SRI International,
December 11, 1997

EVALUATION OF
CHARTER SCHOOL EFFECTIVENESS
Part I

Prepared for:
Mr. Joel Schwartz, Project Director
Evaluation of Charter School Effectiveness
Office of the Legislative Analyst
925 L Street, Suite 1000
Sacramento, CA 95814
Submitted by:
Judith Powell
Jose Blackorby
Julie Marsh
Kara Finnegan
Lee Anderson

Acknowledgments

We wish to acknowledge the contributions of several individuals who provided valuable guidance and support for this study. Special thanks to Joel Schwartz at the Legislative Analyst's Office for his leadership and thoughtful feedback, and the members of the Charter Study Advisory Panel for their expert review. We would also like to thank Mark Kushner of Leadership High School, Elise Darwish of San Carlos Learning Center, and Henry Hurley of San Diego Unified School District for piloting and reviewing data collection instruments.

EXECUTIVE SUMMARY

I. STUDY OVERVIEW AND DESIGN

Introduction
General Approach
Study Design
State databases
Charter school survey
District and county surveys
Case studies
State-level interviews
Challenges
Organization of the Report

II. CHARTER SCHOOL CHARACTERISTICS

General Characteristics of California Charter Schools
Parent Involvement in Charter Schools
Student Characteristics
Instructional Staff Characteristics
Conclusion

III. CHARTER SCHOOL GOVERNANCE, FINANCE,
AND ACCOUNTABILITY

The Development of Charter Agreements
Motivations for charter school petitions
Charter-granting process
Revisions to charter proposals
Charter Opposition
The School-Sponsor Relationship
Charter school independence
Staffing and day-to-day operations
Gaining more independence
Charter school governance
Collegiality and conflict
Charter School Finances
Accountability
Advantages of Charter Status
District- and county-sponsored schools
Charter district schools
Conclusion

IV. TEACHING AND LEARNING IN CHARTER SCHOOLS

Professional Roles and Opportunities
The role of the teacher
Professional development and support
Teacher evaluation
Instructional Programs
Home-based learning and independent study programs
Standards and curriculum frameworks
Subject matter emphasis
Instructional delivery methods
Classroom practices
Conclusion

V. OUTCOMES OF CALIFORNIA CHARTER SCHOOLS

Progress Toward Charter Goals
Parental Involvement and Perspectives
Student Achievement
Why aren't data available?
Philosophical and practical issues
Assessment Approaches for Charter-Noncharter Comparisons
Statewide assessment
Portfolios and demonstrations
Standardized tests
Teacher-assigned grades
Behavioral indicators
Illustrative Examples of Student Achievement at Case Study Schools
National statistics
District comparisons
Comparable-school comparisons
Within-school comparisons
Conclusion

VI. CONCLUSIONS AND RECOMMENDATIONS

Summary of Findings
Recommendations
Recommendations for further research

REFERENCES

APPENDICES

A: Profile of Charter School Survey Respondents
B: Survey Instruments
C: Interview Study Protocols
D: Supplemental Within-District Comparison Data
E: Charter School Law
F: Charter Study Advisory Panel

Executive Summary

I. Study Overview and Design

This report contains the findings of the Evaluation of Charter School Effectiveness, an interim study commissioned by the Legislative Analyst's Office of the State of California in 1997. The study team at SRI International collected and analyzed data on the range of organizational arrangements and instructional activities implemented in charter schools, and sought to assess the impacts of charter schools on teaching and learning. Specifically, we addressed the following research questions:

What is the educational performance of charter schools when compared with noncharter public schools?

What are the characteristics of charter schools on a range of factors, including student body, educational programs and philosophy, finance, governance, contractual arrangements with teachers, independence from sponsoring district, etc.?

What practices or features of charter school educational programs, finance, governance, etc., are associated with particular educational outcomes for charter school students?

What practices of sponsoring districts, county offices of education, and/or the state department of education help or hinder charter schools in implementing their educational program as they intended?

Multiple data sources were employed in this evaluation: state databases; a telephone survey to all charter schools with charters granted before April 1, 1997;⁽¹⁾ a mail survey to all district and county sponsors of charter schools; site visits to 12 charter schools; and interviews with state-level administrators, policy-makers, policy analysts, and stakeholders. Data were collected from June to November 1997.

This executive summary highlights the study findings and recommendations but is not a substitute for the full report.

II. Charter School Characteristics

Differences between Charter and Noncharter Schools. Charter schools had many distinct characteristics, several of which set them apart from most noncharter public schools in California. These differences are discussed in detail below. The basic differences included enrolling students who lived outside of sponsoring district boundaries, serving nontraditional grade-level groupings, and having, on average, smaller enrollments than noncharter public schools in the state.

Similarities between Charter and Noncharter Schools. Charter schools were similar to noncharter public schools in several key respects. For example, charter schools were located in all parts of the state in all types of communities. Charter schools provided about the same number of calendar days of instruction as all public schools in the state. Charter schools served all grade levels, and, like noncharter schools, a higher proportion served lower grade levels.

Possible Admission Criteria. Most charter schools identified several factors that they considered as possible admission criteria in their schools. A space limitation or enrollment cap was often cited by charter schools as a factor that could prevent new students from being admitted to their particular schools. Less often, schools cited that an indication that a student and/or parent was not committed to a school's philosophy was a possible reason for not admitting students (it is possible that some parents may choose not to enroll their children after learning about a school's philosophy). Prior academic performance was rarely taken into account when admitting students.

Parent Involvement. A high level of parent involvement was a defining feature of charter schools. A majority of charter schools required parents to commit to certain levels and types of parent participation, and most schools reported that parent participation was high. Three-quarters of charter schools surveyed required that a parent or adult sign a contract with the school when enrolling a child, and 40% required parents to participate in a minimum number of hours at school. Some schools faced community opposition in trying to institute parental involvement requirements.

Student Characteristics. Overall, charter schools enrolled students with characteristics similar to the state average. For example, charter schools had an ethnic composition similar to that of all public schools in the state. Charter schools served percentages of low-income, special education, and limited English proficient (LEP) students similar to the state averages. Although charter schools served percentages of special education students comparable to those served by the state, special education was an area that was marked by uncertainty for many charter schools and sponsoring agencies.

Instructional Staff Characteristics. Charter school instructional staff were similar to teachers statewide in several ways, although charter school teachers were more likely to be part-time. For example, overall, charter school teachers had an ethnic composition similar to that of teachers statewide. Starting teacher salaries in charter schools were consistent with the state average ($27,200 in charter schools vs. $25,500 in all public schools statewide).

Staff Certification. On average, 71% of charter school instructional staff (full-time and part-time) had credentials. Full-time employees were more likely than part-time employees to have full state certification for the subjects they taught (83% vs. 49%).

Union Membership. Nearly two-thirds of charter schools reported that some or all instructional staff were members of unions or other bargaining units.

Start-up vs. Conversion Schools.(2) In general, start-up charter schools tended to be farther from the district or state norm on several dimensions than were conversion schools. Compared with conversion schools, start-up charter schools were smaller, more often enrolled students from a wide geographic area, and served more nontraditional grade ranges (e.g., grades 5-8). They were more likely to have instructional staff without union representation, and to have fewer traditional contract provisions for instructional staff (e.g., tenure). Start-ups were also less likely to serve special education and LEP students than were conversion schools. At the same time, however, they reported serving more low-achieving students than conversion schools.

Home-Based Study. The 15 survey respondents that identified home-based study with the parent as primary instructor as the predominant--though not necessarily only--method of instruction for most students had a distinct profile.⁽³⁾ These schools were most often located in small towns or rural areas, and were somewhat more likely to require parent contracts, more likely to enroll White students, and less likely to enroll LEP and special education students. The number of schools in this category and the total number of students they served were so small that their impact on overall charter statistics was negligible. However, because they represent a distinct variation on charter schools--one, moreover, that was not envisioned by the authors of the legislation--it is important to point out where they differed from charter schools as a whole.

III. Charter School Governance, Finance, and Accountability

Motivations for Charter School Petitions. There was a wide variety of motivations for establishing charter schools in California. In the case study schools, those motivations included a desire for innovation, alternatives to existing programs, a particular curricular focus, or more parent participation in the education program. Most often among the case study sample, charter schools typically were initiated by some combination of administrators, teachers, and parents.

Charter Sponsorship and the Charter-Granting Process. Most charter schools in California reported district sponsorship (87%). Seven charter schools (7%) reported county sponsorship, and six schools (6%), which were located in the two existing charter districts, reported sponsorship by the California Department of Education. Although most schools were sponsored by districts, the relationships schools developed with their sponsors and the process schools underwent to become charter schools varied tremendously.

Case study schools faced a range of responses to their charter petitions. In some cases, the sponsor and/or teachers' unions raised concerns about attracting students with diverse backgrounds and/or special needs, student assessment, teacher certification, and class size. The process was described by charter school staff as running the range from thorough and fair to horrendous. Almost one-third of charter schools surveyed were required to revise their original charters. One-fourth were required to enter into written side agreements in order to get approval. Most schools thought these changes either helped or had no effect on their ability to implement their educational programs as intended. This is contrary to the assumption held by some charter school observers that any changes to petitions that were requested by the sponsoring agency would have a detrimental effect. Our school survey data indicate that this was not the case.

Charter Opposition. Opposition to original charter petitions took various forms. In the case study sample, the goals and tactics of those who wanted to open charter schools were often seen (and sometimes intended) as open challenges to established interests within public school bureaucracies. In many cases, opponents to the charter were placated by changes to the petition prior to its approval. In other instances, opposition has continued or new opponents have surfaced during implementation of the charter. Despite the visibility of state and local union objections to the charter law and their misgivings about certain charter school practices, collective bargaining units took varying positions on charter schools in California. In general, local circumstances tended to determine local union opposition to charter petitions.

