Skip Repetitive Navigation Links
California State Auditor Logo COMMITMENT • INTEGRITY • LEADERSHIP

Youth Experiencing Homelessness
California’s Education System for K–12 Inadequately Identifies
and Supports These Youth

Report Number: 2019-104

Use the links below to skip to the chapter you wish to view:



Chapter 1

SOME LEAS HAVE NOT ADEQUATELY IDENTIFIED AND SERVED YOUTH WHO ARE EXPERIENCING HOMELESSNESS

Chapter Summary

Some LEAs in California are not identifying and providing appropriate services to youth experiencing homelessness in their jurisdictions, which may affect the performance outcomes of these youth on measures such as chronic absenteeism, suspension, and graduation. Considering that the goal of the McKinney‑Vento Act is to ensure that each youth experiencing homelessness has equal access to the same public education as provided to other youth, the consistent gap in critical performance outcomes between youth experiencing homelessness and their peers is concerning. Significant differences in the level of services that LEAs are providing to some youth experiencing homelessness may affect their chances for success. For example, when we reviewed two comparable LEAs, we found that one—which provided a high level of services—reported that its youth experiencing homelessness performed better than the statewide average on performance outcomes. The other—which provided a significantly lower level of services—reported that the rates of absenteeism, suspension, and dropping out for youth experiencing homelessness were more than twice the statewide average.

Further, most of the six LEAs we visited—Birmingham Community Charter High School (Birmingham Charter), Greenfield Union School District (Greenfield), Gridley Unified School District (Gridley), Norwalk‑La Mirada Unified School District (Norwalk‑La Mirada), San Bernardino City Unified School District (San Bernardino), and Vallejo City Unified School District (Vallejo)—acknowledged that they likely were not identifying all youth who were experiencing homelessness during the academic year. The majority of the six LEAs we visited did not dedicate significant time to their homeless education programs. We found that the more time an LEA’s staff dedicated to administering its homeless education program, the higher the number of youth the LEA identified as experiencing homelessness. Further, although best practices recommend that LEAs use a housing questionnaire to identify such youth, we found the LEAs did not effectively use housing questionnaires. In addition, none of the LEAs followed the federal law requirement to ensure that school personnel who provide services to these youth receive training on their responsibilities or to undertake activities to increase their awareness regarding the educational rights of youth experiencing homelessness. The LEAs we interviewed generally recognized the value of these practices yet have not fully implemented them.

LEAs’ Efforts to Support Youth Experiencing Homelessness Have Not Always Yielded the Desired Results

LEAs have not always been effective in ensuring that youth experiencing homelessness have access to the education and other services that they need to succeed academically. According to Education’s statewide data, youth who were experiencing homelessness during academic year 2017–18 were chronically absent, suspended from school, and dropped out of school at twice the rate of their peers not experiencing homelessness. As Figure 2 shows, LEAs reported that 23 percent of their youth experiencing homelessness were chronically absent, in comparison to 11 percent of their peers. Although numerous considerations can affect student performance outcomes, the consistent gap across different academic indicators between youth experiencing homelessness and their peers suggests that LEAs throughout the State are not ensuring that such youth have equal access to the same free, appropriate public education.

Figure 2
California’s Youth Experiencing Homelessness Are More Likely Than Other Youth to Have Problems That Can Affect Performance Outcomes
Academic Year 2017–18

A bar chart comparing rates of suspension, chronic absenteeism, dropout, and non-graduation for youth experiencing homelessness and those not experiencing homelessness in California in academic year 2017-18.

Source: Data that Education collects from LEAs through CALPADS.


The effectiveness of the homeless education programs for the six LEAs we visited varied significantly. For example, Vallejo and Norwalk‑La Mirada are both located in cities with similar populations and poverty levels and with similar numbers of enrolled students. However, the performance outcomes for Vallejo’s youth experiencing homelessness all fell short of the statewide average, whereas Norwalk La‑Mirada’s youth experiencing homelessness performed better than the statewide average on all measures. During academic year 2017–18, Norwalk‑La Mirada reported that its youth experiencing homelessness had a suspension rate of 4 percent and chronic absenteeism rate of nearly 13 percent, which were better than the statewide averages. However, during the same year, Vallejo’s youth experiencing homelessness were suspended at a rate of 14 percent, two‑and‑a‑half times the statewide average, and they experienced chronic absenteeism at a rate of 60 percent, more than two‑and‑a‑half times the statewide average.

Moreover, youth experiencing homelessness in San Bernardino, the LEA with the largest student enrollment we visited, also performed better on some performance outcomes than the statewide average. During academic year 2017–18, 69 percent of youth experiencing homelessness statewide graduated from high school. That year, San Bernardino reported that nearly 80 percent of its youth experiencing homelessness graduated. Similarly, Norwalk‑La Mirada reported that 85 percent of its youth experiencing homelessness graduated. In contrast, Vallejo’s youth experiencing homelessness graduated at a rate of 40 percent. Vallejo’s superintendent agreed that it needs to do more to improve the performance outcomes of its youth experiencing homelessness; however, he explained that due to budgetary constraints, Vallejo has had to reduce services and personnel, which has affected all of its students, including youth experiencing homelessness.

Although many factors may affect the performance outcomes of youth experiencing homelessness, we found some significant differences in the types of services that these three LEAs collaborate with other entities to provide, which may have played a role in the success of these youth. Specifically, in addition to providing tutoring, transportation, and school supplies within the school setting, Norwalk‑La Mirada has developed collaborations with many service organizations such as those that provide housing assistance, food, health care services, and family counseling, as Table 1 shows. Norwalk‑La Mirada’s local liaison told us that the LEA discovered that youth experiencing homelessness could be more successful when they and their families’ basic needs, including housing, are met. For example, Norwalk‑La Mirada collaborates with nonprofit organizations to arrange for housing assistance and food for the families of these youth. Although, as Table 1 indicates, San Bernardino was unable to provide evidence of its collaboration with providers of social welfare services and health care services, it asserted that it does collaborate with these types of providers. San Bernardino has developed collaborations with other service organizations to coordinate counseling services, meal services, and housing services.

Table 1
Coordination With Service Providers Varied Among the LEAs We Visited

SERVICES		 BIRMINGHAM CHARTER GREENFIELD GRIDLEY NORWALK‑ LA MIRADA SAN BERNARDINO VALLEJO
Counseling X ✔ X ✔ ✔ X
Health care X ✔ X ✔ X ✔
Housing and shelter X ✔ X ✔ ✔ X
Meals X ✔ X ✔ ✔ ✔
Social welfare* X ✔ X ✔ X X

Source: Best practices, LEA interviews, and documentation to demonstrate coordination with service providers.

*   Social welfare services include financial assistance programs and child‑care assistance.

Vallejo’s local liaison told us that he provides youth experiencing homelessness and their families with a list of external organizations that provide services. However, as Table 1 shows, Vallejo does not formally collaborate with other organizations to ensure that the youth and their families receive services related to counseling, social welfare, and housing outside of the school setting. According to its superintendent, Vallejo was unaware that best practices recommend coordinating with local service providers to identify and support these youth. The difference in the level of effort that these three LEAs place on collaborating with outside entities to help youth experiencing homeless and their families to improve their living situations may well have contributed to the difference in performance outcomes of those youth.

Further, Norwalk‑La Mirada and San Bernardino dedicated more time to administering their homeless education programs than did Vallejo. Specifically, during academic year 2017–18, Norwalk‑La Mirada’s local liaison stated that the LEA had dedicated one full‑time and one part‑time staff position to identifying and supporting youth experiencing homelessness. San Bernardino, which had 58,000 students enrolled in academic year 2017–18, or three times the enrollment of Norwalk‑La Mirada, stated that it has four full‑time staff who spend approximately 70 percent of their time identifying and supporting such youth. In comparison, Vallejo’s local liaison is also the director of student services, a position that the superintendent explained has many other responsibilities unrelated to homelessness. Consequently, Vallejo’s local liaison estimated that he spends only a limited number of hours each month on identifying and supporting youth experiencing homelessness.

Although we visited six LEAs, the other three LEAs—Birmingham Charter, Greenfield, and Gridley—are not comparable as their sizes vary and they have identified few youth as experiencing homelessness; thus, comparing their performance outcomes would not result in valid conclusions. For example, Gridley identified only one youth experiencing homelessness during academic year 2017–18, which would result in either a 100 percent or a 0 percent rate of success for each performance outcome. Because they identified such small numbers of youth, we did not include them when comparing performance outcomes and services.

