ACF OPRE: Early Head Start Research and Evaluation Project (EHSRE)

III. EARLY HEAD START IMPACTS ON SERVICE RECEIPT

Early Head Start is a complex intervention that is challenging to implement. As a first step toward understanding impacts on children and families, we need to document program accomplishments and the services families received. Did the 17 research programs provide a fair test of the Early Head Start concept, or did the programs fail to deliver the key services to many families?

Evidence from the implementation study shows that, overall, the research programs succeeded in implementing Early Head Start services and delivered key services to most families who enrolled (Administration on Children, Youth and Families 2001b). As summarized in Chapter I, patterns of implementation varied among the 17 research programs. One-third did not reach full implementation by fall 1999, but most had fully implemented the program by fall 1999. The implementation study also showed that in the initial period following program enrollment, the research programs provided some child development/parenting services to nearly all families who enrolled, and provided intensive services to the majority. Section B below summarizes the nature and intensity of these services.

Control group families were prohibited from receiving Early Head Start services but were free to seek other similar services in their communities. Therefore, it is critical that we document the services they received and assess whether the Early Head Start programs increased the types, amount, or intensity of services that families received. If most control group families received similar services, and if these services were as intensive as the services that Early Head Start families received, we might find few significant impacts on child and family outcomes, even if the Early Head Start research programs were highly successful in achieving their desired outcomes.

Our analysis of program impacts on service receipt shows that even though many control group families received some similar services from other community sources during the initial period after random assignment, program families were much more likely to receive key services:

Even though three-quarters of control families received some key services (home visits, center-based child care, case management, and/or group parenting activities) from other community sources during the first 16 months after enrollment, program families were significantly more likely to receive at least one of these key services (nearly all of them did).
Early Head Start families were much more likely than control families to receive the core child development or parenting-focused services (home visits and/or center-based child care and child development services). In addition, Early Head Start families were more than twice as likely as control families to participate in parent education, parent-child, or parent support group activities.
The Early Head Start programs increased the receipt of intensive child development/parenting services even more dramatically. During the first seven months after enrollment, for example, most Early Head Start families received home visits at least monthly, and nearly half received them at least weekly (an intensity of child development services generally regarded as necessary to produce child or parenting effects). In contrast, very few control families received monthly or weekly home visits. In home-based programs, the majority of Early Head Start families, but very few control families, received weekly home visits. During the first 16 months after enrollment, children who enrolled in center-based Early Head Start programs were in center-based care for almost twice as many hours as control children.
The Early Head Start programs also increased families' receipt of case management and their use of services in the community, such as education and employment-related services and transportation assistance.
Medicaid and State Children's Health Insurance Programs have made health care services widely accessible to low-income families, and nearly all program and control group families reported receiving basic health services.

This chapter presents our analyses of program impacts on families' service receipt. The first section briefly summarizes service receipt by program families, and the second section assesses Early Head Start impacts on service receipt during the first two follow-up periods (16 months, on average, after enrollment). The final section discusses the implications of these analyses for the analyses of impacts on children and families. The data sources for the analyses in this chapter include the parent services follow-up interviews completed an average of 7 and 16 months after random assignment and the Head Start Family Information System application and enrollment data which were described in Chapter II.

A. OVERVIEW OF THE LEVELS AND INTENSITY OF EARLY HEAD START SERVICES DURING THE FIRST 16 MONTHS

This chapter focuses on differences in service receipt by program and control group members. To set the context for examining these differences, this section briefly summarizes the levels and intensity of services received by Early Head Start program families during the first 16 months, on average, after enrollment. A detailed discussion of program participation and service use can be found in the Pathways to Quality implementation study report (Administration on Children, Youth and Families 2001b). At the local level, research-program partners have looked into the home visiting services in more detail. As seen in Box III.1, observational data provide an in-depth picture of home visiting that supplements the cross-site findings listed here.

BOX III.1

AN INSIDE LOOK AT HOME VISITING

Carla A. Peterson, Susan L. McBride, Gayle J. Luze, and Marcia Macedo
Iowa State University

Recent efficacy studies of home-visiting programs have produced mixed and modest results, and home visiting is being questioned as an effective mechanism for service delivery. However, the home is only a location for intervention services. Many recent evaluations of home-visiting programs have employed rigorous experimental designs but have failed to document the actual nature and content of home visits, the diversity of programs and populations being served, or a theory of how and why a program might work.

Iowa State University researchers have collaborated with Mid-Iowa Community Action, Inc. (MICA) to document the process and content of interventions delivered to 77 families through home visits. Here, to illustrate the notion that home visiting as a service delivery model is complex and not homogenous across families even within a single program, we profile two families receiving Early Head Start services.

Observational data describing the process and content of home visits were collected by research staff who accompanied interventionists to families' homes. These data were summarized to present the percentages of overall time spent on content areas (for example, child development topics, family topics) and in specific intervention arrangements (for example, facilitating parent-child interaction, providing information). These data were combined with program documentation of hours of home visiting received to calculate total numbers of hours, or dosage, of specific intervention strategies implemented with individual families.

Rita and Kandy (not their real names) are two young mothers who received home-visiting services from MICA's Early Head Start program; they are very much like many participants in the program. Rita and Kandy were (1) each parenting one child (both children were born in summer 1996); (2) had a high school diploma; and (3) lived in a small, rural community. Both women were single; however, Kandy lived with her son's father during part of this time. Despite these similarities, MICA found it necessary to provide very different Early Head Start services to these two families.

Both families received home visits from a child development specialist (CDS) and a family development specialist (FDS) overall, both families received similar numbers of home visits. However, Rita's family received far more child development services than did Kandy's. Rita's family received 113 home visits (160 hours); 65 of these (99 hours) were CDS visits. Rita's CDS visits focused on child-related content 51 percent of the time, translating into 51 total intervention hours. Rita's FDS visits focused on child-related content 23 percent of the time, accounting for an additional 14 hours of child-related intervention. The CDS spent 19 hours and the FDS spent 5 hours engaging Rita's son and supporting his interactions directly by teaching the child themselves, modeling interactions for Rita, or coaching Rita's interactions with her son. In contrast, Kandy's 109 visits were split almost evenly between CDS visits (55 visits and 68 hours) and FDS visits (54 visits and 61 hours). Kandy received 51 hours of child-related intervention, and interventionists spent 18 hours engaging Kandy's son directly.