The School-Sponsor Relationship. Relationships between charter schools and their sponsors were as varied as all other aspects of charter schools. They ranged from full dependence on the sponsoring agency to nearly complete independence. Some schools, in fact, were dependent to the point that the advantages of their charter status were obscured.

Most charter schools did not or could not take advantage of all aspects of their "automatic waiver" from state and district regulations because their sponsoring districts or counties did not allow it, or because they made concessions to the union or the sponsor during charter negotiations. A smaller portion of charter schools did not associate charter status with certain freedoms because they were permitted by the district before the school had a charter, e.g., flexibility with the daily calendar or student assessment policy.

Many of the case study schools that had dependent relationships with their sponsors were satisfied with these close ties; many schools did not want to take on the responsibility of managing school finances. There were exceptions to these schools' satisfaction with the status quo, however. On the survey, more than half of the charter schools that did not have full control reported wanting more control over purchasing and staff hiring, discipline, and dismissal.

Charter School Finances. In most charter schools, finances were at least partially controlled by the sponsor. According to the school survey, only 27% of schools had financial autonomy (i.e., they had full control over staff salaries and benefits and other budgetary expenses). Start-up charter schools were more likely to report financial autonomy than were conversions (85% vs. 15%).

The most typical financial arrangement among the case study schools (most of which were financially dependent) was one in which schools were essentially funded as they would be if they were not charters. The financially independent schools in our sample had more variable financial arrangements with their sponsors. We found that the financial knowledge base of the school leader made a profound difference in the level of resources that an independent charter school received; school leaders with financial expertise were able to increase their schools' share of sponsor funds.

School survey data showed that unfamiliarity with the financial side of schools was widespread among charter school directors. For example, 24% of the charter school directors surveyed did not know whether they were eligible for Title 1. Similarly, only a small proportion of school directors were able to report funds received in budgetary categories other than the state revenue limit.

Financial Viability. According to officials in several sponsoring agencies, charter schools, as funded by law, do not have the financial resources to operate truly independently. Without help from the sponsor or an outside agency, the schools in the case study sample found that paying for facilities from revenue limit funds was a real challenge.

Liability. Concerns about their own liability kept some sponsors from loosening their ties to charter schools. Several of our case study sponsors reported that their belief that they would be held liable for charter schools' financial or educational failure discouraged them from giving charter schools more freedom. In at least three districts in our case sample, this point of view was shaped by past negative experiences with charter schools. Sponsor liability remains a large gray area in charter legislation. Recommendations issued recently by the California Department of Education to the State Board of Education regarding the financial operation of charter schools are only "a starting point for discussion and debate on the precise nature of changes that need to be made in statute and regulation in order to address these difficult issues." Until clarification is made by the Legislature or the courts, the liability issue remains ambiguous.

Accountability. For the most part, charter sponsors do not seem to hold charter schools to a higher standard of accountability for students' academic performance than they do noncharter schools. Although the schools we examined had measurable (to varying degrees) academic goals stated in their charters, most reported that they were not held accountable for achieving these goals by their sponsoring district or county. In contrast to the lack of accountability to sponsors for academic outcomes, most charter schools reported feeling accountable to parents in this regard.

District and county sponsors were much more diligent about financial accountability than academic accountability. Sponsor and, in some cases, school staff were determined not to let their charter schools become headline stories of fiscal mismanagement.

Advantages of Charter Status. Charter schools reported that their charters provided them with unique opportunities in several areas, ranging from personnel to finance. When school survey respondents were asked specifically what charter status allowed them to do that they could not have done under the traditional district management structure, charter school directors most frequently reported that they were able to allocate resources in a manner different from the district norm (87%).

Respondents in four case study sites had difficulty identifying just what they gained from being a charter. In three cases, it was likely that the schools would have been able to implement their educational programs without being charters, given the reform orientation of their respective districts and the history of reform efforts within the schools prior to becoming charter entities. Charter status for some of these schools, however, insulated them--at least theoretically--from district policy changes that resulted from shifts in the political climate.

IV. Teaching and Learning in Charter Schools

Different Stages of Change. It may be too early to accurately assess teaching practices within charter schools. Charter schools were at different stages of their experimentation and change processes. Some charter schools sought charter status to experiment with curriculum design, instructional programs, and professional development opportunities and succeeded in realizing this purpose. Others had been reforming these areas for years prior to becoming a charter school and used charter status to expand or continue these practices. Other charter schools articulated visions of new teaching techniques and curriculum in their charter petitions, but struggled to bring them to life within their classrooms. A few schools adhered to more traditional approaches to teaching, but argued that they had gained greater freedom to implement their educational program as they defined it.

The Role of Teachers. The role of teachers in our case study sites varied, with most teachers taking on additional administrative or curricular responsibilities in their schools. At most of the schools we visited, teachers served on the school's governing body.

In some cases, teachers described positive effects of taking on more responsibility. They noted that they were more satisfied with these schools because they had more control over their own classrooms. In other cases, these role changes may have produced negative effects, such as increased hours and work, and a lack of clarity and confusion over exact roles and responsibilities.

Professional Development and Support. Charter schools, for the most part, devoted a similar amount of time to professional development as other schools in California. The majority of case study sites reported providing more appropriate professional development opportunities because these were determined on a schoolwide or individualized basis rather than being designed by the district.

Teacher Evaluation. Nearly all charter schools surveyed (89%) reported having annual performance evaluations for teachers. However, in our case study schools, teacher evaluation systems had not yet been fully implemented. Although school directors reported that they evaluated teachers, most seemed to do so informally.

Standards and Curriculum Frameworks. In synch with the local, state, and national push toward standards-based reform, most case study charter schools recognized the importance of standards and high expectations for their students. Many of our case study schools reported that they followed the California Curriculum Frameworks as guides for curriculum and instruction. Several schools mentioned the use of district or county standards and curriculum guides. A small group of charter schools developed--or were in the process of developing--their own standards for some or all disciplines. A few schools articulated "grade-level expectancies." Many schools had developed or started to develop these standards documents before gaining charter status.

Instructional Delivery Methods. The majority (87%) of charter schools responding to the survey reported using the more traditional, classroom-based approaches; 29% of charter schools also employed home-based learning with the parent as primary instructor, 22% used independent study with the teacher as primary instructor, 20% used work- and/or community-based learning, and 11% used distance learning. Almost half (48%) of charter schools reported using more than one instructional delivery method.

Comparisons among charter schools reveal interesting patterns of instructional delivery. Start-up charter schools were more likely than conversion schools to use home-based learning as a primary method. Similarly, rural schools implemented home-based learning more frequently than schools in other locations. Large schools (enrollment of 600+) were most likely to have implemented distance learning as a primary instructional delivery method.

Classroom Practices. In our observations of classrooms, we found the implementation of classroom practices supported by research on effective teaching for meaning and understanding (e.g., thematic and interdisciplinary instruction, team teaching, multi-age grouping, and technology use) to be uneven across--and even within--some case study schools. At some schools, these practices appeared to be consistently implemented schoolwide. At other schools, they seemed to be more isolated within a select number of classrooms. At still other schools, we heard talk of such practices but did not observe them in action.

V. Charter School Outcomes

Data Challenges. When looking at three outcome areas of charter schools, the available data do not allow us to draw definitive conclusions about charter schools' performance, in part because it is too early in the reform process and in part because the available data are insufficient. However, we are able to address the usefulness of each outcome area in considering the success of California's charter school experiment. These three outcome areas are: (1) progress toward charter goals, (2) parental involvement and perspectives, and (3) student achievement.

Progress toward Charter Goals. Assessing schools' progress toward meeting objectives specified in their charters is useful in that it takes into account the entire range of school goals, provides an opportunity for schools to show interim progress, and has local relevance. The limited data in this area indicated that schools approached this task of measuring progress toward goals quite differently, situating themselves along a continuum from concrete and quantitative to informal and process oriented. Several case study schools took a systematic approach, collecting and analyzing data. Others took a less formal approach to documentation. In all cases, schools found that their progress was mixed.

Parental Involvement and Perspectives. The parent perspective in assessing charter schools is important since parents choose to send their children to charters. In a sense, parents "vote with their feet," so their decision to stay at a school can be interpreted as a sign of satisfaction. Available data on parents' opinions of charter schools indicated that they have achieved considerable success. Most schools reported having wait lists and plans for expansion in the coming year.

Student Achievement. Measuring student achievement is made difficult by the diversity of assessment practices, philosophies, and available data among charter schools. Overall, schools surveyed considered the use of both traditional teacher-assigned grades and alternative assessments, such as portfolios and demonstrations, to be essential or important to their assessment of student progress. Standardized tests and local performance-based tests were also used, but were less frequently considered to be essential or important.

Illustrative Examples of Student Achievement. Illustrative examples of student achievement at case study schools yielded mixed results. (Note that the following statements are intended to illustrate the complexities associated with various approaches to measuring achievement and are based on a very limited number of schools.) Some charter schools performed better than noncharter schools when compared with national averages, with all noncharter schools in their sponsoring districts, and with comparable noncharter schools within their districts. Others did not. Likewise, comparisons of charter schools' performance over time (within-school comparisons) yielded mixed results. It is important to realize that each of the above approaches to assessing progress has limitations. For example, comparisons with national averages can be misleading if the charter school population is very different from the national average on academically important issues, such as ethnicity, socioeconomic status, and disability.

VI. Recommendations

SRI makes the following recommendations based on data collected in this study. Most of these recommendations directly address issues that stem from ambiguities in the charter legislation. They concern serving students with special needs, the ethnic diversity of students, and liability. We highlight them here because they may have implications for reauthorization of the law.

Address how charter schools must comply with special education regulations.
Resolve the contradiction between ethnic balance in charter schools and neighborhood preference.
Clarify legal and fiscal liability issues by including a definitive assignment of charter school responsibility in the Education Code.
Provide more technical assistance to charter schools and sponsoring agencies on monitoring pupil learning, providing services to special populations, charter school finance and budgeting, and (for start-up schools) covering facilities expenses.