LEAs Could Provide Better Services Through Data‑Sharing Agreements

To leverage staff time and increase the likelihood that youth experiencing homelessness are receiving the services they need to succeed, Education could provide guidance to LEAs on data‑sharing agreements with service providers. HUD administers the Continuum of Care Program, which is intended to promote a communitywide commitment to ending homelessness, to provide funding for efforts by nonprofit providers and state and local governments to quickly rehouse people experiencing homelessness, and to optimize self‑sufficiency among those experiencing homelessness. Federal regulations define a Continuum of Care as a group that consists of representatives from organizations within a specified geographic area, including nonprofit homeless service providers, victim service providers, faith‑based organizations, governments, businesses, advocates, and public housing agencies. Each Continuum of Care is responsible for operating a homeless management information system—a local information technology system used to collect client‑level data as well as data on the provision of housing and services to homeless individuals and families and people at risk of homelessness.

The Continuum of Care Program uses HUD’s definition of homelessness, which includes those whose primary nighttime residence is a public or private place not ordinarily used as a regular sleeping accommodation, such as streets, vehicles, abandoned buildings, parks, or campgrounds. However, unlike the McKinney‑Vento Act, HUD’s definition of homelessness generally does not include youth who live with others due to economic hardship; therefore, the Continuum of Care Program may not provide assistance to all of the youth who qualify for services under the McKinney‑Vento Act. Nevertheless, the Continuum of Care Program could provide valuable assistance to those youth who may qualify as experiencing homelessness under HUD’s definition because they are unsheltered or live in shelters, motels, or hotels. Up to 16 percent, or more than 43,000, of the youth who LEAs identified as experiencing homelessness may have met this description during academic year 2017–18.

The U.S. ED provides guidance on the use of data to foster interagency and community coordination. In 2015, in coordination with the U.S. Interagency Council on Homelessness, the U.S. ED developed an informational brief that provided several examples of the benefits of interagency collaboration. The brief noted that some educational agencies in other states have established data‑sharing agreements with their local Continuum of Care to authorize those agencies to enter the youth’s information—so long as it is not personally identifiable or, if it is identifiable, the LEA received parental consent to do so—into the Continuum of Care’s local homeless management information system. For example, the brief explained that all school districts in one county in Michigan have a data‑sharing memorandum of understanding with the county’s Regional Educational Service Agency (RESA). The schools use a written consent form to allow them to disclose specific data, and RESA enters the information into its Continuum of Care’s homeless management information system. According to the brief, this data sharing enables the RESA to obtain a more complete picture of students experiencing homelessness, which helps with providing services and funding. According to the brief, the data sharing has resulted in HUD‑funded program personnel being better informed of the needs of homeless families, and this has led to improvements in existing housing options for those families.

Education stated that it had not considered providing guidance to assist LEAs in coordinating with their local Continuum of Care. Although Education believes that this guidance would be a good idea, it will not be able to develop such guidance until it has addressed its other plans for guidance outlined in its state plan. Although we recognize that some counties may have already implemented processes, such as referral hotlines, to assist those experiencing homelessness in obtaining referrals for available services, other counties that have not yet done so could benefit from data‑sharing agreements. Therefore, if Education established guidance for LEAs regarding data‑sharing agreements, LEAs would have the information necessary to streamline the process for youth experiencing homelessness to receive services. Education could collaborate with HUD to develop guidelines to assist LEAs in establishing data‑sharing agreements with their local Continuum of Care.

Some LEAs Have Underidentified Youth Experiencing Homelessness

California’s LEAs are likely underidentifying youth experiencing homelessness. Of the six LEAs we visited, the majority identified very few of their economically disadvantaged youth as experiencing homelessness. The local liaisons of five of the six LEAs believe that families and youth experiencing homelessness fear disclosing their living status for a variety of reasons, including stigma, which hinders the LEAs’ abilities to identify all of the youth. Additionally, a common misunderstanding is that people who experience homelessness live without shelter. However, as Figure 3 shows, LEAs reported that 84 percent of all youth experiencing homelessness in California during academic year 2017–18 shared housing with others, such as extended family or friends, due to various reasons, including economic hardship, while only 16 percent lived in hotels, motels, temporary shelters, or without any shelter.

Figure 3
Most California Youth Experiencing Homelessness Lived in Shared Housing During Academic Year 2017–18

A pie chart describing the share of California youth experiencing homelessness who lived in shared housing, in shelters, in hotels or motels, or who were unsheltered during academic year 2017-18.

Source: CALPADS cumulative end‑of‑academic year 2017–18 data that Education does not publish on its website.

Note: The figure is a summary of all the dwelling types for youth experiencing homelessness during the year, so it includes students who may have more than one dwelling type during this period or have a dwelling type in multiple districts..


Although there is no precise method for determining whether an LEA has identified all its youth experiencing homelessness, we found that most homeless education experts use 10 percent of economically disadvantaged youth as a benchmark to determine an LEA’s effectiveness in identifying them. The U.S. ED told us that it is aware that some state education departments assess whether an LEA has effectively identified students experiencing homelessness based on whether the LEA has identified 10 percent of its economically disadvantaged students as experiencing homelessness during the academic year. Similarly, the National Association for the Education of Homeless Children and Youth (NAEHCY), whose mission is to ensure educational equity and excellence for youth experiencing homelessness, explained that LEAs can expect at least 10 percent of their economically disadvantaged students to experience homelessness at some point during the academic year. In addition, Florida’s state coordinator agreed that the 10 percent benchmark is an industry standard and Texas describes that benchmark on its Homeless Education Office’s webpage. The National Center for Homeless Education, on the other hand, believes a more reasonable measure is that 5 percent of economically disadvantaged students experience homelessness each year. The entities we reviewed, including NAEHCY and Education, define economically disadvantaged students as those who are eligible for free or reduced‑price meals under the National School Lunch Program, which requires a student’s household income to meet certain federal poverty level guidelines.

Education has not yet established a benchmark to assess an LEA’s effectiveness in identifying youth experiencing homelessness. Although it is possible for an LEA to have less than 5 percent of its economically disadvantaged youth that experience homelessness, having such a benchmark can help Education and LEAs better gauge their efforts in identifying these youth. In fact, in academic year 2017–18, CALPADS data shows that 74 percent of LEAs throughout California identified less than 5 percent of their economically disadvantaged students as experiencing homelessness. Therefore, in the absence of an established benchmark, we believe Education should be able to identify all LEAs that likely are not effectively identifying youth experiencing homelessness if it employs a benchmark similar to the ones used by other entities—either 5 or 10 percent of economically disadvantaged students.

As Figure 4 shows, only Norwalk‑La Mirada and San Bernardino identified more than 5 percent of their economically disadvantaged youth as experiencing homelessness in academic year 2017–18. The remaining four LEAs identified far fewer youth experiencing homelessness—reporting between zero and 3 percent.To view the statistics of youth identified as experiencing homelessness by county and LEA throughout California, visit our interactive map. For example, although Greenfield had close to 10,000 economically disadvantaged students during academic year 2017–18, it identified fewer than 80 of those students—or less than 1 percent—as experiencing homelessness. This low number indicates that Greenfield may have hundreds more youth experiencing homelessness who are not receiving the needed services and support to succeed. In addition, all but one of the six local liaisons we interviewed acknowledged that their LEAs likely did not identify all youth experiencing homelessness during the academic year. In fact, the state coordinator acknowledged that the underidentification of these youth is an issue throughout the State.

Figure 4
Four of the Six LEAs We Visited Did Not Identify Many Economically Disadvantaged Youth as Experiencing Homelessness

A map of the state of California showing the locations of the six LEAs we visited and a table that shows each of the six LEAs’ total enrollment, number of economically disadvantaged students, number of students identified as experiencing homelessness, and identification rate.

Source: CALPADS cumulative end‑of‑academic year 2017–18 data that Education does not publish on its website. .

* The identification rate is the percentage of economically disadvantaged youth identified as experiencing homelessness. .


Gridley was the only LEA we reviewed whose former local liaison believed its schools identified all youth experiencing homelessness in academic year 2017–18; however, we believe this assessment is likely inaccurate. As Figure 4 shows, Gridley identified that only one of its nearly 1,700 economically disadvantaged students experienced homelessness. Its former local liaison, who retired in 2019, believes that two factors limit the number of youth who experience homelessness in Gridley. First, families mostly stay in the area and typically enroll their youth in Gridley’s schools from kindergarten through grade 12. In addition, minimal job opportunities exist in the area, causing few new families to move into the community. However, considering the 5 percent measure deemed reasonable by the National Center for Homeless Education, we expected Gridley to have identified more than just one youth experiencing homelessness. Of further concern is that Gridley has generally not implemented best practices to identify its youth experiencing homelessness, such as administering an annual housing questionnaire to all students, disseminating information about the McKinney‑Vento Act in public places, and training staff, as we discuss further later. Although the superintendent, who is the new local liaison for Gridley, agreed with the former local liaison’s description of Gridley’s limited job opportunities and residency, he acknowledged that there could have been more than just one youth experiencing homelessness during academic year 2017–18.