Seemingly, greater emphasis on a specific content area or strategy should translate into more powerful intervention outcomes in the targeted area(s). However, systematic study of links between intervention activities, outcomes, and contexts is necessary to refine intervention services effectively and to guide policy recommendations adequately.

1. Early Head Start Participation

Nearly all families in the program group received some Early Head Start services. Ninety-one percent received more than minimal services (received more than one home visit, met with a case manager more than once, received at least two weeks of center-based child development/child care services, and/or participated in group parenting activities).1 Most families (86 percent) received child development or parenting services during home visits or in program centers.

2. Home Visits

Most families in the program group received at least one Early Head Start home visit. Across all research programs, most families (85 percent) reported receiving at least one Early Head Start home visit by the time of the second follow-up interview, and 75 percent reported receiving more than one. Receipt of home visits was highest among home-based programs (92 percent of families reported receiving at least one Early Head Start home visit, and 89 percent reported receiving more than one) and lowest among center-based programs (64 and 34 percent, respectively).
Slightly more than half of families in home-based programs received weekly home visits. Among the home-based research programs, 57 percent of families, on average, reported receiving Early Head Start home visits at least weekly during the first follow-up period (seven months, on average), and 52 percent reported Early Head Start home visits at least weekly during the second follow-up period (nine months on average). These levels of receipt of weekly home visits are generally consistent with the experiences of other home visiting programs, which have found that, on average, they are able to complete about half the intended number of home visits, regardless of the planned frequency of home visits (Gomby 1999).
Home visits almost always included child development activities. Nearly all families who reported receiving Early Head Start home visits reported receiving child development services during those visits.
Home visits typically lasted at least an hour. Most parents who received Early Head Start home visits reported that the typical visit lasted between one and two hours.
Receipt of Early Head Start home visits remained high throughout the first two follow-up periods but declined modestly in the second period as some families left the program.2 On average, 70 percent of families reported receiving more than one Early Head Start home visit by the time of the first follow-up interview. The level of reported home visit receipt declined between the first and second follow-up interviews (to an average of 58 percent of families) as some families left the program.

3. Case Management

Home visits and case management services overlapped substantially. The receipt and frequency of case management mirrored the receipt and frequency of home visits. Most program families reported receiving both home visits and case management (71 percent in the first follow-up period and 56 percent in the second follow-up period). More than 90 percent of these families reported that the person they met with for case management was the same person who visited them at home.

4. Group Parenting Activities

Participation in group parenting activities was lower than participation in other key services. Overall, slightly more than half of families reported that they had attended an Early Head Start group parenting activity by the time of the second followup.

5. Child Care and Center-Based Child Development Services

Levels of child care use by program families were high across all three program types. Two-thirds of children had received some child care services by the time of the first followup. Nearly 80 percent had by the time of the second.
Program families relied on a wide range of providers for their primary child care arrangement (the arrangement used for the most hours during the follow-up period). Twenty percent of all program families relied primarily on an Early Head Start center, and 14 percent relied on other child care centers. Another one-third of families relied on a relative as their primary child care provider, usually grandparents or great-grandparents. Twelve percent of families used a nonrelative caregiver as their primary arrangement.
The proportion of program families using center-based child care increased over time as children got older. One-third of all program children received care in child care centers during the first follow-up period. By the time of the second followup, the percentage of children who had ever been enrolled in center-based care increased to 43 percent.
Approximately one-fourth of program families received center-based Early Head Start care. On average, 22 percent of program children received Early Head Start center-based child development services during the first follow-up period. By the time of the second followup, 25 percent had received Early Head Start center-based child development services.
The use of multiple child care arrangements was common. On average, children received child care in two arrangements during the first 15 months after enrollment. One-third of program children received care in multiple arrangements concurrently.
Many program children received intensive child care services. One-third of program children were in child care for an average of 20 hours per week or more during the first 15 months after enrollment.3 Not all children who received child care were in care during the entire follow-up period, but more than half of program children received child care for at least 60 percent of the 15-month follow-up period.

6. Services for Children with Disabilities

By the time of the second followup, five percent of program families reported that their child had an identified disability.4 The proportion of children whose parents reported that a disability had been identified ranged from 0 to 13 percent across programs.
The percentage of families who reported that their child had received early intervention services was slightly lower. On average, three percent of families reported that they had received early intervention services. Across programs, the receipt of early intervention services ranged from zero to eight percent.

7. Health Services

All children had received some health services by the second followup. Nearly all children had received some immunizations by the time of the second followup (97 percent of all program children). More than 90 percent had visited a doctor.
Nearly all families (96 percent) had received some health services (besides those they obtained for the focus child) by the time of the second followup. The proportion of families who received any health services ranged from 85 to 100 percent across programs.
By the time of the second followup, when children were 20 months old on average, few children (11 percent) had visited a dentist.
At least one family member in nearly two-thirds of program families had visited an emergency room by the second followup. Many program children (42 percent) had visited an emergency room by the time of the second followup.
By the time of the second followup, 17 percent of program families had received some mental health services, including 16 percent who had received treatment for an emotional or mental health problem, and 3 percent who had received drug or alcohol treatment.

8. Other Family Development Services

Most primary caregivers (83 percent) reported having received education-related services by the time of the second followup. Two-thirds of primary caregivers reported having talked to a case manager about education services, and slightly more than half reported having attended school or a job training program.
Two-thirds of program families reported having received some employment-related services by the time of the second followup. Twenty-two percent of families reported having received job search assistance by the second followup, and 61 percent reported having talked to a case manager about finding a job or job training.
Many families received other important support services. Half of program families had received housing assistance (public housing, rent subsidy, help finding housing, and/or energy assistance) by the time of the second followup. Nearly 30 percent of program families had received transportation assistance.

9. Families' Engagement in Early Head Start Services

On average, program staff judged that slightly more than one-third of the research families became highly involved in program services. The extent to which program staff rated families as highly involved, however, varied substantially across the 16 sites that provided ratings, ranging from 20 to 74 percent. The staffs of three programs reported that at least half the research families enrolled in their program were highly involved.

10. Fathers' Receipt of Early Head Start Services

Although the vast majority of respondents to the parent interviews were mothers, the Early Head Start research also collected information from fathers about their receipt of Early Head Start services when their children were approximately 24 months old. Box III.2 on the following page presents a picture of the range of program activities and services that the fathers of Early Head Start children participated in. Future reports will provide even greater details on Early Head Start fathers and their program participation.

