Follow-up Evaluation of the District of Choice Program

Follow-Up Evaluation of the District of Choice Program GABRIEL PETEK LEGISLATIVE ANALYST FEBRUARY 2021 analysis full gutter AN LAO REPORT LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT Executive Summary Background Legislature Faces Decision About Reauthorizing the “District of Choice” Program. A state law adopted in 1993 allows students to transfer to school districts that participate in the District of Choice program. The program requires participating districts to accept interested students regardless of their academic abilities or personal characteristics. Unlike other interdistrict transfer laws, it also allows students to transfer without seeking permission from their home districts. The program currently consists of 45 participating districts enrolling nearly 9,600 transfer students. The program is scheduled to sunset on July 1, 2023. Legislature Has Made Several Changes to the Program. The Legislature made numerous changes when it last reauthorized the program in 2017. It added new oversight procedures, including a requirement for local auditors to review the districts participating in the program. In addition, it required the California Department of Education (CDE) to collect and publish program data. Regarding the transfer application process, the Legislature required participating districts to make application information easily accessible and prioritize applications from low-income students. These changes were intended to (1) create a mechanism for monitoring compliance with program rules, and (2) increase participation among student subgroups that transfer at lower rates, including low-income students and Latino students. Several members of the Legislature also wanted to know how the program was affecting racial balance among districts. Report Provides Our Second Evaluation. We published our first evaluation of the program in January 2016. The 2017 reauthorization required our office to produce a follow-up evaluation using the newly available data to assess the latest changes and make recommendations regarding further extensions of the program. This report responds to that requirement. Key Findings New Oversight Procedures Uncovered No Major Issues. The new oversight procedures found no systemic problems or districts discriminating against interested students. Local auditors did flag three small districts for missing paperwork. These issues seemed to reflect unintentional oversights or districts being unfamiliar with new requirements, and CDE worked with these districts to ensure they would comply moving forward. Participation by Low-Income Students Has Increased. Similar to our previous evaluation, we found low-income students use the program at relatively low rates compared with other students in their home districts. Participation by these students has increased, however, from 27 percent of transfer students in 2014-15 to 32 percent in 2018-19. Participation by Latino Students Has Increased. Compared with their share of home district enrollment, Latino students use the program at lower rates, whereas Asian students and white students use the program at higher rates. In terms of total participating students, however, Latino students recently overtook white students as the largest users of the program. As of 2018-19, participating students are 40 percent Latino, 28 percent Asian, and 26 percent white (less than 7 percent belong to other groups). Modest Effects on Racial Balance. The program appears to increase racial balance for some districts and reduce it for others, although the changes for most districts are small. Compared to a hypothetical alternative in which all students return to their home districts, the overall effect of the www.lao.ca.gov 1 analysis full gutter AN LAO REPORT program on racial balance is a minor decrease. Compared to a more likely alternative in which only some students return to their home districts, the overall effect appears neutral. Program Provides Students With Additional Educational Options. The program allows students to access educational options that are not offered by their home districts, such as schools with specialized themes or instructional models. In the specific areas we examined—including college preparatory courses, arts and music, and foreign languages—students gained access to an average of five to seven courses not offered by their home districts. Nearly all students transfer to districts with higher test scores than their home districts. Home Districts Often Take Steps to Improve Their Instructional Offerings. As we reported in our previous evaluation, home districts often respond to the program by taking action to gain clarity about the priorities of their communities and by implementing new educational programs. We also found that the home districts most affected by the program have made above-average gains in student achievement over the past several years, although the role of the program in these gains is difficult to determine. Recommendations Reauthorize the Program. The program provides quality educational options for a range of students, the new oversight mechanisms uncovered no notable concerns, and the share of disadvantaged students using the program has risen. We think these strengths merit reauthorization, potentially on a permanent basis. Alternatively, the Legislature could reauthorize the program for at least five more years and reassess trends near the end of that period. In either case, the state could continue monitoring program data and the Legislature could have future oversight hearings to ensure the program continues to meet its goals. Repeal Cumulative Cap. We recommend repealing a cap that limits the cumulative number of students who can transfer out of each district. Once reached, this cap disallows all future participation regardless of student interest or the quality of transfer options available. Without changes to the cap, program participation seems likely to drop over the next several years. In addition, we understand the original purpose of the cap was to mitigate fiscal impacts on home districts, but the cumulative number of transfers over the life of the program does not seem to be a good indicator of a district’s current fiscal condition. Allow Later Application Deadline. Current law requires students to submit their transfer applications for the upcoming school year prior to January 1. This deadline falls before some students and families have begun examining their options, and it limits participation for students who have less existing awareness of