Similarly, some school principals and attendance clerks we interviewed claimed that they have identified all youth experiencing homelessness. However, we question their perspectives given that some of these staff were not aware that the legal definition of homelessness includes families who are living with friends or other family members because of economic hardship. Consequently, some attendance staff at a middle school we visited in Greenfield explained that knowing that students from multiple families had a shared home address would not have prompted them to ask questions or refer the students to the district to determine whether the students qualified to receive services under the McKinney‑Vento Act. Because the attendance staff at this school are primarily responsible for identifying such youth, we believe that this school likely has not identified all of its youth experiencing homelessness.

In addition, a principal and assistant principal at an elementary school in Greenfield believe that the two youth the school identified as experiencing homelessness seemed an appropriate number because the school is located in a higher socioeconomic area than other neighborhoods the district serves. However, nearly 900 of the elementary school’s 1,000 students, or roughly 90 percent, are economically disadvantaged. Therefore, based on the 5 percent measure, we expected the school to have identified about 45 youth who experienced homelessness during the academic year. According to the local liaison, Greenfield may not have identified all youth experiencing homelessness. Further, Greenfield’s superintendent and local liaison stated that some school staff, including principals, may believe their assertions given that historically the LEA more narrowly interpreted the definition of homeless as expressed under the McKinney‑Vento Act because often those who share housing do not view themselves as experiencing homelessness.

The majority of local liaisons we interviewed believe that fear or social stigma may discourage many families and youth from disclosing their homelessness. Five of the six local liaisons we interviewed explained that families or youth fear that they may be subject to stereotypes related to that condition. These stereotypes can include that they have mental health issues or are addicted to drugs. They also fear that these stereotypes could lead others to discriminate against them, exclude them, or criminalize them. In fact, even though Norwalk‑La Mirada has identified 29 percent of its economically disadvantaged youth as experiencing homelessness, its local liaison has encountered barriers to identifying all such youth because their families fear being deported by an immigration enforcement authority or having their children taken away by a child protective services agency. Similarly, the local liaisons of Birmingham Charter, Greenfield, and San Bernardino explained that families within their LEAs also fear involvement by a child protective services agency. According to Birmingham Charter’s local liaison, some families do not want to disclose that they are experiencing homelessness because of the associated stigmas, and Vallejo’s local liaison said that some families or youth may not be aware that their living situations meet the definition of homeless under the McKinney‑Vento Act.

Another contributing factor to underidentification of youth experiencing homelessness for four of the six LEAs is that the staff they have dedicated to administer the homeless education program generally spend limited time on the program. Although the six LEAs described that there could be various personnel, such as teachers, social workers, and enrollment staff, who assist in the identification of youth experiencing homelessness, these personnel do not have a role in administering the homeless education program. We found a strong correlation between the time that staff who administer the homeless education program dedicated to it and the number of youth the LEAs identified as experiencing homelessness, as Table 2 shows. For example, Gridley’s former local liaison estimated that she spent two hours per month on the homeless education program, and her LEA identified only one of its economically disadvantaged youth as experiencing homelessness. In contrast, San Bernardino’s local liaison estimated that she and her staff were spending 482 hours a month on its program and it identified 9 percent of its economically disadvantaged youth as experiencing homelessness. Similarly, according to an estimate by the local liaison for Norwalk‑La Mirada, an LEA with total enrollment that was one‑third that of San Bernardino’s, the LEA was dedicating 247 hours of staff time each month to its program, and it identified 29 percent of its economically disadvantaged youth as experiencing homelessness.

Table 2
LEAs That Reported Dedicating Less Time to Administering the Homeless Education Program Typically Identify Fewer Youth Experiencing Homelessness
LEA ECONOMICALLY DISADVANTAGED
YOUTH		 YOUTH IDENTIFIED AS EXPERIENCING HOMELESSNESS PERCENT OF ECONOMICALLY DISADVANTAGED YOUTH IDENTIFIED AS EXPERIENCING HOMELESSNESS ESTIMATED STAFF HOURS PER MONTH TO ADMINISTER THE HOMELESS EDUCATION PROGRAM* NUMBER OF STAFF ASSIGNED
Birmingham Charter 2,860 43 2% 52 1
Greenfield 9,912 78 1 15 1
Gridley 1,687 1 0 2 1
Norwalk‑La Mirada 14,316 4,174 29 247 2
San Bernardino 52,390 4,971 9 482 4
Vallejo 10,651 344 3 17 1

Source: CALPADS cumulative end‑of‑academic year 2017–18 data that Education does not publish on its website, LEA interviews, and documentation to support the number of staff and estimated time dedicated to their homeless education programs.

* The LEAs generally do not track staff time dedicated to the homeless education program. Therefore, these hours are estimates from the LEAs’ liaisons.


San Bernardino’s and Norwalk‑La Mirada’s higher rates of identifying youth experiencing homelessness suggest that other LEAs could identify more of these youth by dedicating more time and resources to their homeless education program. Only one of the six LEAs described that its resources for homeless education were too low, whereas the remaining five LEAs generally indicated that resources were reasonable. However, based on our observation, some LEAs have not made their homeless education programs a priority. When LEAs do not prioritize identifying all youth experiencing homelessness, they likely do not identify all of these youth, who then do not receive the support and services they need to have equal access to education and succeed in school.

None of the LEAs We Visited Has Followed All Federal Laws or Best Practices to Identify Youth Experiencing Homelessness

None of the LEAs we reviewed were sufficiently following federal law or best practices to ensure that they were effectively identifying youth experiencing homelessness. Toward this end, best practices recommend that LEAs use a housing questionnaire to identify these youth. Additionally, federal law requires local liaisons to ensure that school personnel who provide services to youth experiencing homelessness receive training. It also requires local liaisons to disseminate information on the educational rights of youth experiencing homelessness throughout the community. However, the six LEAs we reviewed did not adequately use housing questionnaires as best practices recommend, or they did not always train school personnel or disseminate information publicly as federal law requires. These LEAs generally agreed that these practices could help them identify more youth experiencing homelessness. However, until they follow best practices and implement these requirements, these LEAs will continue to miss opportunities to identify all such youth.

LEAs Have Not Effectively Used Housing Questionnaires

To assist local liaisons in meeting their responsibilities under the McKinney‑Vento Act and to identify youth who qualify for services, best practices recommend that LEAs use a housing questionnaire to gather information about the living situations of youth in the district. They also recommend providing the questionnaire to all students to prevent singling out those experiencing homelessness. Further, according to best practices and SchoolHouse Connection—a national nonprofit organization working to overcome homelessness through education—LEAs should gather housing information from families and youth at least once a year and preferably at multiple points throughout the academic year. Because the housing situation of a youth can change at any time, LEAs are much less likely to identify all youth experiencing homelessness if they do not ask families about their living situations at least once each academic year.

During the three academic years we reviewed, 2015–16 through 2017–18, four of the six LEAs we visited used housing questionnaires annually, and the different approaches they employed to distribute them to families likely affected their identification rates. According to Birmingham Charter, Norwalk‑La Mirada, San Bernardino, and Vallejo, they follow the best practice of providing housing questionnaires or inquiry forms annually to all families and youth. However, Greenfield provides the housing questionnaires to families and youth only when students first enroll. Similarly, Gridley, which just began using the questionnaire in academic year 2018–19, did not provide the housing questionnaire to all returning students already enrolled in its schools. The practices by Greenfield and Gridley are problematic because students may not be experiencing homelessness when they first enroll in school, and those LEAs do not reassess their living situations upon registration each year as the students progress to graduation.

By not providing all families and youth with housing questionnaires annually, Greenfield and Gridley have likely contributed to their identification of significantly fewer numbers of youth experiencing homelessness during academic year 2017–18 than expected. Specifically, Greenfield only identified 78 youth experiencing homelessness, or less than 1 percent of its nearly 10,000 economically disadvantaged youth, and Gridley identified just one of its approximately 1,700 economically disadvantaged youth as experiencing homelessness. The local liaison for Greenfield could not explain why it only provides the housing questionnaire to families when a student initially enrolls. Gridley’s local liaison explained that the LEA only began using the housing questionnaire in academic year 2018–19 because of regional fires that caused many families to lose their homes in surrounding areas. A number of these families then enrolled their school‑age children in Gridley. Before then, Gridley relied on referrals from homeless shelters or school personnel to raise concerns about whether a student might be experiencing homelessness. Both local liaisons agreed that administering the housing questionnaire annually for all students would be a good practice.