Recommendations for Further Research

Future research should focus on fewer key research and policy questions.
The Legislature should revise the legally mandated timetable for the next charter school study to allow for a longer interval for data collection, analysis, and reporting.

¹ Analyses in this report are based on the 98 responding charter schools that were operational at the time of the survey.

² Start-up charter schools are newly created schools; that is, they were not previously connected to an existing school. Conversion schools are schools that converted from an existing public school, county or district program, or selected grade levels within a preexisting school or program. Survey respondents for this study consisted of 49 start-up and 49 conversion charter schools.

³ Thirteen other schools identified home-based instruction as one of several instructional methods used, but indicated that most students were not taught this way.

I. STUDY OVERVIEW AND DESIGN

Introduction

The California Legislature enacted the Charter Schools Act of 1992 to permit teachers, parents, students, and community members to establish schools that would be free from most state and district regulations (Senate Bill 1448). California was the second state to put this type of law into place�Minnesota enacted charter school legislation in 1991. Since California and Minnesota ventured into this charter school arena, many other states have followed their leads. As of September 1997, 29 states and the District of Columbia have existing charter school laws, though not all have schools in operation (Center for Education Reform, 1997).

Charter school laws in California and other states are based on the proposition that allowing schools autonomy from existing bureaucratic and regulatory constraints will enable them to adopt innovative and productive methods of operation, finance, and governance to support the improvement of teaching and learning.⁽³⁾ In addition, proponents argue that charter schools will increase the number of educational options within the public school system. Most importantly, charter schools represent a shift from rule-based accountability to performance-based accountability. In exchange for freedom from bureaucratic and regulatory constraints, charter schools agree to be held accountable for student performance. As such, they offer California one model for fundamental reform of the state's system of schooling.

Charter school status is granted to schools once they have successfully completed the petition process with their district or county and, ultimately, the state. On approval, the California Department of Education provides each charter school with a number, in order of receipt, to ensure that the total number of charter schools within the state does not exceed the legislative cap of 100. Although the law states a cap of 100, the State Board of Education has been waiving the cap on a case-by-case basis since 1996.⁽⁴⁾ Table I-1 illustrates the number of charters the state has granted each year since the California Legislature enacted this law.

Table I-1
NUMBER OF CHARTER SCHOOLS APPROVED EACH YEAR

Year Approved⁽⁵⁾ Number of Schools

1993 42

1994 28

1995 36

1996 15

1997 3

TOTAL 124

General Approach

The Legislative Analyst's Office of the State of California (LAO) contracted with SRI International to conduct this interim Evaluation of Charter School Effectivness. This study was required by law to provide the Legislature with information prior to a more in-depth evaluation, which is legally mandated to be completed in January 1999. The purpose of SRI's 7-month interim study, conducted from May through December 1997, was to report preliminary findings to the Legislative Analyst's Office relating to the effectiveness of the charter school approach in California. This report will describe the range of organizational arrangements and instructional activities implemented in charter schools and seek to assess the impacts of charters on teaching and learning. It addresses a series of questions regarding charter schools that call for both a description of their chief characteristics and a sophisticated analysis of relationships among charter schools, their sponsoring agencies, their instructional programs, and student outcomes. Specifically, the report addresses the following research questions:

What is the educational performance of charter schools when compared with non-charter public schools?

What are the characteristics of charter schools on a range of factors, including student body, educational programs and philosophy, finance, governance, contractual arrangements with teachers, independence from sponsoring district, etc.?

What practices or features of charter school educational programs, finance, governance, etc., are associated with particular educational outcomes for charter school students?

What practices of sponsoring districts, county offices of education, and/or the state department of education help or hinder charter schools in implementing their educational program as they intended?

To address these questions, we built the evaluation design on the research and analysis of previous reform efforts. Although charter school policies call for a radical shift from traditional rule-based practice, they are actually based on a series of ideas that have been evolving in educational reform for quite some time. Over the past 25 years, policy-makers, practitioners, and researchers have struggled to craft new approaches to the challenges facing American education, including the effective schools movement in the 1970s (Edmonds, 1975; Purkey & Smith, 1983), school-based decision-making and school restructuring in the 1980s (David & Shields, 1991; Elmore, 1990; Newmann, 1991), and parental choice and the provision of greater autonomy in return for increased accountability in the 1990s (Chubb & Moe, 1990). Running through this stream of reforms is a common, if evolving, set of ideas:

The school is the key unit in which to promote improvements in teaching and learning for all students.

Effective schools are organized not by past practice or external regulations, but by the goal of improving teaching and learning.

The appropriate approach to reforming a school will vary across schools.

Effective schools involve teachers, parents, and administrators in critical discussion and decision-making.

Schools must be held accountable for student performance.

Charter schools are meant to possess these characteristics. They receive greater autonomy in return for greater accountability for student performance; they are designed to provide parents, teachers, and administrators with greater flexibility and authority to design programs that make sense for their children; and they are urged to build structures and schedules driven by student learning goals.

Study Design

This evaluation builds on the lessons we have learned from previous reform efforts. These lessons include the complexity and dynamism of the process of change, the significant amount of time required for fundamental shifts in a school and its classrooms, the overemphasis on process and underemphasis on student learning in many reform efforts, and the importance of understanding changes in student achievement in the context of an understanding of school- and classroom-level changes. On the basis of these lessons, our study employed multiple methods and triangulation of data collection from a variety of sources (described in the next part of this section) to address complex research questions. In addition, the evaluation minimized the burden on project participants by (1) concentrating data collection on high-priority evaluation questions; (2) having clear plans for how data would be analyzed and reported, thus eliminating the collection of extraneous data; and (3) maximizing the use of existing data sources to avoid unnecessary duplication of effort.

The study was designed to provide descriptive and analytical information regarding the status of charter schools in the state based on data gathered from five critical sources: (1) state databases, (2) a telephone survey to all charter schools, (3) a mail survey to all district and county sponsors of charter schools, (4) site visits to charter schools, and (5) state-level interviews. This report is based on a synthesis of these study elements. This multifaceted approach allowed the team to compile and analyze quantitative data, including student and teacher demographic data, as well as qualitative data, such as interviews of school staff and observations of classroom activities, in a short period of time. The following is a brief description of each data source.

State databases

To obtain basic descriptive details about charter schools, the study team used the California Department of Education's Web page and the Department's Charter Schools Office as resources. In addition, the study team accessed several statewide databases, including the California Department of Education's California Basic Educational Data System (CBEDS), the Language Census, and the High School Performance Report, to compile data relating to school and student characteristics, as well as student performance data. We also downloaded data for noncharter schools, districts, and the state to allow us to make appropriate comparisons. Achievement data and behavioral indicators were compiled only for those schools that had been in place for at least 2 1/2 years.

Charter school survey

The SRI project team developed and piloted the school survey of the California Charter School Study in May 1997 (a copy of the Charter School Survey is located in Appendix B). The LAO and the LAO's Charter Study Advisory Panel reviewed the survey instrument and provided feedback to the team, which was incorporated into the final instrument. Letters of support from the LAO and the California Network of Educational Charters (CANEC) accompanied the survey to announce the evaluation and ask for charter school support. SRI staff members with prior interviewing experience conducted the phone surveys from June 3 through July 15, 1997. These staff members participated in training sessions led by the study team to prepare them for interviewing charter schools. Considering the complex topic area and detailed survey instrument, the training sessions proved to be critical to gathering reliable and valid data.

Of the 124 eligible charter schools statewide, 111 charter schools responded to the statewide charter school telephone survey (90% response rate).⁽⁶⁾ The participating schools represented 30 counties and 80 districts across the state. Respondents were primarily school directors or coordinators but may have been other school staff, such as teachers who had leadership roles. The telephone surveys were approximately 45 minutes long, and included acquisition of student and teacher data, such as enrollment numbers and staff certification. These data were compiled by the respondent in advance, using a worksheet sent prior to the scheduled survey to lessen the amount of time required on the telephone. The survey was primarily closed-ended.

The descriptive information in this report relating to charter school characteristics, student demographics, teacher characteristics, finances, autonomy, instructional programs, assessment systems, and accountability was provided by the 98 respondents whose schools were in operation during the 1996-97 school year.⁽⁷⁾ Unless otherwise specified, all the survey data refer to the 1996-97 school year.

District and county surveys

The study team developed two charter school sponsor surveys; one version was sent to 71 public school districts, and the other version was sent to 7 county offices of education (copies of these surveys are located in Appendix B). These districts and counties represented the entire population of charter school sponsors for the schools receiving approval as of April 1, 1997.⁽⁸⁾ A first survey mailing occurred in August 1997, with a follow-up mailing in September. Phone calls were made to nonrespondents during October. Only 45 sponsor surveys (39 districts and 6 counties) were returned before the October 15, 1997, deadline. This represents a 58% response rate.

The district and county surveys were similar, with minor differences that enabled the team to ask more appropriate questions of these sponsor agencies. The sponsor surveys were designed both to triangulate charter school survey responses and to acquire district-level data for comparison purposes. The surveys consisted of questions relating to the district's or county's student and teacher characteristics, charter school policies, and school accountability and student assessment data (for both charter and noncharter schools). In addition, sponsors were asked open-ended questions about the impact of the charter schools in their district or county and any concerns they had about these schools. The survey included a school supplement for every school sponsored by the agency. The supplement focused on sponsorship relationships with the schools, including questions relating to school autonomy, the charter-granting process, contractual agreements with teachers, services provided by the sponsor, liability coverage, and charter renewal.