the program, including low-income students. We recommend delaying the deadline, potentially to March 1. Increase Funding for Basic Aid Districts. The 2017 reauthorization significantly reduced funding for students transferring to basic aid districts (districts with high levels of local property tax revenue). We found that this reduction has led these districts to accept fewer transfer students. In addition, the students transferring to these districts are more likely to be disadvantaged than other transfer students. We recommend setting the funding rate closer to pre-2017 levels and providing a higher rate for low-income students and English learners. Continue Collecting Data. We recommend the state continue collecting data about the number and characteristics of students who transfer through the program. Ongoing data collection would help the Legislature monitor the 2017 changes as well as the effects of our recommendations. If the Legislature were interested in developing a greater understanding of why participation varies by student subgroup, it could consider funding a survey of students and parents. The findings from this survey could inform future efforts to promote participation by students who use the program less frequently. 2 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT INTRODUCTION California Has District of Choice Program. evaluation using newly available data. Chapter 15 of California has several laws designed to give parents a 2017 (AB 99, Committee on the Budget) sets forth choice about which school their children attend. One the requirements for this second evaluation: of these laws is called the District of Choice program. “The Legislative Analyst shall conduct, after This program allows a student living in one school consulting with appropriate legislative staff, district to transfer to another school district that has a comprehensive evaluation of the [District of deemed itself a District of Choice. The program differs Choice program] and prepare recommendations from other interdistrict transfer laws because it does regarding the extension of the program. The evaluation shall incorporate the [demographic, not require students to apply with the home districts fiscal, and academic data specified in law] and they are leaving. shall be completed and submitted, along with Legislature Faces Decision About Whether the recommendations regarding extension of the to Extend the Program. Though the state initially program and… implementation of the program adopted the District of Choice program as a to ensure access to the program for all pupils, to five-year pilot, it has reauthorized the program the appropriate education policy committees of several times. The most recent reauthorization the Legislature and to the Department of Finance extended the program until July 1, 2023. As we by January 31, 2021.” describe later in this report, the program currently Report Has Four Main Sections. This report consists of 45 participating school districts enrolling responds to the statutory evaluation requirement. nearly 9,600 transfer students. Participating districts The first section outlines the main features of the are located throughout the state, with the highest program and compares it with other interdistrict concentrations in the Counties of Los Angeles, Kern, transfer options in California. The second section and Sonoma. describes the findings that emerged from our analysis Law Requires Follow-Up Evaluation. At the of the available data. The third section assesses direction of the Legislature, we published our first the program and discusses a few policies limiting evaluation of the program in January 2016. To assist access for interested students. The final section its deliberations about the future of the program, the makes recommendations regarding the future of the Legislature directed our office to produce another program. BACKGROUND In this section, we describe the origins and main be more responsive to community concerns and by features of the District of Choice program. We also allowing parents to choose the instructional setting review the most recent changes to the program and best suited to the interests of their children. Other explain how the program differs from other transfer supporters were concerned about a proposal to options in the state. provide state funding for students attending private schools and thought expanding choice among public History schools was preferable. The Legislature responded Legislature Adopted District of Choice to these concerns by enacting three laws. The Program in 1993. The District of Choice program first, the Charter Schools Act of 1992, allowed the grew out of an effort in the early 1990s to increase the establishment of charter schools that could operate choices available to students within the public school independently from school districts. The second, system. Two main considerations motivated this enacted in 1993, gave students more options to effort. First, many supporters believed choice would transfer to other schools within the same district. The improve public education by encouraging schools to third law, also enacted in 1993, created the District of www.lao.ca.gov 3 analysis full gutter AN LAO REPORT Choice program. Although this law was not the first cases, a district adopts a specific transfer limit for to allow interdistrict transfers, it was designed to be each grade level. much less restrictive. Students Can Leave Their Home District and Legislature Has Reauthorized the Program Six Attend a District of Choice. Once a district has Times. The 1993 legislation implemented the District deemed itself a District of Choice, it may begin of Choice program as a five-year pilot, with the first accepting attendance applications from students transfers occurring in the 1995-96 school year. The in other districts. Interested students apply directly Legislature extended the program for five more years to the District of Choice and do not need to seek in 1999, followed by additional extensions in 2004, permission from the home districts they attend. (As 2007, 2009, 2015, and 2017. The last extension we discuss below, state law allows a home district authorized the program until July 1, 2023. to prohibit transfers under certain conditions.) Students seeking to transfer for the upcoming Program Basics school year must submit their applications prior to A School District May Decide to Become a January 1 of the current school year. A district must District of Choice. Figure 1 summarizes the key give each applicant a provisional notification of its