BOX III.2: FATHER INVOLVEMENT IN EARLY HEAD START PROGRAM ACTIVITIES

More than half of the Early Head Start fathers/father figures interviewed participated in at least one of the seven Early Head Start program activities we asked them about.a Fourteen percent of the fathers reported participating in three or more types of activities, and 40 percent participated in one or two types of activities. As expected, resident biological fathers were more likely than nonresident biological fathers and resident father figures to participate frequently in Early Head Start program activities (see figure below).

Twenty percent of the fathers interviewed reported that they participated in home visits more than once a month. As expected, resident biological fathers participated in home visits more often than nonresident biological fathers (26 percent versus 8 percent). Nine percent of the resident father figures participated in home visits, similar to the participation rates of the nonresident biological fathers.

Almost one-third of fathers dropped off or picked up their child at an Early Head Start center one or more times. Twenty-one percent of fathers reported that they dropped off or picked up their child from an Early Head Start child development center three times or more in the past month. Sixteen percent reported that they dropped off or picked up their child 10 times or more.b

Rates of father participation in the other program activities were lower, ranging from two percent to eight percent of fathers who participated three or more times in the past six months.

Father Involvement in Early Head Start Program Activities

Source: Father interviews conducted when the children were approximately 24 months old.

aThe seven types of activities we asked about include: Early Head Start home visits, dropping off/picking up child at an Early Head Start center, attending Early Head Start parenting classes or events, attending Early Head Start parent-child activities, attending Early Head Start meetings or events just for fathers, attending an Early Head Start Policy Council or governing board meetings, and volunteering to help at the Early Head Start program. The 12 father interview study program sites included all four center-based programs, five of the home-based programs, and three mixed approach programs. None of the fathers in the home-based programs had the opportunity to pick up or drop off their children from an Early Head Start center. Fathers in the center-based and mixed programs had fewer opportunities to participate in home visits than fathers in the home-based programs.(back)

bTen times or more is equivalent to 25 percent of the time if the child attended 20 days per month and needed to be dropped off and picked up 40 times.(back)

B. EARLY HEAD START IMPACTS ON SERVICE RECEIPT

Although control group families could not receive Early Head Start services, they were free to seek similar services in their community. Thus, for understanding program impacts on child and family outcomes, it is important to examine the differences in service receipt5 between program and control families (in other words, the program impacts on service receipt). The following subsections describe the global impacts of the Early Head Start programs on service receipt and then summarize the key variations in these impacts among key subgroups of programs. To illustrate the complexity of the task programs have in meeting the often diverse needs of their families, the local research report in Box III.3 shows the importance of flexible programming.

1. Global Differences in Receipt of Services

Our analyses show that many control families received services similar to those provided by Early Head Start. Nevertheless, program families were much more likely than control families to receive these services. The Early Head Start program impacts on receipt of services persisted through the first two follow-up periods.

a. Impacts on Overall Service Receipt

Early Head Start program families were significantly more likely than control families to receive any key services (home visits, case management, center-based child care/child development services, and group parenting activities) by the time of the second followup (95 compared with 75 percent) (Figure III.1). The impact on receipt of key services was largest during the first follow-up period (31 percentage points, not shown) and then declined to 20 percentage points as more control families began receiving some services during the second follow-up period.

The Early Head Start research programs' impact on receipt of core child development/parenting services was much larger. Nearly all program families (92 percent) had received core child development/parenting services-home visits and/or center-based child care-while only 51 percent of control families had done so by the time of the second follow-up interview.

BOX III.3

DIVERSITY OF EARLY HEAD START FAMILIES AND PROGRAM SERVICES

Michaela Farber, Shavaun Wall, and Harriet Liebow
The Catholic University of America

To understand how United Cerebral Palsy Early Head Start promotes child development and self-sufficiency in families struggling with poverty in Northern Virginia, The Catholic University of America research partners profiled the needs and program services of diverse families. To meet the unique needs of the 75 families served, Early Head Start tailored its array of program services to their demographic profile, birth (immigrant or U.S.-born), and occupational status (military or civilian). Early Head Start served 45 percent immigrant and 55 percent U.S.-born families. The U.S.-born comprised 35 percent military and 20 percent civilian families. To meet the needs of children in these families, Early Head Start developed a flexible mixture of child-focused services, including center-based and family child care and home visiting.

The immigrant families were in their late 20s, married, and more concerned with obtaining basic resources than civilian or military families. To meet their basic needs, Early Head Start sought to mobilize resources in public, faith-based, and voluntary sectors of the community. In addition to poverty, immigrants faced three barriers to economic self-sufficiency: (1) having inadequate English-speaking skills, (2) not completing a high school education, and (3) living in the United States less than five years. To counter these barriers, Early Head Start facilitated referrals to community education.

Three-quarters of the immigrant families were of Hispanic origin, most from Central America, but some from South America and Mexico. The rest were from West Africa, the Caribbean, Pakistan, the Philippines, Vietnam, and Bosnia. To directly serve them, Early Head Start hired bilingual staff. To identify and help remedy the linguistic gaps in the community, Early Head Start staff also participated in community forums. As a result of EHS investment in community collaboration, many immigrants were able to enroll in English classes by showing proof of their participation in Early Head Start.

U.S.-born military and civilian families were younger than immigrants. Military families were more likely to be married, with some college education. Civilian families were the youngest, least likely to be married, and most likely to have a high school education. Although U.S.-born families had more resources than immigrants, they struggled with lack of economic self-sufficiency, family problems, and health care. Civilian families faced the pressing needs of very young families with inadequate health care. Military families faced stresses such as deployment or family separations. To address the needs of young families, Early Head Start integrated Fairfax County's new Nurturing Program into its parent education program. Through community collaboration, Early Head Start staff facilitated a countywide shift in health care for low-income families from a lottery system to universal availability. Early Head Start also established child care in a child development center at a nearby army post and participated in the Special Needs Review Team at the center. To facilitate access to needed mental health and family services, Early Head Start staff collaborated with the military's Family Advocacy, Exceptional Family Member, and New Parent Support Group programs and helped families directly access community services. In addition, Early Head Start staff worked closely with the county's early intervention services to promptly identify and provide support to families of infants and toddlers with special needs.

In conclusion, knowledge of family birth status, occupational status, and demographic needs proved useful in designing and implementing Early Head Start individualized, comprehensive, and culturally sensitive services.