Two of the four LEAs that do administer the housing questionnaire each year require that families and youth complete and return the questionnaires as part of their annual enrollment process, which ensures that the LEAs receive the questionnaire for all youth attending their schools. Specifically, San Bernardino incorporated the housing questionnaire into its emergency card, which every family or youth must complete annually. Likewise, Norwalk‑La Mirada explained that it has historically included the questionnaire in its required enrollment forms. Although Norwalk‑La Mirada told us that all families or youth must complete this questionnaire, the superintendent acknowledged that the LEA did not have a mechanism to ensure that all families did so before academic year 2018–19. However, the local liaison stated that starting in academic year 2019–20, Norwalk‑La Mirada moved its enrollment process to an Internet‑based format, and the superintendent confirmed that this new format will not allow the families or youth to submit their annual enrollment forms without completing the questionnaire. In part to help with this new approach, Norwalk‑La Mirada stated that it has made computers with Internet access available at its school sites for families who need them. Both LEAs cited the housing questionnaire as their primary means of identifying youth experiencing homelessness, and Norwalk‑La Mirada has found that requiring each family or youth to complete it has not required additional resources.

Moreover, some of the LEAs’ housing questionnaires do not follow best practices and might discourage families and youth from properly completing them. According to best practices, to reduce the stigma involved with homelessness, LEAs should avoid using the word homeless in their questionnaires. However, as Table 3 shows, Greenfield’s housing questionnaire still uses the word; the form states that Greenfield will investigate claims of homelessness. In addition, best practices suggest that LEAs inform families and youth that the information being requested will be used to determine whether the student is eligible to receive additional support and services. However, only San Bernardino’s and Vallejo’s housing questionnaires explain that the LEA uses the information for this purpose. The questionnaires of the other four do not explain how they will use the information provided, which could deter families from disclosing their living situations if they fear a child protective services agency or immigration authority may investigate them. When LEAs do not include language in their housing questionnaires to mitigate stigmas associated with experiencing homelessness and do not explain the purpose of collecting critical information, this primary tool for identifying youth experiencing homelessness becomes less effective.

Table 3
LEAs Did Not Consistently Follow Best Practices to Identify Youth Experiencing Homelessness by Using a Housing Questionnaire
HOUSING QUESTIONNAIRE BEST PRACTICES BIRMINGHAM CHARTER GREENFIELD GRIDLEY NORWALK-LA MIRADA SAN
BERNARDINO		 VALLEJO
Uses an annual questionnaire to identify youth experiencing homelessness ✔ X X  ✔  ✔  ✔
Avoids using the word homeless in the questionnaire ✔ X *  ✔ ✔ ✔
Explains that the requested information will be used to determine if the youth is eligible for additional support X X * X ✔  ✔
Explains the rights and protections of youth experiencing homelessness X X * X X X

Source: Best practices, and LEA interviews and questionnaires.

* Gridley did not have or use a housing questionnaire during our audit period; therefore, we did not assess whether its current questionnaire follows best practices. Specifically, Gridley implemented a housing questionnaire in academic year 2018–19 for initial enrollments as a result of the Camp Fire in Butte County.


In addition, the six LEAs did not adequately explain on their housing questionnaires the educational rights and protections afforded to youth experiencing homelessness, even though doing so would likely encourage families and youth to complete it. For example, under the McKinney‑Vento Act, youth experiencing homelessness have the right to immediate enrollment and transportation to school and extracurricular activities. Moreover, state law specifies that a youth experiencing homelessness is not, in and of itself, a sufficient basis for school officials to report child abuse or neglect. Further, federal law generally prohibits LEAs from disclosing information on the youth’s living situation to any individual, agency, or organization without the parent’s or student’s consent. However, none of the LEAs’ housing questionnaires discussed these rights and protections, even though doing so might assuage the fears of some families and youth. According to some of the LEAs, they did not include information on rights and protections of youth experiencing homelessness because it is not required by law or they believed they did not have enough room on the questionnaire as they wanted to limit its length. Yet without such information, the youth and their families may not respond honestly, if at all, to the housing questionnaire.

Finally, Vallejo’s housing questionnaire may inappropriately prevent some youth experiencing homelessness from receiving support and services under the McKinney‑Vento Act. Specifically, the first question on Vallejo’s housing questionnaire asks families or youth whether their current residential address is a temporary living arrangement; however, a family living with relatives because of economic hardship might not consider its living situation temporary and might incorrectly respond no to the question on Vallejo’s questionnaire. In these instances, the questionnaire says to stop filling it out, so the family would not complete the portion of the questionnaire describing the youth’s living situation. If the family does not report on the youth’s living situation, the LEA will not know if the youth qualifies for services and support under the McKinney‑Vento Act. Best practices indicate that the housing questionnaire should not stop a family from completing it based on the response to whether the living situation is temporary or permanent. Vallejo’s superintendent said he was not involved in developing the questionnaire, and the LEA has not reviewed or changed it in years. By instructing families to stop completing the housing questionnaire if they do not consider their living situation temporary, Vallejo may have inappropriately prevented some youth from receiving needed services.

LEAs Have Not Adequately Trained School Staff Involved in Identifying Youth Experiencing Homelessness

None of the six LEAs have trained all school staff who are in a position to identify youth experiencing homelessness. The McKinney‑Vento Act requires local liaisons to ensure that school personnel who provide services to these youth receive training on their responsibilities under the act. Further, best practices recommend that LEAs provide training at least annually for all school staff. In fact, in its state plan Education explains that because youth experiencing homelessness can be identified through student and family relationships with school staff, LEAs will train all school staff on the proper identification and reporting procedures. As a result, we expected LEAs to have a process for ensuring that they are regularly training all school staff. Although some of the LEAs provided training to certain staff, as Table 4 shows, none of the six LEAs have been training all school staff.

Table 4
LEAs Have Not Trained All Staff Who Are in a Position to Identify Youth Experiencing Homelessness
LEA PRINCIPALS, VICE PRINCIPALS, ASSISTANT PRINCIPALS COUNSELORS, PSYCHOLOGISTS, SOCIAL WORKERS ENROLLMENT
STAFF		 TEACHERS SUPPORT
PERSONNEL*
Birmingham Charter  ✔ ✔ ✔ ✔ X
Greenfield ✔  X† ✔  X X
Gridley X X X X X
Norwalk‑La Mirada ✔ ✔ ✔  X X
San Bernardino X X X X X
Vallejo ✔ X X  ✔ X

Source: Best practices, LEA interviews, and training documentation.

* Includes bus drivers and cafeteria workers.

† Greenfield trains its social workers; however, it does not train its counselors or its psychologists.


One LEA was aware that it was not providing adequate training to staff who are involved in identifying youth experiencing homelessness. Specifically, in June 2017, the Solano County Grand Jury issued a report on the educational rights of Solano County’s children and youth who are experiencing homelessness, which included a review of Vallejo and other districts within the county. That report stated that because identifying youth experiencing homelessness is critical to ensuring the delivery of needed services, the strongest emphasis needs to be placed on that process; however, the report concluded that training by the LEAs was limited. Specifically, the report found that the LEAs are emphasizing training for school clerks and administrators, and that this training is often combined with other training. Further, the grand jury found that training for teachers—the staff members who have the most contact with children—is very limited. The grand jury recommended that the county's LEAs, including Vallejo, make it a high priority for teachers to receive intensive training in identifying and reporting students who may be experiencing homelessness.

However, in the two years since that report, Vallejo still has not corrected the problem. When we mentioned the grand jury recommendation, Vallejo’s liaison explained that the LEA coordinated with the Solano County Office of Education in August 2017 to provide a one‑time, mandatory training on the McKinney‑Vento Act to all school staff, including teachers. However, as of September 2019, Vallejo had not held another training on youth experiencing homelessness. Vallejo’s superintendent explained that the division within the LEA responsible for scheduling appropriate training no longer exists because of budget constraints. As a result, the LEA has not provided training on homelessness since 2017. Although he acknowledged that such trainings are important as they provide consistent information to all staff, as of September 2019, the LEA has not determined when it will schedule the next McKinney‑Vento training.

The remaining LEAs gave us a variety of reasons why they have not trained all staff. According to Norwalk‑La Mirada’s local liaison, the LEA relies on front office staff’s knowledge of the homeless education program, but the local liaison agreed that the LEA needs to provide training to all staff annually. Gridley’s superintendent stated that because the LEA historically has had very few identified youth experiencing homelessness, it has not needed to host trainings for all staff, but he agreed that it needs to provide training to all staff going forward. Greenfield acknowledged training as an area for improvement and stated that the LEA recently created a train‑the‑trainers program for its social workers and counselors to train all other school staff on the homeless education program. Birmingham Charter explained that it was not aware that it should train all school staff, but as a best practice, the local liaison has provided training annually to many staff. San Bernardino’s local liaison told us that all their schools have autonomy to create their training calendar for the academic year and it is up to the discretion of each principal at each school to determine topics for training. According to one principal in San Bernardino, the school is short‑staffed, so the staff are unlikely to attend a training unless the LEA requires them to do so. Considering that the law requires local liaisons to ensure that school staff receive training, we expected San Bernardino to require its schools to include training for school staff on their responsibilities under the McKinney‑Vento Act in the schools’ training calendars for the academic year.

