exist.
11An asterisk is used to represent a significant difference between eligible enrolled versus eligible non-enrolled children for each characteristic analyzed. Statistical significance was tested at the .05 level.(back)
12Household refers to all individuals sharing the respondent’s primary residence at the time of the interview. A family includes all those in the household related by blood, marriage or adoption.(back)
13Primary Caregiver refers to the respondent. The NLYS79 CHILDREN sample includes children born to female NLSY79 respondents. These women were between 14 and 21 years of age on January 1, 1979, and, consequently, were between the ages of 31 and 38 during the 1996 survey.(back)
14Education was defined as the highest grade completed. Thus, if a person completed 12(th) grade, it was assumed that the person graduated high school. Anyone reporting more than 12 years of schooling was placed into the ‘Some College’ category(back)
15Employment Status: ‘Employed’ indicates those who were employed or self-employed during the previous week. ‘Part-time’ employment is reported for those who usually work less than 30 hours per week. ‘Unemployed’ refers to those who, during the previous week, were either laid off, or were available and looking for work. ‘Not in Labor Force’ includes retired, disabled, and those not in the labor force for some other reason.(back)
16Total Family Annual Income includes income of related family members (see Footnote 1) reported for the 1995 calendar year. NLSY income includes amounts received for Food Stamps.(back)
17 Frequencies for poverty level were calculated only using families with data on both family size and family income.(back)
18 FACES-Based on number of moves in past 12 months(back)
19 Insurance Coverage is defined as having health insurance coverage at the time of the interview. ‘Private’ includes “health insurance provided either by an employer or by an individual plan.” ‘Medicaid’ indicates that the child was covered by Medicaid at the time of the interview.(back)
20An asterisk represents a significant difference between eligible enrolled vs. eligible not enrolled children in each dataset per characteristic analyzed. Statistical significance was tested at the .05 level.(back)
21Household is defined as the household of the ‘household reference person’ (respondent) and may include related and non-related persons. ‘Family’ refers to a group of two or more people related by birth, marriage, or adoption who reside together.(back)
22The Primary Caregiver is defined as the designated parent or guardian of the child (as reported by the SIPP respondent).(back)
23Education was defined as the highest grade completed. Thus, if a person completed 12(th) grade, it was assumed that the person graduated high school. Anyone reporting more than 12 years was placed into ‘some college.’(back)
24Employment Status: ‘Employed’ indicates those who were employed or self-employed during the previous month. ‘Part-time’ employment is working less than 30 hours per week. ‘Unemployed’ refers to those who were not employed during the previous month but were available for and looking for work. ‘Not in Labor Force’ refers to those who were neither employed nor unemployed.(back)
25Total Family Annual Income includes income of related family members (see Footnote 1) reported for the 1995 calendar year.(back)
26Frequencies for poverty level were calculated only using families with data on both family size and family income.(back)
27FACES-Based on number of moves in the past 12 months.(back)
28Insurance Coverage is defined as having health insurance coverage the previous month. ‘Private’ includes private insurance, CHAMPUS, CHAMPVA, and military coverage. ‘Medicaid’ indicates that the child was covered by Medicaid.(back)
29Child care arrangements include data on the primary child care arrangement (if there were multiple arrangements) used for 10 hours or more per week.(back)
30Refers to Head of Household.(back)
31 Refers to Head of Household.(back)
32 Refers to Head of Household.(back)
33 Refers to Head of Household.(back)
34Based on whether ever worked more than two jobs concurrently in past year.(back)
35 One move means ‘one or more moves’ in the past year.(back)
36 Refers to Head of Household.(back)
37PSID only reports on Medicaid. (back)
| Table of Contents | Previous | Next |