Case studies

The case studies were included in this study design to allow us to describe in greater detail the features of charter schools uncovered in the surveys. Case studies also allowed for an examination of certain issues that could not be adequately addressed in the survey, such as instructional practices. Finally, cases allowed for an exploration of the reasons underlying broader patterns emerging from the data.

A sample of 11 charter schools was randomly selected from the population of 50 charter schools that had been in operation since September 1994.⁽⁹⁾ We limited this sample selection to those schools that had been in operation for 2 1/2 years or longer to ensure that schools had been given the opportunity to implement their educational programs. One school declined to participate and was replaced by a school that was chosen to increase variation in geographic location. In addition, the team included in the case study sample the charter district that had been in operation since September 1994. This allowed the team to study in greater depth the issues faced by schools in districts sponsored by the California Department of Education. This initial sample was analyzed to ensure variation in the location, socioeconomic status, and financial independence of these charter schools. The sample was adjusted slightly to increase the geographic distribution of the sites. In addition, we performed statistical testing was to ensure that the ethnicity of the sample was representative of that of the entire pool of charter schools that had been in operation since September 1994.

The site visits were scheduled with school staff after an initial letter to schools and a phone conversation about the study. Team members conducted site visits during September and October in one- or two-person teams. The visits required approximately 1 to 2 days at the school site and a half day at the sponsoring agency. Staff from the Legislative Analyst's Office accompanied SRI staff on two visits, with the permission of the sites.

During case study visits, team members met with teachers, parents, and school administrators at the charter schools. When possible, the study team also conducted classroom observations. At the sponsoring agency, team members interviewed board members, union officials, business staff, superintendents, and other appropriate district administrators. Team members collected policy handbooks, student assessment data, and school accountability report cards during these visits. Table I-2 presents a list of case study schools. Echo Valley and Blue Sky are pseudonyms for schools that did not want to be identified.

Table I-2
CASE STUDY SCHOOLS

School	Grades	Enrollment 1996-97	Start-up/ Conversion	Location	School Type⁽¹⁰⁾	Sponsor Type
Charter Community School & Extended Day Program	K-12	720⁽¹¹⁾	Conversion	Northern	C, H	County
Charter Oak School	5-8	49	Start-up	Central	C	District
Discovery School	K-6	820	Conversion	Southern	C	District
Eagle Summit Academy	6-12	725	Start-up	Southern	C, I	District
Fenton Avenue Charter	PreK-5	1,281	Conversion	Southern	C	District
Garfield Charter School	PreK-6	600	Conversion	Northern	C	District
Echo Valley Charter School	K-12	1,500	Start-up	Northern	C, H, I	District
International Studies Academy	9-12	520	Conversion	Northern	C	District
Blue Sky Charter School	K-5	615	Conversion	Southern	C	District
Nevada City School of the Arts	3-8	150	Start-up	Northern	C	District
Pioneer Union Elementary School District	K-8	1,200	Conversion	Central	C	State
Westwood Charter School	K-5	720	Conversion	Southern	C	District

State-level interviews

The study team conducted semistructured interviews with a range of state administrators, policy-makers, and policy analysts with responsibility for or knowledge of charter schools in California. Respondents included administrators in the California Department of Education and the Department of Finance, consultants to the education committees of the State Legislature, and representatives of nongovernmental bodies, such as a state teacher's union and a research institute. Specifically, the interviews addressed questions regarding the state policies that help or hinder charter schools, the impact of charter schools on the overall public school system, and changes in state policies or the charter school legislation that might improve the effectiveness of these schools.

Challenges

Charter schools in California vary greatly in terms of school size, demographics, educational purpose, and other key school characteristics. Their charters range from very detailed descriptions of educational programs, financial arrangements, and governance to more generic outlines of schools. As one would assume, these variations in charter petitions have resulted in even greater variations in charter school implementation.

In employing our multimethod design, we faced two formidable challenges, resulting from the relative newness and great variability across charter schools, and a lack of comprehensive and cohesive data on noncharter schools: (1) how to assess charters' impact on student performance, and (2) how to compare charters with noncharters, both in terms of basic descriptive data and in terms of student progress.

Tracking the impact of charter schools on student performance is, to say the least, difficult: most charters are relatively new; by definition, charters are quite varied in their approaches and educational goals; a variety of assessment instruments are used; and data are not always easily obtained. To address this challenge, we excluded from our analyses any school that has not had a charter for at least 2 1/2 years, reducing the sample to approximately 68 schools. We then examined four different data sources to determine student progress: state databases, the school telephone survey, the district/county mail survey, and case study data. Chapter VI of this report provides a detailed analysis of these student performance data.

Comparisons between charters and noncharters present other problems, most notably the lack of comparable data and the difficulty of finding appropriate comparisons for unconventional charter schools. Throughout this report, we make district- or countywide comparisons, as well as state comparisons, using data from the surveys and state databases. Although useful, these comparisons fail to take the context of each charter school into account. To address this challenge, we have used case study data to better illustrate the uniqueness of each charter school.

Organization of the Report

The next chapter of this report (Chapter II) describes in detail the characteristics of charter schools across the state to answer the question: how different are they from regular public schools? The chapter describes school and student characteristics, including school enrollment, grade levels served, student and staff ethnicity, and admission requirements. Chapter III describes motivations for petitioning to become a charter school, and the charter school granting and renewal processes. Also described in this chapter are ongoing relationships between charter schools and their sponsors, including discussions of autonomy, financial arrangements, and accountability. Chapter IV describes teaching and learning in charter schools�the types of curricular and organizational arrangements and the professional opportunities of teachers. Chapter V provides a discussion of the complexities of assessing charter school outcomes, and presents an analysis of available outcome-related data. Finally, Chapter VI summarizes the evaluation findings and presents policy and research recommendations.

II. CHARTER SCHOOL CHARACTERISTICS

This chapter presents a descriptive overview of charter schools in California, including such basic information as the number and types of charter schools, their locations, their admission policies, the students they serve, and the teachers they employ. Overall, charter schools had characteristics that were distinctly different from those of regular public schools. Taken as a whole, charter schools in California were smaller than regular public schools. Charter schools had criteria for admissions that frequently included parents' agreement with the school philosophy and commitment to involvement in the school program, but rarely took into account student achievement level. On the whole, charter schools served a population that was demographically similar to the student population statewide. Within-district comparisons, however, showed that in about 40% of charter schools students were more likely to be White, and in about 60% of charter schools students were less likely to be low-income than other students in their sponsoring districts.

Beyond these generalizations, each charter school was a unique educational experiment. Although the schools needed to be categorized for purposes of analysis, there was almost always a charter school that was the exception to the rule. With that fact in mind, this chapter describes the major characteristics of charter schools, offering comparisons to regular public schools when possible. Where relevant, this chapter highlights the differences between subsets of charter schools, most often between newly started charter schools and existing public schools that converted to charter school status.

Throughout the chapter, we draw on school survey, district and county survey, and case study data. These are supplemented with data from state databases, such as CBEDS. School survey data are based on 98 respondents, and all survey data refer to the 1996-97 school year unless otherwise noted.

General Characteristics of California Charter Schools

Charter schools were located in all parts of the state in all types of communities, with a few geographic areas of concentration.

California charter schools were fairly evenly distributed across the state geographically, with concentrated numbers in Los Angeles, San Bernardino, San Diego, and Nevada Counties. On the basis of self-reports, charter schools in California were located in small towns or communities(33%), urban fringe or suburban areas (28%), urban or central city areas (19%), and rural areas (13%). Seven percent of the schools reported either that they were located in more than one of these locations or that they did not believe that any of these categories were appropriate.

A Closer Look⁽¹²⁾
The 15 survey respondents that identified home-based study with the parent as primary instructor as the predominant (though not necessarily only) method of instruction for most students⁽¹³⁾ were concentrated in small towns or communities (47%) and rural areas(33%). A few reported that they were located in urban or central city areas (13%), and 7% reported either that they were located in more than one of these locations or that they did not believe that any of these categories were appropriate.

Half of California charter schools were newly created, and half were converted from existing public schools.

The California charter school law allows charter developers to convert a public school to a charter school or to create a new school, generally referred to as a "start-up." In some cases, charter developers have converted programs run by either a district or a county, as well as grade levels within a public school. The types of charter school reported by respondents fell into equal proportions: 50% were newly created and 50% were converted.⁽¹⁴⁾ The schools that reportedly converted to charter schools included 39 (40%) that were originally public schools and converted entirely to charter schools, 2 (2%) that were public school programs or grade levels (the original school may still be in operation as a noncharter entity), and 8 (8%) that were originally district or county programs.

Many charter schools enrolled students who lived outside of sponsoring district boundaries.

In many cases, district-sponsored charter schools enrolled students who lived outside the boundaries of the sponsoring district. In fact, approximately 34% of the district-sponsored schools drew 25% or more of their students from other districts�usually from more than four other districts but not statewide. In addition, 13% of these schools reported thatmost or all of their students (76-100%) lived outside the sponsoring district.

The proportion of students who lived within sponsoring-district boundaries was closely linked to whether the school was a conversion or start-up school. Approximately 93% of conversion schools reported that most or all of their students (76-100%) resided within district boundaries, compared with 39% of start-up schools. This pattern is consistent with the state law, which says that charter school admission should not be based on the place of residence of the pupil, although conversion schools must give preference to individuals who reside within the former attendance area of the school (Education Code Section 47605 [d]).

On average, charter schools provided about the same number of calendar days of instruction as all public schools in the state.