FIGURE III.1

IMPACTS ON RECEIPT OF ANY KEY SERVICES AND CORE CHILD
DEVELOPMENT SERVICES BY THE SECOND FOLLOWUP

Figure 3.1: Impacts on Receipt of Any Key Services and Core Child Development by the Second Followup [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after random assignment.

Notes: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant.

Key services include at least one home visit, center-based child care, at least one case management meeting, and/or participation in a group parenting activity. Core services include at least one home visit and/or center-based child care.

* Program impact is significantly different from zero at the .10 level, two-tailed test.
** Program impact is significantly different from zero at the .05 level, two-tailed test.
***Program impact is significantly different from zero at the .01 level, two-tailed test.

b. Impacts on Receipt of Home Visits

All Early Head Start programs are expected to visit families at home on a regular basis. Home-based programs are expected to visit families weekly, and center-based programs must visit families at home at least twice a year (though many do so more often). The Early Head Start research programs had their largest impacts on receipt of home visits. By the time of the second followup, substantially more program than control families had received at least one home visit (87 compared with 33 percent, on average) (Figure III.2). Not only were program families much more likely to have received any home visits by the time of the second followup, they were also much more likely to have received weekly or monthly home visits. Very few control families (four and two percent in the first and second follow-up periods, respectively) received home visits weekly, while more than one-third of program families received them weekly (Figure III.2). A few more control families received home visits at least monthly (11 and 6 percent, respectively), while nearly two-thirds of program families received home visits at least monthly.

Nearly all the families in both groups who received home visits reported that they received child development services during the visits. Thus, the Early Head Start impacts on receipt of home visits largely reflect impacts on receipt of child development services during home visits.

FIGURE III.2

IMPACTS ON HOME VISIT RECEIPT BY THE SECOND FOLLOWUP

Figure 3.2: Impacts on Home Visit Receipt by the Second Followup [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after enrollment.

c. Impacts on Receipt of Case Management

Among the key services provided by the Early Head Start programs, case management was the one that control families were most likely to receive from other sources in their communities. Nevertheless, program families were significantly more likely than control families to have received case management by the time of the second follow-up interview (85 percent met with a case manager at least once, compared with 50 percent of control families) (Figure III.3).

Program impacts on the receipt of frequent case management were large and similar to the impacts on receipt of frequent home visits, which reflects the substantial overlap between home visits and case management services (Administration on Children, Youth and Families 2001b). As was the case for home visits, only a small proportion of control families met with a case manager at least weekly, while more than one-third of program families did so (Figure III.3).

The program impacts on receipt of case management at least monthly were even more dramatic. During the first follow-up period, for example, 68 percent of program families met with a case manager at least monthly, compared with 18 percent of control families (Figure III.3).

FIGURE III.3

IMPACTS ON CASE MANAGEMENT RECEIPT BY THE SECOND FOLLOWUP

Figure 3.3: Impacts on Case Management Receipt by the Second Followup [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after enrollment.

d. Impacts on Receipt of Parenting Information Services and Group Parenting Activities

The Early Head Start programs substantially increased the likelihood that families received parenting information during home visits or group parenting activities. Nearly all program families (93 percent), compared with 56 percent of control families, reported receiving any parenting information services by the time of the second followup (Figure III.4).

Although the Early Head Start programs found it very challenging to achieve high participation levels in group parenting activities (parenting classes, parent-child socialization activities, or parent support groups), they substantially increased program families' participation relative to control families' participation in similar activities in the community. Two-thirds of program families, compared with just under one-third of control families, had participated in a group parenting activity by the time of the second followup (Figure III.4).

Among the group activities we examined, the Early Head Start programs increased participation in parenting classes the most (Figure III.4). By the time of the second followup, approximately half the program families, compared to only one-fourth of control families, reported having participated in parenting classes.

The impact of the programs on participation in parent-child group socialization activities was also substantial. One-third of program families had participated in group activities for parents and children by the time of the second followup, compared with only 10 percent of control families (Figure III.4).

FIGURE III.4

IMPACTS ON RECEIPT OF PARENTING SERVICES BY THE SECOND FOLLOWUP

Figure 3.4: Impacts on Receipt of Parenting Services by the Second Followup [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after enrollment.

Group parenting activities include parenting classes or events, group parent-child socialization activities, and parent support group meetings.

e. Impacts on Child Care and Center-Based Child Development Services

The Early Head Start programs significantly increased families' use of child care. Most families in both the program and control groups had used some child care for the focus children by the time of the second followup, but program children were significantly more likely than control children to have received some child care-80 compared with 74 percent (Figure III.5).

The programs increased families' use of center-based child care/child development services more dramatically. By the time of the second followup, 43 percent of program families, compared with 27 percent of control families had used some center-based child care for their focus child (Figure III.5).

Not only did the Early Head Start programs increase the percentage of families using any child care, they also increased the amount of child care that children received. During the 15 months after random assignment, program children received significantly more hours per week of child care than control children (16.3 compared with 12.9 hours per week, on average) (Figure III.6). Similarly, the programs almost doubled the average hours per week of center-based care that children received (from 3.6 to 7.1 hours per week).

Program families were significantly more likely than control families to use concurrent child care arrangements (more than one child care arrangement at the same time) (Figure III.5). Program families may have had a greater need for multiple arrangements to cover all the hours during which they needed child care, because they used significantly more center-based care than control families, and centers are less likely than some other providers, such as relatives or friends, to provide care during evenings or weekends.

Program families paid significantly less money out of pocket for child care, on average, than control families ($3.34 less per week through the 15th month after random assignment, almost a 40 percent reduction in average out-of-pocket costs) (not shown). Some of the Early Head Start programs provided child care to some or all families free of charge. Other Early Head Start programs did not provide child care directly but helped some families arrange care with other providers and obtain child care subsidies to pay some or all of the costs of those arrangements. The experience of Project EAGLE in Kansas City illustrates program-community partnership strategies for helping families access child care (Box III.4).

FIGURE III.5

IMPACTS ON USE OF CHILD CARE SERVICES FOR FOCUS CHILD BY THE SECOND FOLLOWUP

Figure 3.5: Impacts on Use of Child Care Services for Focus by the Second Followup [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after enrollment.

FIGURE III.6

IMPACTS ON HOURS OF CHILD CARE USE FOR FOCUS CHILD BY THE SECOND FOLLOWUP

Figure 3.6: Impacts on Hours of Child Care Use for Focus by the Second Followup [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after enrollment.

