Executive Summary

In October 1993, the Head Start Bureau (HSB) of the Administration on Children, Youth and Families funded 16 Head Start grantees to implement demonstration projects for homeless children and their families. The objectives of this initiative were to (1) enhance access of homeless families to Head Start services; (2) provide services responsive to the special needs of homeless children and families; (3) identify effective methods of addressing the needs of home-less families; and (4) implement and document replicable strategies for collaboration between Head Start programs and community agencies on behalf of homeless families. At the end of the three-year demonstration period, the HSB transitioned demonstration funds to the Grantees' regular Head Start budgets to promote continuation of services for homeless families. Based on reviews of the projects' final reports and telephone discussions with project administrators, this report provides the following information about the demonstration projects.

• The characteristics of the projects and families served.
• Critical issues relevant to project implementation.
• Challenges encountered and effective practices with respect to each issue.
• Key lessons learned.

Description of Demonstration Projects and Families Served

Four of the 16 demonstration projects integrated Head Start services into ongoing shelter or transitional housing facility services for homeless families; three offered traditional Head Start services and collaborated with other community agencies to provide housing and self-sufficiency-related services; and nine provided housing and self-sufficiency-related services, as well as traditional Head Start services. Project characteristics with regard to duration of services, type of Head Start program, site location, ages of children served, and services provided by project staff are presented in Table 1.

Duration of Services: There was general agreement among all grantees that full-day, full-year Head Start services were necessary to adequately meet the needs of homeless families. Most grantees that were unable to provide full-day, full-year services during the demonstration period were considering options for doing so in the future.

Type of Center: Some grantees suggested that it was necessary to establish special Head Start Centers for homeless children because traditional Head Start classrooms do not meet homeless children's developmental and socio-emotional needs. Other grantees suggested that it was more appropriate to integrate homeless children into existing Head Start Centers because integration provides homeless families with opportunities to interact with families who are not homeless. The grantee implementing a Family Child Care Provider model suggested that this was a particularly appropriate service approach for homeless families because it offers homeless children the consistency and individualized attention they need, fosters the development of close personal relationships between providers and parents, and ensures a home-like environment in which siblings can be served together.

Ages of Children: Although most projects served children between three and five years of age, there was general consensus among grantees that meeting the comprehensive needs of homeless families requires serving infants and toddlers as well as pre-school-aged children.

Location of Center Facilities: The advantages to locating Head Start services in a shelter or transitional housing facility were: (1) transportation was not a problem; (2) families had easy access to classrooms, parent events, and project staff; and (3) services coordination was fostered by daily contact between project and shelter facility staff. A disadvantage of this location was that it was difficult to retain children and families in Head Start project services when families left the shelter, even when projects encouraged continued involvement.

An advantage of locating Head Start class-rooms and services in community sites was that retention of families was not contingent on duration of shelter stays. Two disadvantages to this type of location were: (1) families often lacked transportation to access the classrooms and/or parent activities, and (2) it was more difficult to coordinate services with collaborating agencies. In most cases, projects lacked resources to provide transportation for the families.

Project Services: Nine grantees hired special project staff to help families obtain housing, jobs, and self-sufficiency-related services. In the other seven projects, these services were provided by collaborating agencies. Eight projects provided families with transportation services, either directly through use of a van or bus, or indirectly through vouchers or cash for public transportation. Three grantees maintained funds to help families when they needed money to pay utility bills or security deposits, or to purchase appliances, window coverings, or bedding when they could not obtain them elsewhere.

Family Characteristics: Grantees located in large cities in the East and Midwest tended to serve primarily African American families. Two projects served a large percentage of Caucasian families, while other projects served a mixture of Caucasian, Hispanic/Latino, and African American families. In most projects, participating families were primarily single-parent and female-headed, with only three projects serving a significant percentage of two-parent families. The majority of families served were public assistance recipients at the time of project intake and were residing in some type of shelter or transitional housing facility. Shelter facilities varied with respect to duration of shelter stay (ranging from 30 days to two years) and the populations served (victims of domestic violence, teen mothers, homeless women with young children, and two-parent families). Some projects also served families who were residing with relatives or friends ("doubled up"), living in motels, and/or about to be evicted because they could not pay their rent. Frequently cited reasons for homelessness were domestic abuse (eight projects), substance abuse (seven projects), the breakup of a woman's relationship with a male partner (eight projects), and lack of money to pay rent (five projects).