Charter schools, on average, provided 183 calendar days of instruction. In comparison, the average public school had approximately 175 instructional days.⁽¹⁵⁾ However, charter schools reported a much wider range of instructional days (163 to 250) than noncharter schools (144 to 187). From our case studies, we learned that many charter schools have developed unique daily and weekly schedules. Therefore, these days may have a different meaning from instructional days in regular public schools. A higher proportion of charter schools than regular public schools operated on a year-round or modified year-round calendar�one-third of the schools, compared with only 17% of public schools in the state.⁽¹⁶⁾

A Closer Look
Less traditional calendars and schedules were most commonly found in schools reporting financial autonomy,⁽¹⁷⁾ schools in urban settings, and large schools. Approximately 50% of financially autonomous charter schools operated on year-round calendars, compared with 23% of financially nonautonomous charter schools. Similarly, both geographic location and size seemed to correlate with whether or not the school operated on a year-round calendar: urban schools and large schools were the most likely to implement this type of schedule. Forty-seven percent of urban schools utilized a year-round calendar, compared with 30% of suburban, 22% of small-town, and 8% of rural schools. In addition, 44% of large schools operated year-round, compared with 31% of medium/large schools, 29% of medium/small, and 13% of small schools.⁽¹⁸⁾

Charter schools served all grade levels, often in nontraditional grade-level groupings.

Charter schools served all grade levels from pre-Kindergarten through 12th grade. Like regular public schools, more charter schools served lower grades than upper grades. In contrast to regular public schools, half of charter schools served middle grades, compared with 24% of noncharter schools. Figure II-1⁽¹⁹⁾ illustrates the differences in the percentage of charter schools serving each grade level.⁽²⁰⁾

Consistent with national patterns, California charter schools often did not fit the traditional grade-level groupings--K-6, 6/7-8, and 9-12 (U.S. Department of Education, 1997). Charter schools, at times, served only portions of the grade-level groupings that are used in typical public schools; for example, a school may have served only 2nd through 6th grades. More frequently, however, charter schools served grades spanning two or more traditional grade-level groupings. For example, rather than serving only the standard elementary, middle, or high school levels, some schools served kindergarten through 12th grade. Figure II-2 shows the percentage of charter schools that served each grade-level combination, compared with the percentage of schools in the state.⁽²¹⁾

Start-up schools were more likely than conversion schools to have nontraditional grade-level groupings. As shown in Figure II-3, start-up schools were more likely than conversion schools to serve grades K-8 (middle/high) (27% vs. 13%), and more likely to serve grades K-12 (elementary-high) (15% vs. 6%). Conversion schools were more likely than start-ups to serve grades K-6 (elementary) (52% vs. 23%).

On average, charter school enrollment was smaller than the average state enrollment.

California charter schools came in a wide range of sizes. Enrollment varied from 5 students to 2,406 students, with an average enrollment of 434 students. This is much smaller than the average state enrollment of 767 (this figure includes charters and noncharters).⁽²²⁾ Figure II-4 shows differences between charter school and statewide school enrollment.⁽²³⁾ One upshot of the relatively small size of most charter schools and their vast range in student enrollments is that unweighted school-level data can yield different results when compared with student-level data. That is, if one averages data at the school level, a school with an enrollment of 50 carries the same weight as one with 1,000 students. As an example, the average LEP enrollment of the smaller school with 10% LEP students and the larger school with 50% LEP students would be 30%. If one took into account the number of students being served by each school, the average LEP enrollment would be much higher--48%. Although both school- and student-level analyses have their place, we have chosen, where possible, to report student-level data when describing key demographic characteristics (e.g., ethnicity, LEP, special education).

Start-up schools tended to be smaller than conversion schools. Start-up schools had an average enrollment of approximately 244 students, compared with approximately 620 students for conversion schools. So, although start-ups were 50% of all charter schools in the state, they reached a much smaller proportion of the students. Altogether, start-ups enrolled 28% of all students in charter schools.

A Closer Look
The 15 survey respondents who identified home-based study with the parent as primary instructor as the predominant method of instruction for most students varied widely in size. They served from 5 to 1,756 students, with a mean enrollment of 309 students. In some cases, the larger schools were arranged as umbrella groups that provided support for large numbers of previously unaffiliated home schoolers.

Charter schools identified several factors that they consider as possible admission criteria in their schools.

Charter school developers were required to include information about admission requirements in their proposals only if they thought it was "applicable." Therefore, many schools did not describe the process in their charters but have had to develop and implement these admission criteria once faced with high demand for limited space. In fact, nearly two-thirds (63%) of the schools reported having more applicants than they could serve in 1996-97.

Although charter schools had not necessarily refused to admit students on the following grounds, when asked "Which of the following factors can prevent new students from being admitted to your school?" they identified a number of factors, most often a space limitation or enrollment cap (78%). Such caps could either be self-imposed, contained in a provision of the charter, or dictated by the size of facilities. Student and/or parent not committed to school's philosophy was cited by 44% of charter schools. As worded on the survey, this option could have been interpreted to include those parents and students who themselves selected not to attend a particular school because of philosophical differences. The factors that could prevent students from being admitted to charter schools are listed in Table II-1.

Table II-1
FACTORS THAT COULD PREVENT STUDENT ADMISSION

Percentage of Schools Number of Respondents

Space limitation or enrollment cap 78% 98

Student and/or parent not committed to school's philosophy 44% 97

Evidence that parent or adult cannot fulfill involvement requirements 32% 97

Student's special needs because the school does not provide services such as special education or primary language instruction 30% 93

Residency outside of school or district boundaries 29% 97

Student ethnicity; in other words, the school considers ethnicity in order to achieve racial diversity 17% 98

Student's prior academic performance 3% 97

Prior academic performance was rarely taken into account when admitting students, contrary to what many feared would be the case. On the other hand, 30% of charter schools reported that being unable to meet student's special needs could be a reason not to admit a student (see page II-15 for further discussion of special education students). Likewise, the percentage of schools that cited evidence that parents cannot fulfill involvement requirements (32%) as a potential reason for not admitting a student indicates that some charter schools may have been exercising their right to be selective in their admission practices.

Twenty percent of charter schools reported that other factors could prevent students from being admitted to their schools. These factors included transportation (school does not provide), lack of student volition/motivation, lack of computer access (distance-learning program), student not identified as chemical dependence or at-risk for dependence (charter serves only this population), or expulsion from the sponsoring district or another district.

As illustrated in Table II-2, start-up schools were more likely than conversion schools to have admission requirements in several areas, most strikingly in relation to admission of students with special needs. On the other hand, only 12% of start-ups said they might take into account residency outside of school or district boundaries, compared with 46% of conversion schools. As mentioned earlier, this is consistent with charter law requirements that conversion schools must give preference to neighborhood students.

Table II-2
FACTORS THAT COULD PREVENT STUDENT ADMISSION:
START-UP VS. CONVERSION SCHOOLS

Percentage of Start-up Schools
Number of Respondents Percentage of Conversion Schools
Number of Respondents

Space limitation or enrollment cap 76% 49 80% 49

Student and/or parent not committed to school's philosophy 50% 48 39% 49

Evidence that parent or adult cannot fulfill involvement requirements 37% 49 27% 48

Student's special needs because the school does not provide services such as special education or primary language instruction 44% 45 17% 48

Residency outside of school or district boundaries 12% 49 46% 48

Student ethnicity; in other words, the school considers ethnicity in order to achieve racial diversity 16% 49 18% 49

Student's prior academic performance 6% 48 0% 49

A Closer Look
A comparison of school locations uncovered differences in admission policies and practices. Urban charter schools were more likely to limit admission to those students living within school or district boundaries--44% of urban charter schools reported this as a possible admission factor, as opposed to 30% of suburban, 25% of small-town, and 23% of rural charter schools. Charter schools in small towns and rural locations were more likely to enforce parent requirements for admission--62% of rural charter schools and 41% of small-town charter schools reported that evidence that parent/adult cannot fulfill requirements was a factor that could prevent a student from being admitted, whereas only 21% of urban and 22% of suburban schools reported this as a factor.

Parent Involvement in Charter Schools

Charter schools reported high levels of parent participation.

A high level of parent participation was a defining feature of charter schools. The vast majority of charter schools (88%) included parents on their governing bodies (see Chapter III for a discussion of charter school governance). In addition, charter schools reported that most parents participated in parent-teacher conferences (88%) and ongoing monitoring of homework (76%). Charter schools also reported that, on average, 40% of parents assumed instructional roles.

In most of the case study schools, there was a strong parent volunteer presence. Parents were busy assisting teachers in the classroom, photocopying classroom materials, organizing field trips, washing school uniforms, staffing the library, and working in many other capacities. Parents were also involved in large-scale fundraising in several schools, formal evaluation of the school's instructional program in at least one case, and the investigation of fiscal autonomy in another. Parents also had teaching roles in charter schools, especially those with home study and/or independent study programs.

Several schools reached out to parents by providing training for them. For example, one conversion school invited parents to virtually all on-site professional development events. Another stood out because of the comprehensiveness of its training opportunities for parents and community members. This school had a smaller parent education component prior to attaining charter status, but it has expanded it since then. Parents and other members of the community (usually relatives of students) could enroll in literacy, English, computer, and parenting classes, most of which had daytime and evening sessions to choose from. Child care was provided on-site in conjunction with these classes. Families were also invited on many of the school's field trips and weekend outings.

Although statewide data on parent involvement in noncharter schools are not available, there is anecdotal evidence that parent involvement is higher in some charter schools than in noncharter schools. In focus group interviews, parents frequently made favorable comparisons between their experiences at the charter school and other schools--public and private--where they felt less included. In most cases, they felt more welcome in the charter schools and felt that they had a voice in school decision-making.

A high number of charter schools required parents to commit to certain levels and types of parent participation.

According to the school survey, most charter schools emphasized the importance of parent involvement in school activities and promoted participation with either voluntary or mandatory requirements. Perhaps the most interesting innovation in these schools was mandatory parent contracts. Three-quarters (75%) of charter schools required that a parent or adult sign a contractwith the school when enrolling a child. In the case study schools, contracts typically covered parents' acceptance of school rules and parent involvement requirements, if there were any. Some charter schools also required parents or adults to participate on committees/governance boards or attend parent meetings (41%) or participate in a minimum number of hours at school (40%). Many schools did not have consequences if the parent or adult failed to fulfill these requirements; however, 23% of schools with parent involvement requirements reportedly had asked students to leave because of parents' failure to comply with these rules.