Notes: All means are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant.

BOX III.4

CHILD CARE PARTNERSHIPS

Martha D. Staker
Project EAGLE Early Head Start

Early Head Start programs are charged with developing comprehensive initiatives to support infants, toddlers, pregnant women, and their families. How they do this is up to them as long as the program meets or exceeds the Head Start Program Performance Standards. This flexibility allows Early Head Start sites to select program options and design services that respond to community and individual family needs.

However, most Early Head Start programs can enroll only a fraction of those eligible for the program. If Early Head Start programs partnered with existing agencies to support families, they could share resources and strengthen systems. Early Head Start could affect more children and families by purchasing services from other agencies that serve children, pregnant women, and families and by anchoring the partnership through training and joint case management efforts.

Project EAGLE Early Head Start of the University of Kansas Medical Center does this. Project EAGLE decided to invest in the community by purchasing developmentally appropriate child care from 25 existing centers and family child care homes. Over three years, Project EAGLE paid the tuition and fees for 55 child care providers from these sites to attend the local community college and work on their Child Development Associate (CDA) credential. Project EAGLE purchased textbooks, gave stipends to the child care providers who needed child care for their own children while they were in class, and awarded bonuses when providers completed nine college credits. Three interagency agreements support the partnership. These agreements are with (1) the community college to deliver three college courses (each three credits) that would meet the requirements of the Infant-Toddler CDA credential, (2) each child care provider asking them to commit to the class schedule and assignments, and (3) each center or home that reflects the administrator's support for the child care providers' continuing education and for the center's compliance with the performance standards. This last agreement also allows Early Head Start staff to visit the child care site unannounced and provide weekly or biweekly reflective supervision and support.

With this approach, every stakeholder benefits:

Child Care Centers/Homes. Early Head Start pays the centers/homes their full tuition rates, staff further their education, and centers adopt and reflect the Head Start Program Performance Standards.
Child Care Providers. Providers enhance their education and receive ongoing peer support; shared Individualized Family Support Plans (IFSPs) guide the provider's understanding of each child's strengths and needs.
Families. Parents learn how to identify and choose quality child care, access quality child care, and advocate for their child through the development of one coordinated IFSP.
Early Head Start Programs. Sites have flexibility in establishing program options and services that meet individual family needs. They ensure that families have access to quality infant-toddler care on a flexible schedule, promote shared responsibility and accountability, and leverage funds. They strengthen the child care system for all children in the community.

This approach contains challenges. It takes time for centers and homes to meet the Head Start Performance Standards. Money is often needed to upgrade facilities, and monitoring home providers is a difficult task. However, Project EAGLE is making a positive impact on the community. Last year, it arranged child care for 350 infants/toddlers in addition to the 200 it is funded to serve. Partnerships create systems change, and the whole community benefits.

f. Impacts on Receipt of Services for Children with Disabilities

The Early Head Start programs did not significantly increase the percentage of children with identified disabilities, but they did increase the percentage receiving early intervention services. The percentage with identified disabilities (as reported by parents) was low for both program and control children (four and three percent, respectively) through the second follow-up period (Figure III.7).

The percentage of children who parents reported as receiving early intervention services was also low for both program and control children, but program children were significantly more likely than control children to have received disability services by the time of the second followup (three percent of program children, compared with two percent of control children) (Figure III.7).

FIGURE III.7

IMPACTS ON RECEIPT OF EARLY INTERVENTION SERVICES FOR FOCUS
CHILD BY THE SECOND FOLLOWUP

Figure 3.7: Impacts on Receipt of Early Intervention Services for Focus Child by the Second Followup [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after enrollment.

Notes: All means are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant.

g. Impacts on Child Health Services and Child Health Status

Nearly all children in both the program and the control groups received some health services, which reflects the accessibility to health services afforded by Medicaid and State Children's Health Insurance Programs. Few impacts on children's receipt of health services or their intensity were significant during either follow-up period (Table III.1). The few impacts on health services that were significant during the first follow-up period (the impacts on immunizations, hearing testing, and any health services) were small and did not persist through the second follow-up period (not shown).

By the time of the second followup, program children had visited a doctor for treatment of illness significantly more often than control children (four compared with three visits, on average) (Table III.1). Program families may have gained better access to health care for their children than control families, or they were more likely to take their children to a doctor when they were ill. Alternatively, program children may have become ill more often than control children and needed more frequent treatment.

Parents' reports of the health status of their children at 14 months of age suggest that program children were not as healthy as control children at that age (Table III.1). The impacts of programs on parents' reports of their children's health status were largest among center-based programs, where the impacts on use of center-based child care were the largest. Early Head Start children in those sites were probably exposed to more contagious illnesses in child care centers and became ill more often than control group children in those sites.

A closer look at how health services are provided at the local level suggests both the complexity of delivering these services and program strategies that can be effective, particularly with families who have no other access. These are described in Box III.5. In Box III.6 we see a special local focus on nutrition and health status.

TABLE III.1
IMPACTS ON CHILD HEALTH CARE OUTCOMES DURING THE FIRST 16 MONTHS FOR THE FULL SAMPLE
Outcome	Program Group	Control Group	Estimated Impact per Eligible Applicanta
Percentage of Focus Children Who Visited a Doctor:
For any reason	92.4	92.9	-0.5
For a check-up	87.6	87.9	-0.3
For treatment of an acute or chronic illness	70.9	69.8	1.2

Average Number of Doctor Visits:	4.2	4.1	0.2
For checkups
For treatment of an acute or chronic illness	4.0	3.4	0.6**

Percentage Who Had Sufficient Well-Child Doctor Visits During Their:
First year	77.0	74.8	2.2
Second year	74.1	72.5	1.7

Percentage of Focus Children Who Visited An Emergency Room	42.0	39.8	2.6
			&
Average Number Of Emergency Room Visits:
For any reason	0.9	1.0	-0.0
For treatment of accident/injury	0.1	0.1	-0.0

Average Number of Hospitalizations During Child's:
First year	0.4	0.3	0.1
Second year	0.2	0.2	0.0

Average Number of Nights Hospitalized During Child's:
First year	1.4	1.2	0.2
Second year	0.5	0.8	-0.3
Average Percentage of Focus Children Who: Visited A Dentist	10.6	9.8	0.8
Received Any Immunizations	97.4	96.7	0.7