Table 1: Number of Demonstration Projects Exhibiting Particular Characteristics

PROJECT CHARACTERISTICS	# OF PROJECTS
Duration of Educational Services
Full Day (at least 8:30 a.m. to 5:30 p.m.)	7 projects
Extended day (at least 6 hours)	1 project
Half-day (eg., approximately 3 hours)	5 projects
Both full-day and half-day classes	3 projects
Full week	14 projects
3-4 days per week	2 projects
Type of Program
Special Centers for homeless children	9 projects
Integration of homeless children into community Head Start Centers	3 projects
Both special and integrated Centers	3 projects
Family child care providers (no classrooms)	1 project
Ages of Children Targeted for Services
Three- to five-year-olds only	11 projects
Both three- to five-year-olds and infants and toddlers	5 projects
Location of Center Facilities (Classrooms)
In a shelter or transitional housing facility	4 projects
In various community sites	10 projects
In both housing facilities and community sites	1 project
Services Provided by Project Staff
Mainly Head Start comprehensive services	7 projects
Self-sufficiency and housing-related services in addition to Head Start services	9 projects
Transportation services	8 projects
Emergency funds	3 projects

Implementation Issues, Challenges Encountered, and Effective Practices

Critical issues with respect to implementing Head Start services for homeless families were:

• Building effective collaborative relationships.
• Recruiting, enrolling, and retaining homeless families.
• Involving homeless parents in Head Start.
• Meeting the unique needs of homeless children and parents.

Building Effective Collaborative Relationships: Grantees developed three types of collaborative relationships: (1) collaborations involving coordination of services, (2) collaborations developed to recruit families to the project, and (3) collaborations established to access services for project families. All grantees reported that collaboration was a challenging task because of the necessity to build trust with partners, convince partners of the importance of Head Start services for homeless families, and establish open and effective communication with partners.

Strategies reported to be effective in building trust were:

• Including potential partners early in the project's planning phase.
• Developing a project advisory committee including representatives from partner agencies.
• Representing the project on existing coalitions of homeless services agencies.
• Emphasizing the benefits of collaboration to potential partners.
• Inviting partner agency staff to participate in project activities.

Strategies reported to be effective in convincing partners of the importance of developmental services for children were:

• Informing partners about the importance of early childhood development services for the well-being of parents as well as children.
• Educating partners about the efficacy of a family perspective in providing services to homeless families.
• Identifying homeless children's needs, the lack of services for homeless children in the community, and the ways that Head Start can fill these gaps.

Strategies that promoted open communication between project and partner agency staff were:

• Developing interagency case management teams or joint planning and case staffing teams.
• Co-locating Head Start staff and collaborating agency staff in the same sites.
• Identifying a staff person from the partner agency with a similar philosophy to serve as a primary contact.
• Establishing formal agreements specifying collaborating partners' roles and responsibilities.

Recruiting, Enrolling, and Retaining Homeless Families: Several demonstration projects encountered problems recruiting, enrolling, and/or retaining homeless families in project services, particularly during the first year of project operations. The strategy that was found to be most effective in enhancing recruitment was educating other agencies about Head Start, the projects' services, and the eligibility requirements. Strategies that tended to enhance enrollment were those that reduced the amount of time necessary to obtain required documents, such as:

• Enrolling children in classes as soon as they have immunization records and establishing a 30-day grace period for obtaining other documentation.
• Helping families navigate the bureaucracy to obtain necessary documentation.
• Establishing relationships with health care agencies to provide free physical examinations and immunizations for homeless children on an expedited basis.
• Providing services to parents even before a child's enrollment in the classroom.