School staff for the six LEAs were not always aware of the McKinney‑Vento Act, the rights and services afforded to youth experiencing homelessness, or what living conditions qualify under the act. We interviewed school staff at the six LEAs, including administrators, counselors, registrars, and teachers. When we interviewed two teachers at each of two schools in Greenfield, none had any knowledge of the McKinney‑Vento Act, nor did they understand the educational rights and services guaranteed to youth experiencing homelessness. Further, although the teachers we interviewed at some LEAs we visited had a general knowledge of the definition of homelessness, almost half shared that a specific hypothetical scenario of a youth’s living situation we presented to them would not have prompted them to notify the local liaison for further review. One of the two teachers we interviewed at Birmingham Charter, and three of the four teachers we interviewed at Greenfield, told us that they would not have considered notifying the liaison to request further review of the youth’s living situation even though the circumstances could potentially qualify the youth for homeless support and services.

Moreover, the quality of the training that LEAs provided varied. According to best practices, training should focus on the definition of homelessness, signs of homelessness, the impact of homelessness on youth, and the steps that staff should take once an LEA has identified a youth as possibly experiencing homelessness. Although Greenfield provides training to its administrators and enrollment staff, the training materials did not include information on all these topics, such as the signs and impact of homelessness. Similarly, the training that Birmingham Charter provided to counselors and registrars did not include the signs of homelessness. According to Greenfield, the missing best practices were an oversight, and Birmingham Charter explained that it was not aware of the best practices. Both LEAs acknowledged that they plan to include these details in their trainings going forward. Unless LEAs provide adequate training to all school staff that includes the information necessary for staff to identify the signs of homelessness, there is an increased likelihood that youth experiencing homelessness will not be identified and receive needed services and support.

Some LEAs Have Not Adequately Disseminated Information About Their Homeless Education Programs

Federal law requires and best practices stress that LEAs should undertake certain activities to increase awareness among families and communities regarding the educational rights of youth experiencing homelessness. Specifically, the McKinney‑Vento Act requires local liaisons to disseminate information about the educational rights of these youth in public places frequented by their families, including schools, shelters, public libraries, and food pantries. Further, the local liaison must display this information in an understandable manner, such as in posters or flyers at each site. State law also requires local liaisons to ensure that public notice of the educational rights of these youth is disseminated in schools within the LEA. Best practices also recommend that LEAs include the rights of these youth and their local liaisons’ contact information on their websites. Disseminating information on the homeless education program, including the educational rights of eligible youth, increases public awareness and the opportunities to identify youth in need.

Although not specified in federal law or best practices, we believe that LEAs should also describe the protections afforded to these youth in the information they disseminate. Specifically, five of the six LEAs we visited noted that families and youth may not disclose that they are experiencing homelessness because they fear that a child protective services agency or immigration authority could take their children. Yet in California homelessness is not, in and of itself, a sufficient basis for school officials to report child abuse or neglect. Informing families and youth about the protections afforded to them under the law could alleviate many of these concerns and might result in greater identification of youth in need of services.

However, only one of the six LEAs we visited—San Bernardino—disseminated information explaining the educational rights of youth experiencing homelessness in public places, as Table 5 shows. According to local liaisons for four of the five LEAs that have not been disseminating such information, they were unaware that the law required them to do so. The fifth LEA, Norwalk‑La Mirada, explained that at the outset of its homeless education program, it directly posted the educational rights of homeless youth in public places, but as the program has evolved, the LEA began to rely on its schools and other service providers to disseminate it. However, the LEAs generally indicated that it would be a good idea to start disseminating information about the educational rights of youth experiencing homelessness in public places as the law requires.

Moreover, although Education makes a poster available that LEAs could use at their district offices and school sites, only four of the six LEAs displayed posters or flyers at each school we visited and in their district offices, and neither Gridley nor Vallejo displayed such posters, as Table 5 shows. According to these two LEAs, they were unaware that they should be displaying posters at the district and schools. However, we expected these local liaisons to have familiarized themselves with the law and best practices and to have ensured that their LEAs were adhering to them. Although the remaining four LEAs used posters developed by Education, we noted that Education’s poster does not include information regarding the legal protections afforded to youth experiencing homelessness.

Table 5
The LEAs We Visited Inconsistently Follow Federal Law and Best Practices for Disseminating Information to Youth Experiencing Homelessness and Their Families
METHODS TO
DISSEMINATE INFORMATION		 BIRMINGHAM CHARTER GREENFIELD GRIDLEY NORWALK‑LA MIRADA SAN BERNARDINO VALLEJO
Required by Federal Law
Provides information about the educational rights of youth experiencing homelessness in public locations frequented by their families, which may include shelters, public libraries, and food pantries X X X X ✔ X
Displays information about the educational rights of youth experiencing homelessness in an understandable manner, such as in posters or flyers, at each school  ✔  ✔ X ✔ ✔ X
Best Practices
Includes the local liaison contact information in an easy‑to‑find location on its website  ✔ X X  ✔  ✔ X
Includes information about the educational rights of youth experiencing homelessness on its website X X X  ✔ X  ✔

Source: Federal law, best practices, LEA websites, interviews, and documentation to evidence dissemination of information.


Further, some LEAs have not clearly disclosed their local liaisons’ contact information or information on the educational rights of youth experiencing homelessness on their websites. Specifically, Greenfield, Gridley, and Vallejo listed the local liaisons by their formal position titles, such as director of student services or director of special education, rather than clearly disclosing that these individuals were the local liaisons. As a result, an individual visiting their websites would not know whom to contact regarding the LEAs' homeless education programs. Additionally, Birmingham Charter, Greenfield, Gridley, and San Bernardino did not include information about the educational rights of youth experiencing homelessness on their websites. Further, none of the LEAs’ websites included information about the legal protections afforded to those experiencing homelessness, and we believe that doing so may result in greater identification of these youth. When LEAs do not make their local liaisons’ contact information or critical information regarding the rights and protections of youth experiencing homelessness readily available, families and youth needing support may struggle to find the appropriate person to contact for assistance or may not be aware that they can enroll immediately regardless of their housing status and receive support services.

Recommendations

Legislature

To ensure that LEAs effectively identify and serve youth experiencing homelessness, the Legislature should require LEAs to follow best practices, as follows:

LEAs

To comply with federal law and best practices, Birmingham Charter, Greenfield, Gridley, Norwalk‑La Mirada, San Bernardino, and Vallejo should, before academic year 2020–21, do the following:

Education

To ensure that youth experiencing homelessness have access to the necessary services to help them succeed in school, by August 2020 Education should establish guidance for implementing data‑sharing agreements between the LEAs and other organizations that provide services to these youth.



Chapter 2

EDUCATION HAS NOT PROVIDED ADEQUATE OVERSIGHT OF THE LEAS’ HOMELESS EDUCATION PROGRAMS

Chapter Summary

Education has not adequately monitored LEAs’ policies and processes for identifying and supporting youth experiencing homelessness, which has contributed to the LEAs’ various shortcomings that we describe in Chapter 1. Specifically, of the nearly 2,300 LEAs in the State, Education has reviewed only about 20 LEAs every year to ensure that they are complying with requirements of the McKinney‑Vento Act. Although Education has said it lacks the resources to perform additional reviews of LEAs, it has not leveraged available data to help target its monitoring efforts. For example, it could use the data that LEAs report on the number of their youth experiencing homelessness and the number who are economically disadvantaged to determine whether LEAs are likely underidentifying those experiencing homelessness. Education told us that although it has not performed such analyses using available data, it plans to do so. Education also has not developed training modules for all LEA staff as its state plan requires, and it provides only limited in‑person training to LEAs. In the absence of such training for all LEAs, Education relies on county offices of education to provide more frequent training to LEAs and local liaisons. However, these resources do not always align with best practices or with the goals that Education set for itself in its state plan.

Education noted that it lacks the staff to effectively fulfill its duties related to the homeless education program, yet it has recognized this shortcoming for at least a decade and has not used existing resources to better support the program. Further, it has not conducted a staffing analysis that would establish whether it needs additional staff. Until it conducts such an analysis, it cannot effectively justify the need for additional resources.