Start-up charter schools were more likely than conversion schools to require parent contracts�86% of start-ups, compared with 64% of conversions. Likewise, start-ups were more likely to require parents to participate a minimum number of hours at the school�46% of start-ups, compared with 34% of conversions. When the 15 home-based study schools are excluded from the analysis (home-based study schools make up 24% of the start-up schools), these comparisons are essentially unchanged: 87% of start-ups vs. 62% conversions required parent contracts, and 54% of start-ups vs. 32% of conversions required a minimum number of volunteer hours. One might expect that home school charters would be more likely to have parent contracts and requirements for volunteering, but that does not appear to be the case. Perhaps these schools do not consider parent involvement to be volunteer activity.

The parent involvement requirements in contracts were often quite precise. One case study school, for example, required that a parent volunteer at the school for 3 hours per week; another called for 4 hours per year. Two other schools each called for 5 hours per month. Two of these schools actually enforced provisions of the parent contract, leading, in one case, to a lawsuit against the school and, in the other, to a request that several families leave the school. Other schools have struggled with the question of how binding to make their contracts with parents.

In two case study schools, the parent contract idea was dropped in the face of community opposition. Some educators and advocates objected to mandatory parent involvement on the grounds that working, low-income, or single parents (many of whom were expected to reside outside the schools' historical attendance areas) might not be able to meet the terms of the contracts. Respondents in other charter schools argued that mandatory requirements were not discriminatory if they were flexible and included options for how a parent or guardian could fulfill requirements.

A Closer Look
Almost all small schools (96%) required parent contracts, and more than half of these schools (59%) required parents to participate in their schools for a minimum number of hours. Large and medium/large schools were less likely than medium/small and small schools to require parent contracts or minimum hours. Similarly, urban charter schools were less likely than schools in other locations to implement parent involvement contracts--58% of urban charter schools required parents to sign a contract with the school, compared with 85% of rural, 84% of small-town, and 74% of suburban charter schools.
Charter schools in rural locations were more likely than others to ask a student to leave when a parent or guardian did not fulfill the participation requirements--55% of rural charter schools reported taking these measures, as opposed to 24% of small-town, 17% of urban, and 14% of suburban charter schools.
The contrast between rural and urban schools is particularly striking regarding reports of parent involvement. Compared with urban schools, rural schools reported much higher levels of parent participation in instructional roles (53% rural vs. 22% urban), monitoring of student homework (93% rural vs. 68% urban), and parent-teacher conferences (97% rural vs. 83% urban). Part of this contrast may be accounted for by the 15 charter schools that are primarily home-based (i.e., have high parent involvement by definition); these schools made up 42% of all rural schools.

Student Characteristics

Overall, charter schools had a similar ethnic composition to all public schools in the state.

Demographically, students in California charter schools were similar to students throughout the state. In total, approximately 48% of charter school students were White, followed in order by Hispanic (34%), Black (9%), Asian (5%), Filipino (3%), American Indian (2%), and Pacific Islander (0.5%). As illustrated in Table II-3, in all racial/ethnic categories, the difference between students in charter schools and students statewide was less than 9 percentage points.

Table II-3
RACIAL/ETHNIC STUDENT DEMOGRAPHICS⁽²⁴⁾

Race/Ethnicity Percentage of Students⁽²⁵⁾

Charter SchoolsTotal students: 37,436 (94 schools) Statewide Total students: 5,612,965

White, not of Hispanic origin 48% 40%

Black, not of Hispanic origin 9% 9%

Hispanic 34% 40%

Filipino 3% 2%

Asian 5% 8%

Pacific Islander .5% .6%

American Indian/Alaska Native 2% .9%

A Closer Look
The ethnic composition of the schools that identified home-based study with the parent as primary instructor as the predominant method of instruction for most students was strikingly different from that of other charter schools. The vast majority of students at home-based charter schools were White (81%), compared with slightly fewer than half for all other charter schools (45%).⁽²⁶⁾ Twelve percent of students in home-based schools were Hispanic, compared with 36% in other charter schools.

Comparisons of individual charter schools with their sponsoring districts⁽²⁷⁾ provide a more valid description of school demographics in context. Such within-district comparisons of ethnic composition revealed a pattern similar to that found in the state comparison. Overall, most charter schools mirrored the ethnic ratios of their districts' noncharter public schools. The majority of charter schools deviated within only 10 percentage points from noncharter schools in their sponsoring districts with regard to the percentage of American Indian, Asian, Pacific Islander, and African-American students (see Appendix D for a full presentation of ethnic representation data).

The within-district comparisons for Hispanic and White students revealed slightly greater variation, however. Although approximately 50% of schools deviated within only 10 percentage points of the district figure, in 19% of schools White students exceeded the district percentage by more than 25 points (see Figure II-5). In 18% of schools, Hispanic students were less than the district percentage by more than 25 points (see Figure II-6). On average, the percentage of White students in charter schools was 9.2 percentage points more than the percentage for noncharter schools in their sponsoring districts, and the percentage of Hispanic students in charter schools was 8.6 percentage points less. It is likely that at least part of this discrepancy reflects patterns of residential segregation, as opposed to any efforts on the part of charter schools to be exclusionary. For example, conversion schools that are located in ethnically homogeneous communities will almost certainly lack diversity, given that they must give preference to neighborhood students. Another possible explanation is that the relatively small size of most charters may not allow them to offer the breadth of programs that would appeal to students with diverse interests and backgrounds.

Using free/reduced-price lunch program eligibility as an indicator, charter schools served levels of low-income students similar to the state average.

Approximately 43% of all students enrolled in the charter schools were eligible for the National School Lunch Program (providing free or reduced-price school lunches to students, based on the income level of their parents), compared with 47% statewide--a difference of only 4 percentage points.⁽²⁸⁾ (It should be noted that 12 out of the 73 schools that responded to this survey item said that 0% of their students were eligible; it is possible that some of these schools did not participate in the lunch program. Likewise, it is possible that some of the 25 charter schools that did not respond to this question did not participate in the lunch program.)

A comparison at the district level, however, revealed slightly greater differences between charter school and noncharter school students. On average, the percentage of students eligible for the lunch program at charter schools was 17 percentage points less than the total percentage for noncharter students in the sponsoring districts. As Figure II-7⁽²⁹⁾ illustrates, in approximately 36% of the charter schools, the proportion of students eligible for the lunch program was more than 20 percentage points less than that in their district's noncharter schools.

A Closer Look
Of the schools in which home-based study was the primary mode of instruction for most students, only 6% of the students were reported to be eligible for the National School Lunch Program.⁽³⁰⁾ However, since the number of students in these schools was small (n=698), the effect on average free/reduced-price lunch eligibility in all charter schools was negligible (mean in non-home-based charters = 44% vs. mean in all charter schools = 43%).
Urban charter schools enrolled a greater number of students who were eligible for the lunch program than schools at other locations. On average, 73% of students at urban charter schools were eligible, compared with 36% for rural, 41% for suburban, and 30% for small-town charter schools.

Charter schools were also asked whether or not they were eligible for Title I�the federal program serving low-achieving students. Funding allocations to eligible applicants are based on the proportion of students from low-income families served in these schools and districts. Charter schools varied on whether or not they were eligible for Title I funds: 36% reported they were eligible, 2% reported their eligibility was pending, and 36% reported they were not eligible. More importantly, however, 24% of charter schools did not know whether they were eligible for this large-scale federal funding program. There was a pattern among charter schools of a low level of knowledge about school finances. Further evidence and implications are discussed in Chapter III.

Charter schools served percentages of special education students similar to the state average.

Approximately 8% of all students enrolled in the charter schools received special education services, compared with 9% of the total public school population in California.⁽³¹⁾ Special education services were provided either by the school or by the sponsoring agency. Most of the case study schools that provided special education services did so through their district or county sponsors. The charter schools usually contracted with the sponsoring agency for these services as part of their charter agreements.

Start-up schools were less likely to provide special education services (either directly or indirectly through their sponsors) to students in their schools than were conversion schools. In fact, 26% of start-ups said they did not have any students receiving special education services, compared with only 6% of conversion schools. This difference is not surprising, given that conversion schools were linked into the complex system of special education legislation and regulation before becoming charter schools. Case study conversion schools either provided the services that were previously delivered at the school site or maintained previous connections to district services. Unfortunately, many start-up school leaders were not as familiar with special education and believed, usually incorrectly, that the district automatically took care of this area.

Although charter schools served percentages of special education students comparable to those in noncharter schools, special education was an area that was marked by uncertainty for many charter schools and sponsoring agencies. This situation is not surprising, since special education is a complex maze of regulations that is difficult for staff at noncharter public schools to comprehend. For charter schools, the situation is further complicated because they must determine how their educational programs and the waivers of the charter school law mesh with federal special education laws and regulations. Most of our case study charter schools had little knowledge in this area. Even in cases where services were provided by sponsors to charter school students, many charter school directors were not involved in�or even aware of�the special education referral, assessment, and placement process. Some charter schools reported being uncertain whether or not students had Individualized Education Plans (IEPs) before attending their schools. In these cases, they reported problems with obtaining student files from regular public schools. They also believed that parents withheld information regarding their children's special education needs from charter school staff when they enrolled them because they thought this label stigmatized them.

Charter schools reported a lack of information from the California Department of Education about what their responsibilities were regarding special education. One charter school located in a district other than its sponsoring district was unsure of whose responsibility it would be if a student with severe needs showed up at the school. This charter school worried that meeting the needs of one child with severe disabilities could bankrupt them. The sponsoring district was also very concerned about this liability issue and was waiting for guidance from the California Department of Education. Another district realized that charter school directors in the district were not aware of any of the special education rules and regulations. The district was very concerned about students with disabilities who were not being linked to appropriate services and was working to make charter directors more aware of the special education requirements by developing a training course for them.