components of the District of Choice program. To decision to accept or deny the transfer application participate in the program, a district must register by February 15, with final notification required by with the California Department of Education (CDE) May 1. Although transfer students are not guaranteed and its county board of education. The governing attendance at specific schools, most districts allow board of the district must also adopt an annual students to rank their preferred schools and honor resolution specifying the maximum number of any preferences that do not displace students already transfer students it is willing to accept. Each district enrolled. Transfer students accepted into a district determines its own transfer limit, typically after are not obligated to attend. They may withdraw their assessing multiple factors including facility space, applications at any time and return to their home staffing requirements, and overall enrollment. In most districts. The law waives the regular application deadline for students from military families relocated Figure 1 Key Components of the District of Choice Program 9 District Participation. A district opts into the program by registering with the state and its county board of education. The district also adopts a resolution specifying the maximum number of transfer students it will accept. 9 Transfer Rules. A student’s “home district” must allow the student to transfer unless the transfer would affect the home district in one of the following ways: • Exceed an annual cap equal to 3 percent of the home district’s student attendance for the year.a • Exceed a cumulative cap equal to 10 percent of the home district’s average annual attendance over the life of the program.a • Exacerbate severe fiscal distress. • Hinder a court-ordered desegregation plan. • Negatively affect racial balance. 9 Admission Procedures. A District of Choice must accept all interested students up to its locally approved limit and conduct a lottery if oversubscribed. An oversubscribed district must give priority to the siblings of students who already attend the district, low-income students, and students from military families. 9 Funding Allocations. When a student transfers, the home district no longer generates funding for that student and the District of Choice begins generating the associated funding.b a For districts with more than 50,000 students, the annual cap is 1 percent and the cumulative cap is not applicable. b Different rules apply for basic aid school districts. 4 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT within the past 90 days. These students may transfer their academic performance, athletic ability, or any midyear to any district in the program with available other personal characteristic. Finally, the law requires space. districts to provide translated application information A District of Choice Must Accept All Interested if they receive students from districts in which Students up to Its Local Limit. A distinguishing more than 15 percent of students speak a primary feature of the District of Choice program is the language other than English. requirement for a district to accept all interested …And With Home Districts. The law requires students up to the number specified in its board participating districts to share information with the resolution. The law explicitly prohibits giving any districts students are leaving at various points. consideration to academic or athletic performance Specifically, a District of Choice must provide each or to the cost of educating a student. If the number home district a preliminary list of approved transfers of transfer applications exceeds a district’s locally for the upcoming school year by February 15 and approved limit, the district must conduct a lottery an updated list by May 2. These communication at a public meeting to determine which students requirements are intended to help home districts it will accept. This lottery must give priority to develop budgets and staffing levels that account for students whose siblings already attend the district, the number of students transferring out. (State law low-income students, and students from military requires districts to make certain preliminary staffing families (in that order). The law only allows districts to decisions for the upcoming year by March 15 and deny transfer requests on an individual basis under final decisions by May 15.) two limited circumstances. First, a provision dating Home Districts Can Prohibit Transfers for a to 1993 allows a district to reject an application that Few Reasons. The law allows a home district to would require it to create a new program to serve an prevent a student from transferring out under a few incoming student. A subsequent reauthorization of circumstances. Specifically, a home district may the program, however, largely negated this provision prohibit a transfer that would affect it in one of the by prohibiting districts from rejecting English learners following ways: and students with disabilities. Second, the law • Exceed a 3 Percent Annual Cap. A home provides for a district to deny an application that district may limit the number of students would displace a student who already attends or transferring out each year to 3 percent of its resides in the district. average daily attendance for that year. (The Accepted Students Do Not Need to Reapply annual cap is 1 percent for home districts with Each Year. Another distinguishing feature is the more than 50,000 students.) provision allowing students to remain in the program • Exceed a 10 Percent Cumulative Cap. In until they graduate. Once a district accepts a student addition to the 3 percent annual cap, the law through the program, it can revoke the transfer only allows a home district to deny transfers that through formal expulsion proceedings. The law also exceed a cumulative cap. (Home districts with allows a district to revoke all transfers if it withdraws more than 50,000 students cannot invoke this from the program. In the case of withdrawal, the cap.) In a 2011 dispute over the calculation of district must allow high school students to remain the cap, an appellate court determined that until they graduate. the cap is equal to 10 percent of a district’s Program Sets Rules for Communication With average annual attendance over the life of the Students… The law requires each participating program. Every student who has transferred district to post application information on its website, since the inception of the program counts including the procedures for submitting an