Education Has Not Sufficiently Monitored the LEAs or Assessed Their Effectiveness in Identifying Homeless Youth

Federal law requires Education to monitor the activities of LEAs to ensure that they comply with requirements of the McKinney‑Vento Act. Best practices recommend that states monitor each LEA’s compliance regularly, with many states conducting on‑site monitoring of each LEA every three years and more frequently for LEAs that receive McKinney‑Vento grant funding. Each academic year, Education selects approximately 130 LEAs to monitor for compliance with requirements of certain programs, including homeless education. However, Education reviewed less than 1 percent of all LEAs’ homeless education programs during each academic year in our audit period. Specifically, of the nearly 2,300 LEAs in California, the state coordinator only reviewed between 12 and 21 LEAs for compliance with homeless education program requirements each year between academic years 2015–16 and 2017–18. The state coordinator indicated that time and resource constraints limit the number of LEAs Education selects. The state coordinator did tell us that it plans to review a larger number of LEAs each year starting with academic year 2020–21.

Considering the large number of LEAs in California and the severity of the State’s homelessness issue, as well as the number of LEAs in our review that were not complying with federal requirements, Education’s current effort to review less than 1 percent of LEAs for their compliance with homeless education program requirements is inadequate. As we discuss later, half of the LEAs we visited had outdated policies for homeless education and, as a result, did not always reflect key updates to federal law. Further, as we discuss in Chapter 1, four of the six LEAs we visited did not always coordinate with other organizations to identify youth experiencing homelessness and to provide them with needed services. Thus, the problems we found at LEAs highlight the need for Education’s homeless education program to monitor more LEAs to ensure that they comply with federal requirements and that they have effective homeless education programs.

For those LEAs the state coordinator does not review, it said it relies on staff who monitor the compensatory education program to include a review of some aspects of LEAs’ homeless education programs; however, we found that this review is limited. LEAs with high numbers of students from low‑income families receive federal funding to provide compensatory education services to help ensure that all students meet state academic standards. Federal law requires LEAs that receive these funds to reserve some portion as necessary for their homeless education programs. During academic year 2018–19, Education’s compensatory education staff planned to review nearly all of the approximately 130 LEAs selected for monitoring. However, this review was limited to ensuring that the LEAs described the services they provided to support the enrollment, attendance, and success of youth experiencing homelessness, in coordination with the services the LEA provides under the McKinney‑Vento Act. As a result, this review provides Education with minimal assurance about the adequacy of LEAs’ homeless education programs. Specifically, the review did not include any areas critical to an LEA’s homeless education program that the state coordinator reviews as part of its monitoring. For example, the state coordinator reviews training for local liaisons, coordination of services with external organizations, questionnaires and other registration forms, and referrals for health care services. Consequently, Education cannot rely on the reviews that its compensatory education staff perform to supplement the low number of LEAs that the state coordinator reviews.

Moreover, Education has an inadequate process for selecting the LEAs whose homeless education programs it plans to monitor. Until we pointed out its lack of policies for selecting the LEAs, the state coordinator had not documented its methodology for doing so. The protocols it finalized in September 2019 to select the LEAs to review explain that it examines statewide data that LEAs self‑report in the Consolidated Application and Reporting System (CARS)—a system through which LEAs report various information, including their use of funds to support youth experiencing homelessness and to provide assurances that they are complying with legal requirements of the program. For example, these protocols include determining whether an LEA has reserved sufficient federal funds for homeless education. Further, those protocols state that for LEAs that receive McKinney‑Vento Act grant funds, the state coordinator reviews the timeliness of the fiscal reports that they submit to it and their grant expenditures to identify potential risks. However, as we discuss later, Education has not established a method for detecting LEAs that may be underidentifying homeless youth, and as a result the state coordinator’s protocols do not include criteria to target such LEAs. The state coordinator could use performance outcomes—including graduation, suspension, and chronic absenteeism rates for youth experiencing homelessness—as additional criteria to help identify which LEAs to monitor. These data are readily available through the California School Dashboard, a web‑based system that Education is required by state law to develop and maintain. The dashboard displays the performance of LEAs, schools, and student subgroups, including youth experiencing homelessness, on various academic and other performance outcomes.

Additionally, Education’s other method for monitoring LEAs’ homeless education programs does not sufficiently ensure that these programs comply with laws and provide adequate support for youth experiencing homelessness. Education administers an annual collection of LEA‑reported data using CARS. The data collection asks LEAs to self‑report on topics such as their training of LEA staff and their policies related to their homeless education programs. However, because Education does not verify the accuracy of the responses submitted, the information in CARS may not always be accurate. In fact, we found instances of such inaccuracies when reviewing the six LEAs. Specifically, in 2017 five of the six LEAs we visited reported that their local liaisons had provided training on homeless education to teachers. However, as we discuss in Chapter 1, we found that three of these five LEAs had not actually trained teachers on the topic.

Further, Education has not always used all information that the LEAs provide through CARS to identify those LEAs that may not comply with requirements. For example, in 2017 all six of the LEAs we visited reported that their policies were several years old, and two reported that they had not updated their policies since 2003. However, when reviewing CARS data for all LEAs the state coordinator only ensured that the LEAs had a policy in place. The State coordinator did not review LEAs' responses regarding when they last updated these policies and did not follow up with any LEA that indicated that its policies might be out of date. As a result, some of the LEAs’ policies did not always reflect key updates to federal law related to the homeless education program. For example, Congress amended the McKinney‑Vento Act effective October 1, 2016, to make local liaisons responsible for ensuring that school personnel who provide services under the act receive training. However, only three of the six LEAs’ board policies included a requirement for local liaisons and other appropriate staff to participate in training. By not ensuring the accuracy of the responses, such as by reviewing a selection of LEA responses, and by not reviewing all responses that they provide, Education may inaccurately conclude that LEAs have complied with requirements.

Education Has Not Targeted Its Efforts to Assist Struggling LEAs by Maximizing Available Data

Education uses some data to monitor and provide assistance to improve LEAs’ homeless education programs; however, these efforts are limited. At the beginning and end of each academic year, Education requires LEAs to report the number of youth they have identified as experiencing homelessness through the California Longitudinal Pupil Achievement Data System (CALPADS)—a system composed of student demographic and enrollment data. Education uses these data to identify those LEAs that report having zero youth experiencing homelessness enrolled in their schools and to send them a letter offering technical assistance. However, the letter that Education disseminates is limited to describing federal reporting requirements, defining homelessness under the McKinney‑Vento Act, and asking the LEA to ensure that future counts of youth experiencing homelessness are accurate. Further, the state coordinator does not conduct any follow‑up to ensure that the LEAs take subsequent steps to improve their identification methods. Besides this letter, the state coordinator does not provide any other technical assistance or support to LEAs that report zero homeless students. As a result, Education is not effectively using these data to provide struggling LEAs with the guidance they need to improve their identification methods.

Education could better analyze available data to assess whether LEAs may be inadequately identifying youth experiencing homelessness. As Chapter 1 describes, Education has not developed a benchmark to assess whether LEAs are effectively identifying these youth; however, we found that other entities have developed such benchmarks. For example, Florida’s state coordinator analyzes data to identify LEAs that report less than 5 percent of their economically disadvantaged youth as experiencing homelessness, and it offers the LEAs technical assistance to improve their identification practices and rates. Additionally, the National Center for Homeless Education considers an LEA at high risk of underidentifying such youth if it reports less than 5 percent of its economically disadvantaged students, and less than 2.5 percent of its total student enrollment, as experiencing homelessness. If Education had used these data and applied the method that the National Center for Homeless Education uses, it would have found that in academic year 2017–18, two‑thirds of California’s nearly 2,300 LEAs may have underidentified these youth. By not performing a similar analysis, Education is missing the opportunity to identify a significant number of LEAs that may be struggling to identify these youth and to help the LEAs improve their homeless education programs.

Education collects a variety of data that it can use to inform its homeless education program’s approach to providing technical assistance and guidance. For example, as Table 6 shows, CALPADS contains the LEAs’ graduation, suspension, and chronic absenteeism rates for students identified as experiencing homelessness. The state coordinator could use these data to help gauge the effectiveness of the supports the LEAs are providing these students and provide targeted technical assistance as needed to underperforming LEAs. For example, the state coordinator would have identified that youth experiencing homelessness in Vallejo had worse performance outcomes than those statewide and could have provided guidance to help Vallejo improve student performance outcomes. In addition, the state coordinator could use the data that LEAs self‑report in CARS regarding staff training and turnover in local liaison positions to identify those LEAs that might need additional training. For instance, in academic year 2017–18, one‑third of local liaisons had only one year of experience in the role. Without conducting specific data analyses, Education cannot make informed decisions about which LEAs need technical assistance and what topics to include in the guidance it provides. 