Average Percentage of Children Who Received:
Any screening test	54.9	52.9	2.0
A hearing testing	30.1	28.8	1.2
A lead test	21.8	23.4	-1.6

Average Percentage of Children Who Received Any Health Services	99.5	99.4	0.1

Average Parent-Reported Health Status Of Child:
When child was 14 months old	3.6	3.7	-0.1***
When child was 24 months old	3.8	3.9	-0.1

Percentage Who Were Reported By Parents To Be In Fair Or Poor Health
When child was 14 months old	18.7	16.0	2.7*
When child was 24 months old	12.2	12.9	-0.8
Sample Size	1,139	1,097	2,236
SOURCE: Parent services follow-up interviews conducted approximately 7 and 16 months after random assignment and parent interviews conducted when children were 2 years old. NOTE: All impact estimates were calculated using regression models, where each site was weighted equally. aThe estimated impact per eligible applicant is measured as the difference between the regression-adjusted means for all program and control group members. bPrimary caregivers rated their children's health status on a scale of 1 (poor) to 5 (excellent). Significantly different from zero at the .10 level, two-tailed test Significantly different from zero at the .05 level, two-tailed test. **Significantly different from zero at the .01 level, two-tailed test.

BOX III.5

VENICE FAMILY CLINIC CHILDREN FIRST PROGRAM HEALTH SERVICES PROVE SUCCESSFUL

JoEllen Tullis and Karen Lamp
Venice Family Clinic Children First Early Head Start

The mission of the Venice Family Clinic (VFC) is to provide affordable, accessible, and compassionate comprehensive primary health care for people with no other access to such care. One of the clinic's guiding principles is that clients are partners in their health care and that health care happens in the context of the cultural, social, physical, emotional, and economic needs of the client. As a result of this commitment, VFC sought and received funds to operate the Children First Early Head Start program.

The program's mission is to optimize the quality of life for children prenatal to age 3 by strengthening families and communities. To achieve this, children and families must be healthy. The first steps toward reaching the desired outcome of healthy children and families are to help families access insurance and to connect them to a medical home. Proven consequences of being uninsured include limited and delayed access to needed services, poorer physical and mental health, premature death, and a diminished capacity to contribute to one's family and community. Children First Early Head Start helps all its families determine whether any family members are eligible for any insurance programs. VFC becomes the medical home for families that are not insurable. At VFC, families receive free quality primary health care and can access a variety of services. These services include health education, developmental screening, diagnostic tests, chronic care treatment, medication, nutrition counseling, ophthalmology/optometry (including free glasses), case management, and social work. They also include mental health services, which provide crisis, individual, and family counseling, and group support and education programs (for example, parenting, prenatal, battered women). The clinic also has a warm line to answer basic child development concerns and questions about parent/child classes. Because Children First Early Head Start home visitors understand the scope of services at the clinic and (with family permission) have access to their physician and multidisciplinary case conferences, the families are more likely to take advantage of these services, seek care in a timely manner, and adhere to treatment plans.

Having Early Head Start as part of the clinic has led to operational changes at the clinic that provide advantages to all patients. Children First Early Head Start has enhanced the ability of VFC staff to (1) understand the importance of the early years and how those years affect an individual in the future, (2) see patients in the context of their families rather than individuals in a state of disease, and (3) look beyond the medical model and embrace the services of social work. The relationship has also led VFC to create a literacy program for pediatric patients, to strengthen the Health Education Department with its focus on primary prevention and community outreach, and to infuse resources into behavioral modification/risk reduction and identification of victims of domestic violence. All physicians screen for domestic violence, and the clinic now has a domestic violence specialist, an advocate to help victims through the court system, and an ongoing support group.

Substantial quantitative and qualitative data show that this comprehensive approach to health care makes a difference. Compared to county averages, Children First Early Head Start families fare much better in rates for both number of uninsured and incomplete immunizations.

"He who has health, has hope; and he who has hope, has everything." --Arabian proverb

BOX III.6

DIET QUALITY BY FOOD INTAKE AND MEALS IN LIMITED-INCOME
MOTHER-INFANT PAIRS IN JACKSON, MICHIGAN

Seung-yeon Lee, Sharon Hoerr, and Rachel Schiffman
Michigan State University

Low-income families are at high risk for poor nutritional status and health. Low socioeconomic status (SES) groups show higher incidences than high SES groups of premature and low-birthweight babies, growth and developmental retardation in infants/toddlers, and chronic diseases such as heart disease, stroke, and some cancers. Poor diet is a factor in these conditions that is sometimes overlooked by child development specialists. Furthermore, despite the importance of diet to growth, limited research exists on the dietary quality of infants and toddlers.

Participants for this study were 181 mother-infant pairs eligible for Early Head Start. Mothers were interviewed in their homes about many aspects of parenting, service use, and family health habits. Interviewers obtained 24-hour dietary recalls of both the mothers (average age 23.3 years, + 5.2) and their infants at or near the time of enrollment (average age of infants was 6.4 months, + 3.3) and again when the infants were about 14 months old. Questions were asked at the first interview about consumption of nutritional supplements and participation in the Special Supplemental Nutrition Program for Women, Infants and Children (WIC), Food Stamp, and Medicaid programs. Foods were entered by type and subdivided by the major food groups. The dietary quality of infants at the first data collection was classified according to the U.S. Department of Agriculture's guideline for WIC, including the age-/amount-appropriate intake of formula, juice, milk, grains, vegetables and fruits. The dietary quality of mothers and 14-month-old children was examined by food group and by skipping breakfast.

For the two time points, 119 cases could be matched. The percentage of mothers using WIC and Medicaid was 87.5 and 88.7, respectively. Only 58.3 percent of mothers reported receiving food stamps. Most of the mothers (91.5 percent) had inappropriate diets. About two-thirds of mothers consumed a vegetable or dairy food, but fruit consumption was very low at both time points. Mothers' diets were also fairly consistent from the first to the second time point, with only about half of mothers consuming foods from four or five of the food groups. Most infants (82.5 percent) were not fed according to the WIC guidelines. Infants consumed formula in inappropriate amounts and were fed juice, fruit, grains, and vegetables at younger ages than recommended (only 11 infants were breast-fed). Fruits and vegetables were the least frequently consumed food groups for toddlers, but more than 50 percent of toddlers consumed from the five food groups. The percentage of skipped meals was higher for mothers than for toddlers. Forty-one percent of mothers skipped breakfast, but toddlers rarely missed a meal. A poor diet for a mother usually predicted a poor diet for her infant at both time points. There was no relationship between services received and dietary quality.