Since charter schools often provided small classes and individualized instruction (two components of special education programs), the needs of students in schools that were not providing special education services may have been addressed in a less formal manner. However, the issue of how charter schools�which are based on a theory of waivers from regulations�are supposed to comply with the highly regulated area of special education is an important issue that charter schools and the state must address.

A Closer Look
Of the schools in which home-based study was the primary mode of instruction for most students, only 3% of the students received special education services.⁽³²⁾ However, since the number of students in these schools was relatively small (n=2,569), the effect on the overall percentage of students receiving special education services in all charter schools was negligible (mean in non-home-based charters = 9% vs. mean in all charter schools = 8%).
Almost one-third (30%) of all small schools reported that no students received special education services, whereas 14% of medium/small schools, 23% of medium/large schools, and none of the large schools reported that no students received special education services.

Overall, charter schools served slightly fewer LEP students than the state average.

Twenty percent of all students enrolled in charter schools were identified as limited English proficient (LEP), compared with 24% of the public school population statewide.⁽³³⁾ A district-level comparison, however, revealed that, on average, the proportion of LEP students at charter schools was almost 8 percentage points less than the total proportion among noncharter students in sponsoring districts (see Appendix D for a full presentation of LEP representation data). In fact, the percentage of LEP students was more than 20 points less than the total percentage for noncharter students in the sponsoring district in 24% of charter schools (see Figure II-8).

Start-up schools were less likely to have students identified as limited English proficient in their schools than were conversion schools. In fact, more than half of the start-ups (53%) reported that no students were identified as LEP, compared with only 17% of conversions.

A Closer Look
Small schools (fewer than 100 students)--which are more often start-up schools--were less likely to have LEP students: 77% of these schools reported that none of their students were identified as LEP.
Of the schools in which home-based study was the primary mode of instruction for most students, only .4% of the students were identified as LEP.⁽³⁴⁾ However, since the number of students in these schools was relatively small (n=2,519), the effect on the overall percentage of LEP students in all charter schools was negligible (mean in non-home-based charters = 21% vs. mean in all charter schools = 20%).
Urban charter schools enrolled a greater percentage of LEP students than did schools in other locations. On average, 31% of students at urban charter schools were identified as LEP, compared with 23% at suburban, 18% at rural, and 8% at small-town charter schools.

Languages spoken in charter schools were similar to those spoken by students statewide.

Most charter school LEP students spoke Spanish (87%). As illustrated in Table II-4, this was higher than the proportion in the statewide LEP population, which was 79% Spanish-speaking. Other languages represented in charter schools were similar to the statewide LEP population, except for Vietnamese, Cantonese, and Tagalog, which were slightly underrepresented in charter schools.

Table II-4

LANGUAGES SPOKEN BY LEP STUDENTS IN CHARTER SCHOOLS

Language Spoken Percentage of LEP Students

Charter Schools (n=90) Statewide⁽³⁵⁾

Spanish 87% 79%

Vietnamese 1% 4%

Hmong 3% 2%

Cantonese .5% 2%

Tagalog .8% 2%

Korean .9% 1%

Lao .3% .8%

Russian 1% .5%

All other languages 5% 9%

On average, 41% of charter school students were low achieving.

One intention of the Legislature in enacting the charter law was to "increase learning opportunities for all pupils, with a special emphasis on expanded learning experiences for pupils who are identified as academically low achieving" (Education Code Section 47601 [b]). Unfortunately, there is no agreed-on definition for the term "academically low achieving." On the basis of schools' own definitions, they reported, on average, that 41% of their students were categorized as academically low achieving.⁽³⁶⁾ The criteria most charter schools used to determine the percentage of low-achieving students were scores on standardized tests (80%) and grades from previous schools (52%). Schools also reported using other methods, such as teacher evaluations/assessments/observations, portfolios, performance-based assessments, attendance, and parent assessments. We were not able to compare low-achieving students in charter schools with statewide or district averages because these data are not compiled in regular public schools.

Start-up schools reported enrolling a greater percentage of low-achieving students than conversion schools�37% of start-ups reported that 51-100% of their students were low achieving, compared with only 25% of conversions. These statistics are not necessarily contradictory with reports that start-up schools serve relatively fewer LEP and special education students. In our case study schools, we found that many start-up schools were designed to serve at-risk students, such as those who have been expelled, have dropped out, or have failed to succeed in traditional public schools. These students are not necessarily LEP or special education students; rather, they are students with special needs.

A Closer Look
Urban charter schools reported a greater percentage of low-achieving students (54%) than other schools (suburban, 33%; small-town, 34%; rural, 43%). More than half of the urban charter schools (53%) reported that 51-100% of their students were low achieving, compared with 24% for suburban, 13% for small-town, and 36% for rural charter schools.
Similarly, county-sponsored charter schools reported a greater percentage of low-achieving students than did district-sponsored charter schools. On average, 57% of students in county-sponsored charter schools were low achieving, compared with an average of 39% in district charter schools. Along the same lines, almost half of the county-sponsored schools (43%) reported that most or all of their students (76-100%) were low achieving. Only 14% of district-sponsored charter schools reported this high a concentration.⁽³⁷⁾ These statistics are not unexpected, since many county-sponsored programs--charter and noncharter--were specially designed to serve special-needs students.

On average, charter schools enrolled a slightly higher percentage of students considered gifted than the percentage of students enrolled in statewide gifted programs.

Approximately 9% of all students in charter schools were considered to be gifted and talented students; in other words, charter schools reported that these students would qualify for the state's Gifted and Talented Education (GATE) program or a similar program.⁽³⁸⁾ In comparison, 7% of the statewide student population were enrolled in the GATE program.⁽³⁹⁾ This statistic reinforces other evidence (see section on school admission policies) that charter schools are not "creaming" high-achieving students. This parity with the state average holds true when looking at start-up and conversion schools, with 8% of start-up and 10% of conversion students reported to be qualified for GATE.

Instructional Staff Characteristics

Charter schools employed both full-time and part-time instructional staff, most of whom had credentials.

On average, two-thirds of charter school instructional staff were employed full-time and one-third were part-time. Their sponsors, in contrast, employed on average 85% full-time instructional staff. Full-time charter school employees were more likely than part-time employees to have full state certification for the subjects they taught in their schools (mean = 83% and 49%, respectively). Part-time employees, however, were more likely than full-time employees to have special education credentials (mean = 11% and 4%, respectively). On average, charter schools reported that 71% of all instructional staff (full-time and part-time) had credentials.

Conversion charter schools were more likely to have full-time staff�53% of conversions reported that most or all of their staff (76-100%) worked full-time, compared with only 33% of start-ups. Teachers at start-up schools were less likely to have credentials than their counterparts in conversion schools. On average, 66% of teachers at start-up schools had credentials, compared with 77% of teachers at conversion schools (see Table II-5). Start-up schools also were less likely than conversion schools to have teachers with special education credentials. More than three-fourths of start-up schools (77%) reported that no full-time staff at their schools had special education credentials, compared with only 25% of conversion schools.

Table II-5
CHARTER SCHOOL TEACHERS WITH CREDENTIALS:
START-UP VS. CONVERSION SCHOOLS

Overall (n=96) Start-up (n=48) Conversion (n=48)

Mean percentage of all full-time and part-time staff with credential 71% 66% 77%

Overall, charter school teachers had an ethnic composition similar to teachers statewide.

Charter school instructional staff were primarily White, not of Hispanic origin (71%). Non-White instructional staff represented the following ethnic groups: Hispanics (17%), Blacks (not of Hispanic origin) (6%), Asian or Pacific Islanders (4%), American Indian or Alaska Natives (1%), and Filipinos (1%). These percentages are similar to those found in all public schools in the state (see Table II-6).

Table II-6
RACIAL/ETHNIC TEACHER DEMOGRAPHICS⁽⁴⁰⁾

Race/Ethnicity Charter School Teachers (n=2,790) Teachers Statewide(n=250,527)

White, not of Hispanic origin 71% 78%

Black, not of Hispanic origin 6% 5%

Hispanic 17% 11%

Filipino 1% 1%

Asian or Pacific Islander 4% 4%

American Indian/Alaska Native 1% 1%

Start-up schools were less likely to have an ethnically diverse teaching staff than were conversion schools. In fact, Hispanic, Filipino, and Asian/Pacific Islander teachers appeared to be underrepresented in start-up schools. One-half of the start-up schools reported that they did not have any Hispanic teachers, compared with only 23% of conversion schools. In addition, greater percentages of start-ups reported that they did not have instructional staff who were Filipino (94%) or Asian/Pacific Islander (77%), compared with 75% and 52%, respectively, of conversion schools.

Charter schools depended on volunteers to fill both instructional and support roles.

According to the school survey, charter schools depended on nonpaid instructional staff as part of their educational programs. Some of the roles played by volunteers are typically paid positions in noncharter schools; other roles are generally filled by volunteers in charters and noncharters alike. Many charters, especially those with home-based learning, relied heavily on parents; fewer schools also depended on people from the community. Charter schools reported that the most common roles of nonpaid instructional staff were tutoring and helping in the classroom with reading groups. Charter schools also reported that volunteers helped with playground/recess supervision, art classes, and library assistance. Some charter schools had educational programs built around community expertise. For example, one school had parent and community volunteers teach elective courses.

Nearly two-thirds of charter schools reported that some or all instructional staff were members of unions or other bargaining units.

Although charter schools are not required by law to join bargaining units or unions, many schools (or at least some teachers in many schools) maintain these connections. Nearly two-thirds of charter schools (63%) reported that some or all of their instructional staff were members of unions or other bargaining units. (See Chapter III for further discussion.)