Table 6
Education Collects Data That Could Inform the Guidance It Provides to Improve LEAs’ Homeless Education Programs
DATA SOURCE APPLICATION
LEAs with zero or very few identified students experiencing homelessness End‑of‑academic year, cumulative data that LEAs submit through CALPADS Determine whether an LEA is at risk of underidentifying youth experiencing homelessness.
Number of students that are economically disadvantaged or eligible for free or reduced‑price meals
Student enrollment
 
Graduation rates California School Dashboard using data that LEAs submit through CALPADS Assess the outcomes and success of LEAs’ homeless education programs.
Suspension rates
Chronic absenteeism rates
 
Liaison, principal, enrollment staff, teacher, and counselor training CARS Identify whether LEAs’ staff that are in a position to identify youth experiencing homelessness received training.
Date of approval of the LEA’s homeless education policy CARS Identify LEAs that have outdated homeless education policies.

Source: Data that Education collects from LEAs and best practices.


Education explained that it has not had the capacity to conduct such data analyses; however, it recently reassigned one staff member from a different department to the state coordinator. The new staff person will analyze CALPADS and CARS data related to LEAs’ homeless education programs. These efforts will include analyzing the percentage of economically disadvantaged students whom LEAs have identified as experiencing homelessness and identifying LEAs that may require further assistance.

Education Has Not Provided Adequate Guidance to Most LEAs

Although Education provides some guidance to LEAs by making resources available on its website and providing in‑person training to a limited number of LEAs, these resources and trainings are largely inadequate and do not always align with best practices. Education’s website includes sample documents to aid LEAs in identifying youth experiencing homelessness, including a housing questionnaire and training modules for certain school staff. However, these documents do not align with best practices and contain language that may discourage qualified youth from disclosing that they are experiencing homelessness and thus preventing them from receiving services under the McKinney‑Vento Act. Further, although Education provides annual in‑person training to a small number of LEAs that receive grant funds, it provides only infrequent training to a limited number of those LEAs that do not receive such funds. As a result, not all LEAs receive the training they need from Education to inform their efforts to effectively identify and support youth experiencing homelessness.

Education’s Guidance on Housing Questionnaires Does Not Incorporate Best Practices

Education’s online resources are outdated and insufficient, and they do not always fulfill the goals for training that Education established in its state plan. Education’s website includes a sample housing questionnaire that LEAs can reference when developing their own questionnaires. However, this sample housing questionnaire does not address or consider some key barriers that LEAs face when trying to identify youth who may be experiencing homelessness. Specifically, the housing questionnaire instructs families or youth to stop filling out the form if they “live in a fixed, regular, adequate nighttime residence.” Yet as we describe in the Introduction, the McKinney‑Vento Act’s definition of homelessness includes sharing housing with other people because of loss of housing, economic hardship, or a similar reason.

In fact, Education’s sample housing questionnaire could hinder an LEA’s ability to identify youth experiencing homelessness. The majority of the LEAs we visited indicated to us that many families and youth who live with other people for one of these reasons consider themselves to be living in a fixed, regular, adequate nighttime residence. If an LEA modeled its housing questionnaire after the one Education makes available on its website, families or youth who live with others may not identify themselves as experiencing homelessness and therefore may not receive services under the McKinney‑Vento Act even if they qualify.

Moreover, Education’s sample housing questionnaire does not sufficiently disclose the rights of such youth, including available key services. For example, under the McKinney‑Vento Act, youth experiencing homelessness have the right to immediate enrollment and to remain in their school of origin. Best practices recommend that to encourage families to complete a housing questionnaire, the forms should describe these rights, which may include some services, such as immediate enrollment in the school they last attended, transportation to school and extracurricular activities, and free meals. However, Education’s sample housing questionnaire does not include either the youth’s rights or these available services—information that could encourage a youth or family experiencing homelessness to complete the questionnaire.

The state coordinator explained that the sample documents it makes available on Education’s website, including the housing questionnaire, were developed by various LEAs. It also believed the existing information on the questionnaire was sufficient and did not want the questionnaire to exceed one page. However, we found a sample housing questionnaire from the National Center for Homeless Education that contained all relevant information on a single page. Further, the state coordinator believed that families of youth experiencing homelessness should be able to obtain additional information about rights and services from school personnel or posters at the schools. We question the reasonableness of this explanation, particularly since three of the six LEAs we reviewed did not ensure that enrollment staff received training on the homeless education program. Further, two of the six LEAs did not display posters in schools, and the remaining four LEAs displayed the posters that Education makes available to LEAs. As we discuss in Chapter 1, these posters could be more effective if they described the legal protections afforded to youth experiencing homelessness.

Education’s sample questionnaire also does not mention any of the protections afforded to youth experiencing homelessness. As we discuss in Chapter 1, many of the local liaisons we interviewed acknowledged that some families fear that if they identify as homeless, a child protective services agency may investigate them for child neglect or an immigration authority may investigate their residency status. However, state law specifies that a youth’s homelessness is not, in and of itself, a sufficient basis for school officials to report child abuse or neglect. Further, in the United States, all children are entitled to a free public elementary and secondary education regardless of their or their parents’ immigration status. Therefore, we believe that it is a best practice to state such protections in the housing questionnaire to encourage families and youth to identify themselves as experiencing homelessness. Without disclosing the rights and protections afforded to these youth, Education’s sample questionnaire is not as effective as it could be in assisting LEAs in identifying and supporting a larger number of these youth.

In addition, Education has not emphasized that LEAs should use the housing questionnaire annually with all students. Best practices recommend using a housing questionnaire to identify youth experiencing homelessness, and in our review of best practices, we found that of the five states we reviewed, the state coordinators for four—Florida, Georgia, Texas, and Washington—told us that they emphasize distributing these forms annually, and some emphasize distributing these forms more than once during the academic year. In fact, in 2014 Washington State implemented legislation that strongly encourages schools to use a variety of methods each year to notify students and families about services and support available to them if they experience homelessness, including distributing and collecting an annual housing questionnaire. Further, Education’s state coordinator believes that it is a best practice for LEAs to distribute the housing questionnaire to all students at the beginning of each academic year. Consequently, we expected Education to recommend to LEAs that they do so. However, the state coordinator could not provide any documentation of such a recommendation.

In fact, as we discuss later, the training modules that Education has made available on its website do not include any discussion of a housing questionnaire. This lack of guidance may have contributed to two of the six LEAs we visited not distributing the housing questionnaire annually. The state coordinator agreed that such a practice should be emphasized in the guidance it provides to LEAs and told us that it plans to incorporate this practice into the training modules. By not encouraging LEAs to distribute the housing questionnaire annually to all students, Education lacks assurance that LEAs are aware of this best practice to help them identify youth experiencing homelessness.

Education Has Not Developed Adequate Training for LEAs as Required

Education has not developed adequate training modules and posted them on its website for all stakeholders as its state plan stipulates. California’s state plan, which the U.S. ED approved in 2018, states that Education will develop training modules with stakeholder input on various homeless education topics for principals, teachers, local liaisons, health care providers, outside agencies, preschool staff, and enrollment staff; and it will post these modules online and disseminate them during the 2017–18 academic year. However, as of September 2019, Education had posted training modules only for enrollment staff, school counselors, and teachers. When we asked the state coordinator why it had not developed the remaining training modules, including those for principals and preschool staff, as Education confirmed it would do in its state plan, the state coordinator explained that it will complete the training modules as time permits and plans to do so by April 2020. The state coordinator also indicated that it had developed a fourth training module that was intended for local liaisons, but stated that this module had been removed from Education’s website in February 2019 for unknown reasons. Education did not post the module to the website again until late September 2019.

Further, none of the training modules mention the best practice of distributing a housing questionnaire to identify youth experiencing homelessness. This omission is concerning because best practices and the five states we reviewed generally identify the housing questionnaire as their most effective method for identifying these youth. The state coordinator agreed that the training modules for relevant stakeholders should emphasize the importance of distributing a housing questionnaire and told us that it will work to revise the modules to ensure that they include this information. By not making training modules available to all stakeholders as specified in its state plan and by not ensuring that these modules contain relevant best practices, Education is not ensuring that LEAs have access to adequate information to assist them in identifying youth experiencing homelessness. Although in September 2019 Education developed and posted on its website a fourth training module for local liaisons, this module also does not discuss or recommend distributing a housing questionnaire.

Further, these training modules are electronic slide presentations and are not interactive. By using interactive webinars and posting recordings of them on its website, Education can reach more LEAs with the limited resources it has available and maximize the effectiveness of its training. This alternate approach to in‑person interactive training would also allow greater participation by eliminating the time and costs associated with travel, and it ultimately could result in better information sharing among LEAs with similar issues.