Even though these limited-income families received health services and most were in WIC, diet quality of most mothers was poor and remained so. Fruits and vegetables were the food groups least likely to be consumed by mothers and toddlers. Infants were often fed inappropriately, although, by 14 months of age, the quality of the children's diets had improved slightly.

h. Impacts on Receipt of Family Health and Other Family Development Services

Because almost all program and control families received some health services, none of the estimated program impacts on receipt of family health s ervices was significant (Table III.2). The Early Head Start programs also did not have a significant impact on families' receipt of mental health services (Table III.2).

The estimated program impact on primary caregivers' self-reported overall health status was negative and significant when their children were 14 months of age. The primary caregivers of program children may not have been as healthy as the primary caregivers of control children, or may have been more aware of health issues and more likely to take them into account when rating their overall health status. The early impact on primary caregivers' self-reported health status, however, did not persist when children were 24 months of age (Table III.2).

An important focus of Early Head Start services was families' self-sufficiency goals and their efforts and progress toward them. The programs substantially increased primary caregivers' receipt of education-related services (school or job training program participation and/or discussion of education topics with a case manager): 83 percent of program families compared with 51 percent of control families received education-related services during the first 16 months after enrollment (Figure III.8). 6 Significantly more program than control families participated in an education or job training program (48 compared with 44 percent). In addition, significantly more program than control families received employment-related services (job search assistance and/or discussion of employment with a case manager): 67 compared with 29 percent by the time of the second followup (Figure III.8).

The Early Head Start programs increased families' receipt of some kinds of assistance designed to help families become self-sufficient and facilitate their access to other critical support services. Most important, the programs increased families' receipt of transportation assistance. Significantly more program than control families received transportation assistance (29 compared to 19 percent) by the second followup (Figure III.8). The programs' impact on receipt of transportation assistance increased over time-from 5 percentage points at the first followup (not shown) to 10 percentage points at the second followup.

TABLE III.2
IMPACTS ON FAMILY HEALTH CARE AND HEALTH STATUS
Outcome	Program Group	Control Group	Estimated Impact Per Eligible Applicanta
Percentage of Families Who Received Any Family Health Services	98.0	97.9	0.0

Percentage of Families Who Received Any Mental Health Services	17.2	16.2	1.1

Average Self-Reported Health Status Of Parent Or Guardian:
When child was 14 months old	3.5	3.6	-0.1**
When child was 24 months old	3.5	3.5	0.0
Sample Size	1,139	1,097	2,236
SOURCE: Parent services follow-up interviews conducted approximately 7 and 16 months after random assignment and parent interviews conducted when children were 2 years old. NOTE: All impact estimates were calculated using regression models, where each site was weighted equally. aThe estimated impact per eligible applicant is measured as the difference between the regression-adjusted means for all program and control group members. bPrimary caregivers rated their own health status on a scale of 1 (poor) to 5 (excellent). Significantly different from zero at the .10 level, two-tailed test Significantly different from zero at the .05 level, two-tailed test. **Significantly different from zero at the .01 level, two-tailed test.

FIGURE III.8

IMPACTS ON FAMILY DEVELOPMENT SERVICES BY THE SECOND FOLLOWUP

Figure 3.8: Impacts on Family Development Services by the Second Followup [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after enrollment.

Education-related services include school, job training, and/or discussion of education and training with a case manager. Employment-related services include job search assistance and/or discussion of employment issues with a case manager.

2. Differences in Program Impacts on Receipt of Key Services Among Targeted Subgroups of Programs

Beyond the overall impacts on service receipt described in the previous sections, it is important to explore variations in impacts on service receipt among targeted subgroups of programs.7 Variations in program impacts on service receipt may help explain differences in program impacts on child and family outcomes for subgroups of programs and may highlight successes and challenges that the particular groups of research programs experienced in providing services to families.

Caution must be used in interpreting the variations in impacts on service receipt among subgroups of programs. The subgroups are defined on the basis of a single program characteristic but may differ in other characteristics. These other unaccounted-for variations in program characteristics may also influence the variations in impacts on service receipt. Thus, in our analyses, we focus on patterns of impacts across outcomes and consider the potential role of other differences in characteristics that may have influenced the outcomes being examined (Table II.7 and Appendix Tables E.VII.31 and E.VII.32 show the configuration of family characteristics across the research sites).

The subgroup analyses show that the impacts of the Early Head Start research programs on service receipt were broad-based and not limited to a particular subset of programs. The estimated impacts on families' receipt of key services were large and significant in nearly all the program subgroups we examined.

Although the impacts on service receipt were large for all groups of programs, the magnitude of the impacts varied among subgroups, usually in expected directions. The variations in the size of the impacts sometimes reflect differences among key groups of programs in the extent to which program families received services and sometimes reflect differences in service receipt by control families among the subgroups, probably as a result of differences in the availability of services across communities. The following sections highlight variations in impacts on service use among subgroups of Early Head Start research programs, variations that can inform our understanding of what program features may promote higher levels of participation and service receipt.

a. Variations in Impacts by Initial Program Approach

As described earlier, the Early Head Start programs adopted different basic approaches to providing child development services, based on the unique needs of the children and families in their communities. In 1997, four programs offered center-based services only, seven offered home-based services only, and six offered both home- and center-based services (in other words, took a mixed approach).

We expected to find differences in program impacts on service receipt that reflected the different approaches these programs took to serving families and children. In general, the variations in impacts are consistent with our expectations. Home-based programs had the largest impacts on the receipt of home visits and group parenting activities, and center-based programs had the largest impacts on the receipt of center-based child care and the amount of center-based care received. Mixed-approach programs tended to produce impacts that were between those of home- and center-based programs, but were often closest in magnitude to the impacts of home-based programs (Figure III.9).

Overall, home-based and mixed-approach programs had the largest impacts on the receipt of any key services (home visits, center-based care, case management, and/or group parenting activities), and home-based programs had the largest impacts on the use of core child development services (home visits and/or center-based care) (Figure III.9 and Appendix Table E.III.1). These differences reflect both lower receipt of services by program families in center-based sites and greater receipt of key services by control families in these sites.

Only center-based and home-based programs had significant impacts on the identification of children with disabilities and receipt of early intervention services. Center-based programs increased identification of children with disabilities by four percentage points and increased receipt of early intervention services by four percentage points (see Appendix Table E.III.1).