application toward the cap, including students who are no and the relevant deadlines. The law also requires longer enrolled. Upon reaching this cap, a home districts to make public announcements about the district may prohibit all further transfers. availability of the District of Choice program. Any • Exacerbate Severe Fiscal Distress. If a home communication from the district must be “factually district receives a negative budget rating from accurate” and avoid targeting students based on www.lao.ca.gov 5 analysis full gutter AN LAO REPORT its county office of education (COE) (meaning year. Specifically, the state credits all districts with the the district will be unable to meet its financial higher of their current- or prior-year attendance for obligations for the current or upcoming year funding purposes. without corrective action), the district may limit Lower Funding Rate for Basic Aid Districts. the number of students transferring out. A About 10 percent of school districts have local home district also may limit transfers if its COE property tax revenue exceeding their allotments determines that the district will fail to meet state calculated under the Local Control Funding Formula. standards for fiscal stability exclusively due to The state allows these districts to keep this additional the impact of the program. revenue and treat it like other general purpose • Hinder a Court-Ordered Desegregation funding. These districts are known as basic aid Plan. If a home district is operating under districts (a term derived from the section of the State a court-ordered desegregation plan, it may Constitution guaranteeing all school districts at least prohibit transfers that would hinder the plan. $120 per student from the state.) Under the District of • Negatively Affect Racial Balance. A district Choice program, funding for basic aid districts works may limit transfers that would “negatively differently than it does for other districts: impact” its voluntary desegregation • Transferring From Nonbasic Aid to Basic Aid plan or racial and ethnic balance. Any District. If a basic aid district enrolls a student usage of this provision must comply with from a home district that is not a basic aid Proposition 209 (1996), a constitutional district, the basic aid district receives 25 percent amendment that prohibits schools and other of the base grant the student would have agencies from using race and ethnicity (among generated in his or her home district. The other factors) as a factor in public programs. student does not generate any supplemental or Outside of these five provisions, the law prohibits concentration funding that would apply in the home districts from adopting any policies to block home district. These types of transfers therefore or discourage students from transferring out. In generate state savings relative to other types of addition, home districts may not prohibit transfers by transfers. students from military families for any reason. • Transferring From Other Basic Aid Districts. If a basic aid district enrolls a student from Funding another basic aid district, the law provides for Funding Follows Students to Their District no exchange of funding between the districts. of Choice. California funds school districts based From the state’s perspective, these transfers are on student attendance, with per-pupil funding cost neutral. rates determined by the Local Control Funding Formula. This formula, adopted in 2013, establishes Oversight a base grant for all students that varies by grade Law Requires Participating Districts to Collect span but is otherwise uniform across the state. and Report Data. The law requires each participating Low-income students and English learners generate district to track the following information: (1) the total a “supplemental grant” equal to 20 percent of the number of students applying to enter the district each base grant. In districts where these students make year; (2) the outcome of each transfer application up more than 55 percent of the student body, the (granted, denied, or withdrawn), as well as the formula also provides a “concentration grant.” The reason for any denials; (3) the race, ethnicity, gender, total allotment for each district is funded through socioeconomic status, and home district of each a combination of state aid and local property tax student transferring into the district; (4) the number revenue. When a student transfers, the home district of English learners and students with disabilities no longer generates funding for that student and the transferring into the district; and (5) the number and District of Choice begins generating the associated characteristics of students receiving transportation funding. Another state policy, however, tends to from the district. The law also requires each district mitigate the fiscal effect on home districts for one 6 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT to prepare an annual report summarizing this CDE to investigate any complaints it receives about information. No later than October 15 each year, the districts (1) participating in the program without district must provide all adjacent districts with a copy registering or (2) failing to report required data. If of this report, as well as a notice indicating its intent the department substantiates a complaint, it must to remain in the program for the following school year. withhold funding generated by the district’s transfer In addition, the law requires districts to report this students until the district has registered for the information to CDE. The department collects district program and provided any missing data. data on applications and school transportation as Changes to the Program part of an annual survey it administers to all districts. It collects data on the characteristics of transfer 2017 Reauthorization Made Several Notable students through the California Longitudinal Pupil Changes. Several important components of Achievement Data System (CalPADS). The law the program reflect changes the Legislature requires CDE to publish a summary of the information adopted in 2017 when it reauthorized the it collects on its website. Whereas the requirement for program. Figure 2 (see next page) shows how districts to produce a local report dates to the original the current version of the program differs from version of the program, the requirement for CDE the previous version. The most notable changes to collect and publish data began in 2018-19. The involved (1) making districts subject to annual state currently provides CDE an annual General