We also found that Education provides limited in‑person training to a small number of LEAs, and it does so infrequently. Specifically, during academic year 2017–18, Education offered annual training to 95 LEAs, including 54 county offices of education that received McKinney‑Vento Act grant funds. These 95 LEAs represent only 4 percent of the LEAs in California. Between 2015 and 2018, the state coordinator also provided training to one LEA and 14 county offices of education; the participants mostly included local liaisons from LEAs within those counties, according to the state coordinator. However, it does not record attendance at these trainings and, therefore, could not identify which local liaisons attended those trainings. The state coordinator also explained that it has presented information related to the homeless education program at conferences and training for other LEA staff. The state coordinator’s infrequent training of local liaisons is of special concern considering the high turnover rate in these positions; according to LEA‑reported data in CARS for academic year 2017–18, more than half of local liaisons had two years or less of experience in the position. The state coordinator indicated that it relies on county offices of education to provide more frequent training to LEAs. In fact, in academic year 2017–18, Education provided funding to 54 of California’s 58 county offices of education to provide training and technical assistance to LEAs’ local liaisons. However, Education provided funding for this purpose to less than half, or 20, of the county offices of education in the preceding two academic years. Further, Education does not require county offices of education to report to the state coordinator on the trainings they provided to LEAs. As a result, the state coordinator lacks assurance that all liaisons received adequate and regular training.

Further, the training Education provides to local liaisons of LEAs that do not receive grant funds is not as comprehensive as the training for local liaisons of LEAs that receive grant funds. For example, Education’s 2017 training for local liaisons of grant‑receiving LEAs included a session on trauma‑informed practices for schools, which provided participants with an understanding of the trauma that youth experiencing homelessness face and its impact on their academics, behavior, and relationships. In contrast, the trainings that Education provided to local liaisons of LEAs that did not receive grant funds were generally briefer and typically did not include such trauma‑informed practices. Further, although Education presented slightly different topics at each training, not all of the trainings covered certain key information. For example, these trainings focused on various topics, such as state legislation related to the homeless education program and the use of Title I, Part A, funds for the homeless education program. However, not all trainings included topics such as the emotional and academic effects of homelessness, strategies for enrolling and supporting unaccompanied youth, and the importance of coordinating with community resources to increase services for youth experiencing homelessness. By not providing comprehensive training to all local liaisons, Education has missed an opportunity to ensure that local liaisons are aware of ways to train school staff to identify more youth experiencing homelessness, create a more accepting environment for these youth, and ensure that they receive the services they need. As we discuss in Chapter 1, the six LEAs we visited were unaware of certain requirements of the McKinney‑Vento Act and best practices, which in part may be the result of inadequate training from Education.

Although Education claims it lacks the resources to provide guidance and training to more LEAs, we found approaches to training activities in other states that could help Education leverage its available resources. For example, Texas and Georgia both stated that they conduct webinars and Georgia uploads to its department website materials from previous trainings for local liaisons to review. Further, best practices recommend using alternative methods of interactive training, such as webinars, to overcome the difficulties and cost associated with in‑person training and to efficiently and effectively provide training to as many participants as possible. However, Education has only conducted three webinars in the last three years, all of which were held in 2018, and only one of the webinars was scheduled to include 50 participants, whereas the remaining two webinars were scheduled to include 10 or fewer participants. Further, the state coordinator explained that it typically conducts webinars when an LEA or county office of education requests it. Education has also not posted recordings of these webinars on its website for other local liaisons to review.

Education Has Not Devoted Adequate Resources to Fulfilling Its Responsibilities

In addition to the activities we discuss earlier, the state coordinator also collaborates with other Education programs and with external organizations that are involved in working with, or improving services provided to, youth experiencing homelessness. However, until recently Education only had 2.5 positions to administer the homeless education program. Although it now has more staff, Education has not performed a staffing analysis to determine the number of staff it needs to fully meet all of its responsibilities.

In fact, at least 10 years ago, the federal government noted that Education’s staffing of its homeless education program was significantly inadequate. The U.S. ED periodically reviews states to assess the extent of the leadership and guidance they provide to LEAs for implementing policies and procedures that comply with federal requirements. As part of its 2010 review of compliance with the McKinney‑Vento Act, it evaluated Education’s guidance and technical assistance to all LEAs as well as its administration of California’s homeless education program. The U.S. ED found that Education had insufficient capacity to oversee the LEAs receiving grant funds as well as all other LEAs in the State. In making this conclusion, the U.S. ED explained that although Education is allowed to reserve up to 25 percent of its McKinney‑Vento Act allocation for state‑level activities supporting the implementation of the homeless education program in all LEAs, Education had reserved less than 2 percent of its allocation for this purpose. In fact, the U.S. ED stated in its 2010 monitoring report that this was the third time it had raised the concern about insufficient capacity to Education—meaning that Education has known that it has inadequately staffed the homeless education program for longer than 10 years.

Other states we reviewed employ more staff for the number of LEAs they oversee, as Table 7 shows. For example, as of July 2019, Georgia, which has about 200 LEAs, employed four staff and spent 17 percent of its McKinney‑Vento Act grant money on state‑level administration. Moreover, despite also having fewer LEAs than California, Texas has committed seven staff and spent 25 percent of its McKinney‑Vento Act funds on state‑level administration and providing technical assistance to LEAs. If California devoted more resources to state‑level administration of the program, it could provide greater oversight and guidance to its LEAs, which likely would enable them to better identify and support youth experiencing homelessness.

Table 7
California Has Devoted Fewer Staff Than Other States, Despite the Large Number of LEAs It Oversees

STATE		 PERCENT OF GRANT RESERVED FOR STATE‑LEVEL ADMINISTRATION STAFF LEAs LEAs
PER STAFF MEMBER
California 5.5%     2.5* 2,272 909
Florida 6.5 3.5 75 21
Georgia 17 4 202 51
Michigan 18 2 901 451
Texas 25 7 1,206 172
Washington 25 2 332 166

Source: Unpublished CALPADS cumulative end‑of‑academic year 2017–18 data for the number of California LEAs, data from the National Center for Homeless Education for the number of LEAs in the remaining states during academic year 2016–17, interviews with the respective state coordinators, and budget documentation from Education.

Note: Although the staffing levels are from the 2018–19 academic year, the number of LEAs, with the exception of California, is from academic year 2016–17, which is the most recent information available from the National Center for Homeless Education.

*   Education only recently added one additional staff member in July 2019, for a total of 3.5 staff.

Although Education has known about the limited resources for more than a decade, it has not made adding more resources a priority. Education has engaged in deliberations, some of which are confidential, over whether its staffing for the homeless education program is adequate. However, until recently, Education’s staffing for its homeless education program had remained unchanged. Further, Education has not fully considered using other existing resources to meet its staffing needs. In fact, Education did not reassign an additional staff member to the homeless education program from another division until after we began this audit. Education stated that it was able to make this change because it reorganized the division that oversees the homeless education program. Education also explained that the newly assigned staff member will analyze data and assist with other duties, including providing guidance and monitoring LEAs. Moreover, the state budget for fiscal year 2019–20 allocated an additional 1.5 full‑time‑equivalent positions to Education for the homeless education program. As of September 2019, Education was in the process of filling one of these positions. Education explained that it is using the remaining position authorization to make an existing part‑time office technician a full‑time staff member of the homeless education program. Once Education fills the new full‑time position, the state coordinator will have five full‑time staff.

The state coordinator asserted that Education still needs additional staff; however, we found that it has not clearly identified how many staff it needs to adequately meet its responsibilities under the homeless education program, including its obligation to monitor LEAs. Taking into account the U.S. ED’s concerns and Education’s acknowledgment of its inadequate staffing, we expected to find that Education had conducted a staffing analysis to determine how many additional staff it needs. A staffing analysis would allow Education to evaluate its responsibilities, establish whether it can meet those responsibilities with existing resources, and determine the number of additional staff it would need to meet all of its responsibilities. If Education had conducted a staffing analysis, it would have been able to determine staffing needs and better justify any request for additional resources, yet it has not done so.

Recommendations

Legislature

To ensure that Education provides effective oversight for the education of youth experiencing homelessness, the Legislature should require Education to do the following:

Education

To ensure that it has the resources necessary to effectively meet its responsibilities under federal law, Education should complete a staffing analysis by May 2020 to determine the resources needed to meet its responsibilities for homeless education. This analysis should consider the resources needed to implement all of the recommendations in this report. If Education determines that it needs additional resources, it should take the necessary steps, including reallocating existing resources within the department, to secure the needed resources.

To effectively monitor LEAs and help them identify additional youth experiencing homelessness, Education should do the following:

To ensure that all LEAs receive necessary guidance and training, Education should perform the following:

We conducted this audit under the authority vested in the California State Auditor by Government Code 8543 et seq. and according to generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives specified in the Scope and Methodology section of the report. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives.
Respectfully submitted,

ELAINE M. HOWLE, CPA
California State Auditor

November 7, 2019





Back to top