Retention of families was fostered by strategies that addressed the high mobility of homeless families, including:

• Developing working relationships with shelter staff that allowed demonstration project staff to intervene with families if an eviction was imminent.
• Encouraging families to continue bringing their children to Head Start classes even when the family is no longer in the shelter.
• Providing transportation to all families to facilitate access to Head Start classrooms.
• Establishing close personal relationships with parents so that they perceive project services as enhancing their own well-being as well as their children's.

Involving Parents in Head Start: Although most grantees experienced difficulties involving homeless parents in Head Start activities, parent involvement was reported to be improved by the following practices:

• Holding parent meetings at times and locations that accommodate parents' schedules.
• Structuring parent meetings as social occasions or gatherings.
• Developing creative ways in which parents can be involved without being in the classroom, such as assisting on the bus as a monitor or recording books and songs on tape to be played in the classroom.
• Coordinating with social services departments or shelters so that parents can obtain "work" or "volunteer" credits for the hours they work in the classroom.
• Hiring a homeless parent to serve as a program advocate because that person was effective in reaching out to homeless families.
• Establishing a contract with parents to contribute a specific number of hours.

Meeting the Multiple Needs of Homeless Families: Grantees described the following practices as being effective in meeting the multiple service needs of homeless children:

• Implementing a Primary Caregiver Model in the class-room in which each child is assigned to a caregiver responsible for supporting and nurturing that child.
• Providing children with a lot of physical attention and validation of feelings.
• Establishing a classroom environment that has limited amounts of materials and toys, food available at all times, consistent and repetitive routines, and space for quiet times.
• Hiring a child development specialist to work with teachers, children, and families.
• Hiring teachers who have qualifications that go beyond the early childhood education (CDA) certificate requirement of Head Start.

For the most part, service needs of homeless parents were addressed by referring them to other community agencies. However, several projects were directly involved in providing housing-related services to homeless parents. Effective practices for helping parents obtain housing were:

• Employing a staff person to work with realtors, apartment managers, and housing administrators to advocate on behalf of project families.
• Providing transportation to visit apartments and meet with realtors, writing letters of recommendation on behalf of families, and helping families complete forms.
• Providing emergency funds to aid families in paying utility bills and security deposits and to help them purchase appliances, window coverings, or bedding.

Lessons Learned from the Demonstration Projects

The key lessons learned in the process of implementing services for homeless children and families were the following:

• Homeless parents need as much support and nurturance as their children.
• The needs of homeless children often are overlooked by agencies serving homeless families.
• Transportation services are essential.
• Homeless families require children's programming for at least nine hours a day.
• Head Start is a "newcomer" in the homeless services area, and time is needed to establish credibility before implementing project services.
• It is sometimes difficult to obtain services for some families, because they are not considered to be homeless according to definitions of many community services agencies and the McKinney Act (e.g., families doubled up with other families are not considered homeless).
• Some changes in welfare policies and practices have often made it more difficult to serve homeless families and more difficult for homeless families to access the services they need.

Implications for Serving Homeless Families

The experiences of the 16 demonstration projects suggest that there are a number of effective approaches to serving homeless families, and that the efficacy of any particular approach often depends on the characteristics of the community with respect to its resources, local policies, and service delivery systems for homeless families. The projects' experiences also indicate that Head Start has a critical role to play in serving homeless families, and in many communities it may be the only agency serving homeless families that focuses on children as well as parents. In addition, the fact that Head Start employs a family-based, comprehensive services approach to serving families means that it is in a unique position to provide homeless families with the kinds of services necessary to meet their multiple needs. Finally, a key implication of the demonstration projects' experiences is that Head Start programs cannot "do it all." Collaboration with other agencies serving homeless families was a critical element of each demonstration project. However, collaboration often proved to be a challenging task, and many Head Start programs desiring to expand their services to homeless families may require training and technical assistance in developing effective partnerships in their communities. Collaboration also was found to be problematic in communities in which services to homeless families were limited. This suggests that Head Start Centers may have a role in advocating for greater services for these families in their communities or may need to expand their own services to meet existing service gaps.