Home-based programs had slightly smaller significant impacts on these outcomes (three and two percentage points, respectively).

The home-based and mixed-approach programs had the largest impacts on receipt of education services, employment-related services, and transportation assistance. The center-based programs, which were located in areas where control families were much less likely to report receiving housing assistance, significantly increased receipt of such assistance (Appendix Table E.III.1).

FIGURE III.9

SELECTED IMPACTS ON SERVICE RECEIPT BY PROGRAM APPROACH IN 1997

Figure 3.9: Selected Impacts on Service Receipt by Program Approach in 1997 [D]

Source: Parent Services Follow-Up Interviews conducted approximately 7 and 16 months after random assignment.

Notes: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant.

Core services include at least one home visit or center-based child care.

* Program impact is significantly different from zero at the .10 level, two-tailed test.
** Program impact is significantly different from zero at the .05 level, two-tailed test.
***Program impact is significantly different from zero at the .01 level, two-tailed test.

b. Variations by Degree of Program Implementation

Based on the ratings developed in the implementation study, the research programs differed in their patterns of overall program implementation. As summarized in Chapter I and reported more fully in Pathways to Quality (Administration on Children, Youth and Families 2001b), six programs were rated as fully implemented in fall 1997 (early implementers), six were not rated as fully implemented in fall 1997 but were rated as fully implemented overall in fall 1999 (later implementers), and five were not rated as fully implemented at either time (incomplete implementers). The incomplete implementers either emphasized family support (with less emphasis on child development) or faced difficult implementation challenges (such as early staff turnover in leadership positions or partnerships that did not work out well).

Early implementation was associated with larger impacts on receipt of core services. Although programs in all three groups significantly increased service receipt, for core services (home visits and center-based child care), the impacts were consistently largest among programs that became fully implemented early (Figure III.10 and Appendix Table E.III.2). The differences in impacts by level of implementation were largest for receipt of frequent home visits and for use of center-based child care. In addition, only early implementers increased receipt of housing assistance, and only early and later implementers increased receipt of transportation assistance.

FIGURE III.10

SELECTED IMPACTS ON SERVICE RECEIPT BY PATTERN OF IMPLEMENTATION

Figure 3.10 : Selected Impacts on Service Receipt by Pattern of Implementation [D]

The large impacts of early implementers on receipt of core services are due to higher levels of service receipt in the program group, not lower levels in the control group.

Incomplete implementers had larger impacts than later implementers did on several measures of service receipt, including receipt of frequent home visits, case management, and frequent case management. Several of the incomplete implementers focused on family support and provided case management to many families in home visits but did not fully implement child development services. Control families in the sites served by programs that were fully implemented early were much more likely to receive any case management, and the impact on case management in these sites was smaller than in the other groups of sites.

In the first follow-up period, some impacts on receipt of health services were significant for early implementers. By the time of the second followup, however, most impacts on receipt of health services were not significant. Although this suggests that early implementers helped some families receive health services a few months earlier than they otherwise would have, the impacts were not sustained through the second follow-up period (not shown).

The research programs also varied in their implementation of child development services, which were rated separately as part of the overall implementation rating process.8 Six programs fully implemented child development services early and sustained full implementation over time. Five programs fully implemented child development services in fall 1997 or fall 1999 but not both. Some were later implementers of child development services, and some fully implemented child development services early but were not able to sustain full implementation. Because the numbers of each were small, we combined these programs into one group: the single-period implementers. Six programs did not fully implement child development services by fall 1999. Ten programs received the same rating for implementation of child development services that they received overall, while the remaining programs received higher or lower ratings on this area of program services. The overlap in ratings makes it difficult to know whether it is overall implementation or implementation of family development services that accounts for the subgroup findings.

The program impacts on the use of center-based child care, the use of concurrent arrangements, and the out-of-pocket costs of child care were larger in programs that fully implemented child development services early and were small in programs that did not fully implement them (Appendix Table E.III.3). The patterns of impacts on the other measures of service use are not as intuitive. In particular, the programs that never fully implemented child development services provided home visits to a higher proportion of families and provided intensive home visits to more families than the programs in the other groups. The similar pattern of case management receipt among program families in the three groups suggests that it is the provision of case management services during home visits, not the provision of child development services, that is responsible for the relatively large impacts on receipt of home visits among programs that never fully implemented child development services.

Based on the ratings of implementation of family development services that were developed in the implementation study, the research programs also varied in their implementation of family development services.9 Seven programs fully implemented family development services early and sustained full implementation over time. Seven programs fully implemented such services in fall 1997 or fall 1999 but not both. Three programs had not fully implemented family development services by fall 1999. For 12 programs, the rating of implementation of family development services reflected the program's overall implementation rating, while for most of the remaining programs, the rating of implementation of family development services was higher than the rating for overall implementation. Again, it is difficult to determine whether variations in overall implementation or in implementation of family development services account for variations in impacts.

Programs that fully implemented family development services early had substantially larger impacts on the receipt of case management (and home visits) at least weekly (Appendix Table E.III.4). They also had slightly larger impacts on receipt of education-related services, family health services, and transportation assistance.

c. Variations by State Work Requirements for Mothers of Infants on Welfare

Seven research programs were located in states that require mothers who have infants under age 1 and who receive welfare cash assistance to meet work requirements, and 10 were located in states that exempt mothers with infants from work requirements. Parents who are required to work are likely to have a greater need for child care and employment-related services. Among the seven programs located in states with work requirements for parents of infants, three were center-based programs, one was a mixed-approach program that offered center-based care to some program families, and three were home-based programs.

Early Head Start programs had their largest impacts on families' receipt of any key services in locations without early work requirements, mainly because they had a much larger impact on receipt of case management (Figure III.11 and Appendix Table E.III.5). They also increased the use of any child care in states without early work requirements, but did not significantly increase the use of any child care in states with early work requirements.

The Early Head Start programs had somewhat larger impacts on the use of center-based care in sites with early work requirements. They increased the use of any center-based care, increased the hours per week of center-based care used, and reduced the out-of-pocket costs of child care of families in both groups of sites (Figure III.11 and Appendix Table E.III.5). However, the impacts on the amount of center-based care used and out-of-pocket child care costs were larger in the sites with early work requirements.

To meet parents' needs, more case management and child care services appear to be available from community sources in locations where mothers of infants are required to