Teacher contracts in a large percentage of schools contained a high number of contract provisions typically found in contractual agreements with unions (although the exact nature of the provisions may have been different). Table II-7 illustrates the provisions that charter schools included in either their contractual agreements (union and non-union) or their personnel policies for instructional staff.

Table II-7
CHARTER SCHOOL CONTRACT PROVISIONS

Percentage of Schools Number of Respondents

Annual performance evaluations 89% 96

Coverage by district retirement benefits or equivalent 85% 89

Benefits package equivalent to or better than the district's package 79% 85

Due process for dismissal proceedings 76% 95

Minimum hours of work 66% 96

One-year at-will contract 58% 93

Granting of tenure at the school or acknowledgment of teacher's right to return to tenured position in district 55% 87

Start-up schools were less likely to include tenure, an equivalent benefits package, and minimum hours of work, compared with conversion schools. Start-ups were more likely to have 1-year at-will contracts�68% of start-ups, compared with 48% of conversions. Conversion schools were more likely to have staff members who belonged to unions (see Table II-8)�79% of conversion schools reported that some or all of their instructional staff were members, compared with 47% of start-ups. Half of the conversion schools reported that all of their staff belonged to unions, compared with only 16% of start-up schools. These comparisons remain robust when home-based study schools are excluded: 71% of start-ups vs. 47% of conversions had 1-year at-will contracts; 78% of conversions vs. 49% of start-ups had some or all staff members who belonged to unions; and in 51% of conversions vs. 16% of start-ups, all staff belonged to unions. One might expect that primarily home-based schools would be less likely to have teacher contract provisions, and thus would bias the start-up figures (24% of the start-ups were home-based schools), but that is not the case.

Table II-8
CHARTER SCHOOL TEACHER UNION MEMBERSHIP:
START-UP VS. CONVERSION SCHOOLS

Overall (n=97) Start-up (n=49) Conversion (n=48)

All teachers were members of bargaining units or unions 33% 16% 50%

Some teachers were members of bargaining units or unions 30% 31% 29%

No teachers were members of bargaining units or unions 37% 53% 21%

A Closer Look

Teacher contracts and personnel policies at charter schools with financial autonomy differed greatly from those at charter schools that reported more ties to their sponsoring agency. For example, 70% of financially autonomous charter schools, compared with 26% of other charter schools, reported that none of their teaching staff belonged to unions. Conversely, only 8% of the financially independent charter schools reported that all of their teachers belonged to unions, compared with 43% of the non-financially independent schools.

Financially autonomous (FA) schools were less likely to report the inclusion of various provisions in their contractual agreements or personnel policies:

Granting of tenure/acknowledged right to return (32% of FA schools vs. 63% of non-FA)

Coverage by district retirement benefits or equivalent (58% of FA schools vs. 95% of non-FA)
Benefits package equal to or better than district's (62% of FA schools vs. 84% of non-FA)
Minimum hours of work (48% of FA schools vs. 71% of non-FA).

Starting teacher salaries in charter schools were consistent with the state average.

Starting teacher salaries reported by charter schools were consistent with the state average of $25,500.⁽⁴¹⁾ The average salary for teachers new to the teaching profession in charter schools was approximately $27,200, with a range from $20,000 to more than $35,000. Some charter schools, primarily those in independent/home study charters, compensated teachers on the basis of the number of students they taught. There was little difference between the salaries of teachers in start-up and conversion charter schools.

A Closer Look
On average, starting salaries were slightly higher at financially autonomous schools--approximately $28,600 at these schools, compared with $26,700 at schools that did not report financial independence. Starting salaries were also slightly higher at urban schools, which reported an average beginning salary of approximately $29,600, compared with $27,000 (suburban), $27,000 (small-town), and $25,000 (rural).

Conclusion

The data show that charter schools were smaller, were slightly less diverse, and had somewhat more restrictive admission policies than other public schools in California. The data also show that there were striking differences between start-up and conversion charter schools. In general, start-ups tended to be farther from the district or state norm on several dimensions than were conversion schools. Compared with conversion schools, start-up charter schools were smaller, more often enrolled students from a wide geographic area, and served more nontraditional grade ranges (e.g., grades 5-8). They were more likely to have instructional staff without union representation and to have fewer traditional contract provisions for instructional staff (e.g., tenure). Start-ups were also less likely to serve special education and LEP students than were conversion schools. At the same time, however, they reported, on average, serving more low-achieving students than conversion schools. Often, start-up schools were designed to serve students who do not succeed in regular school programs. Of course, start-up status did not determine the eventual characteristics of the charter school. Other factors, such as the charter school's relationship with its sponsoring agency and support from the local teachers, school board, and community, played a role as well. Still, start-up charter schools ended up with decidedly different characteristics than both regular public schools and conversion charter schools.

The 15 survey respondents who identified home-based study with the parent as primary instructor as the predominant method of instruction for most students had a distinct profile.⁽⁴²⁾ They were most often located in small towns or rural areas, and were somewhat more likely to require parent contracts, more likely to enroll White students, less likely to enroll LEP students, and less likely to have special education students. The number of schools in this category, and the number of students they serve, is so small that their impact on overall charter statistics is negligible. However, because they represent a distinct variation on charter schools--one, moreover, that was not envisioned by the authors of the legislation--it is important to point out where they differ from charter schools as a whole.

Next, Chapter III presents more detailed information on motivations for starting a charter school, the development of a charter agreement, and school-sponsor relationships.

Continue to SRI--Evaluation of Charter School Effectiveness Part II

Return to LAO Home Page

Year Approved⁽⁵⁾	Number of Schools
1993	42
1994	28
1995	36
1996	15
1997	3
TOTAL	124

	Percentage of Schools	Number of Respondents
Space limitation or enrollment cap	78%	98
Student and/or parent not committed to school's philosophy	44%	97
Evidence that parent or adult cannot fulfill involvement requirements	32%	97
Student's special needs because the school does not provide services such as special education or primary language instruction	30%	93
Residency outside of school or district boundaries	29%	97
Student ethnicity; in other words, the school considers ethnicity in order to achieve racial diversity	17%	98
Student's prior academic performance	3%	97

	Percentage of Start-up Schools	Number of Respondents	Percentage of Conversion Schools	Number of Respondents
Space limitation or enrollment cap	76%	49	80%	49
Student and/or parent not committed to school's philosophy	50%	48	39%	49
Evidence that parent or adult cannot fulfill involvement requirements	37%	49	27%	48
Student's special needs because the school does not provide services such as special education or primary language instruction	44%	45	17%	48
Residency outside of school or district boundaries	12%	49	46%	48
Student ethnicity; in other words, the school considers ethnicity in order to achieve racial diversity	16%	49	18%	49
Student's prior academic performance	6%	48	0%	49

Race/Ethnicity	Percentage of Students⁽²⁵⁾
Race/Ethnicity	Charter SchoolsTotal students: 37,436 (94 schools)	Statewide Total students: 5,612,965
White, not of Hispanic origin	48%	40%
Black, not of Hispanic origin	9%	9%
Hispanic	34%	40%
Filipino	3%	2%
Asian	5%	8%
Pacific Islander	.5%	.6%
American Indian/Alaska Native	2%	.9%

Language Spoken	Percentage of LEP Students
Language Spoken	Charter Schools (n=90)	Statewide⁽³⁵⁾
Spanish	87%	79%
Vietnamese	1%	4%
Hmong	3%	2%
Cantonese	.5%	2%
Tagalog	.8%	2%
Korean	.9%	1%
Lao	.3%	.8%
Russian	1%	.5%
All other languages	5%	9%

	Overall (n=96)	Start-up (n=48)	Conversion (n=48)
Mean percentage of all full-time and part-time staff with credential	71%	66%	77%

Race/Ethnicity	Charter School Teachers (n=2,790)	Teachers Statewide(n=250,527)
White, not of Hispanic origin	71%	78%
Black, not of Hispanic origin	6%	5%
Hispanic	17%	11%
Filipino	1%	1%
Asian or Pacific Islander	4%	4%
American Indian/Alaska Native	1%	1%

	Percentage of Schools	Number of Respondents
Annual performance evaluations	89%	96
Coverage by district retirement benefits or equivalent	85%	89
Benefits package equivalent to or better than the district's package	79%	85
Due process for dismissal proceedings	76%	95
Minimum hours of work	66%	96
One-year at-will contract	58%	93
Granting of tenure at the school or acknowledgment of teacher's right to return to tenured position in district	55%	87

	Overall (n=97)	Start-up (n=49)	Conversion (n=48)
All teachers were members of bargaining units or unions	33%	16%	50%
Some teachers were members of bargaining units or unions	30%	31%	29%
No teachers were members of bargaining units or unions	37%	53%	21%

EVALUATION OF CHARTER SCHOOL EFFECTIVENESS Part I

Contents

EXECUTIVE SUMMARY

I. STUDY OVERVIEW AND DESIGN

II. CHARTER SCHOOL CHARACTERISTICS

III. CHARTER SCHOOL GOVERNANCE, FINANCE, AND ACCOUNTABILITY

IV. TEACHING AND LEARNING IN CHARTER SCHOOLS

V. OUTCOMES OF CALIFORNIA CHARTER SCHOOLS

VI. CONCLUSIONS AND RECOMMENDATIONS

APPENDICES

Executive Summary

I. Study Overview and Design

II. Charter School Characteristics

III. Charter School Governance, Finance, and Accountability

IV. Teaching and Learning in Charter Schools

V. Charter School Outcomes

VI. Recommendations

I. STUDY OVERVIEW AND DESIGN

II. CHARTER SCHOOL CHARACTERISTICS

EVALUATION OF
CHARTER SCHOOL EFFECTIVENESS
Part I

III. CHARTER SCHOOL GOVERNANCE, FINANCE,
AND ACCOUNTABILITY