Fund audit, (2) requiring CDE to collect and report data, appropriation of $117,000 to support these activities. (3) adding transfer priority for low-income students, School Districts in California Undergo Annual (4) reducing funding for basic aid districts, and Audits. Every school district in California undergoes (5) requiring districts to make application information an annual audit to assess the accuracy of its financial available online. Most of these changes took effect statements and determine whether it complied with in 2018-19. The funding reduction for basic aid various state and federal laws. A district hires its districts, however, became effective in 2017-18. auditor from a list of approved firms. The auditor Other Transfer Options then conducts an independent review following procedures set forth in the school district audit State Has a Few Other Interdistrict Transfer manual developed by the state. If the district is out Options. In addition to the District of Choice of compliance with any of the applicable laws, the program, the state has a few other laws allowing auditor communicates the finding in its audit report. students to transfer to other school districts. Each Depending on the nature of the finding, the district’s law functions independently of the others and COE or CDE will work with the district to implement a includes different requirements for the students and corrective action plan. districts involved. Many districts accept students Parts of the District of Choice Program through more than one option. (All of the interdistrict Subject to Audit. Since 2018-19, the state has transfer options are separate from an intradistrict required auditors to ascertain whether districts transfer law that allows students to attend other complied with certain parts of the District of Choice schools within the same district if space is available.) program. Specifically, an auditor must verify that Interdistrict Permit System Allows Transfers a participating district (1) registered with the state With the Agreement of the Districts Involved. A and its county board of education, (2) accepted longstanding state policy allows a student to transfer all interested students up to its locally approved from one district to another when both districts sign limit, (3) conducted a lottery if it received more a permit consenting to the transfer. With respect applications than it could accept, and (4) collected to students leaving, most districts have a limited all data required by law. CDE generally is responsible set of reasons for which they will release a student. for addressing any audit findings identified in these Examples include the availability of child care in the areas. other district or the attendance of a sibling already Law Authorizes Special Complaint Process. enrolled in the other district. Students who are Separate from the audit process, the law requires denied a release may appeal the decision to their www.lao.ca.gov 7 analysis full gutter AN LAO REPORT Figure 2 State Made Notable Changes to District of Choice Program in 2017 Issue Prior Law Current Law Oversight Auditing District not subject to regular audit process. Local auditors required to review district compliance with program rules each year. Registration Not required. Districts required to register with the state and their county boards of education. Data reporting CDE not required to publish data about the CDE required to maintain a list of Districts of Choice and program. publish data about the program. Application Process Transfer priority Siblings of students who already attend the Siblings of students who already attend the district, district and students from military families. low-income students, and students from military families. Application forms No specific requirements. Forms and application information must be posted online, sometimes in multiple languages. Other Funding for basic aid districts 70 percent of LCFF base rate. 25 percent of LCFF base rate. Information shared with home A summary of the transfer requests approved A preliminary list of students transferring for the upcoming districts for the upcoming year by May 15. year by February 15 and an updated list by May 2. CDE = California Department of Education and LCFF = Local Control Funding Formula. county board of education. After students receive transfers that exceed certain annual limits, which vary their permits, their new districts typically require based on district size. Prior to 2016, this program them to maintain satisfactory attendance, behavior, operated under a series of temporary extensions. and grades as a condition of remaining enrolled. Chapter 106 of 2016 (AB 2537, O’Donnell) made the Chapter 781 of 2019 (AB 1127, Rivas) made several law permanent. changes to the interdistrict permit system. Most Open Enrollment Act Formerly Provided notably, the law now prohibits districts from granting Another Transfer Option. In 2010, the state or denying a permit application based on a student’s enacted the Open Enrollment Act, a law designed academic performance, athletic ability, or any other to provide additional options for students attending personal characteristic. The law also contains a new schools with low test scores. The law required the requirement for districts to provide “transportation state to rank schools according to the performance assistance” to low-income students upon request. of their students on state assessments and place Districts May Accept Transfers Based on the 1,000 schools with the lowest scores on the Parental Employment. A state law adopted in 1986 “Open Enrollment List.” A student attending one of allows a district to admit any student who has at these schools could transfer to another school within least one parent or legal guardian employed within or outside of the district, provided that school had the boundaries of that district for at least ten hours higher test scores and available space. In 2016-17, during the school week. (The law is frequently known the state implemented a new accountability system as the “Allen Bill” after its author, Assembly Member under which it publishes multiple measures of Doris Allen.) If a district participates in this program, performance for each school based on test scores, it may not deny a transfer based on a student’s graduation rates, attendance rates, and suspension personal characteristics, including race, ethnicity, rates, among other indicators. The new system does parental income, and academic achievement. It may, not specify a mechanism for ranking schools, and the however, deny the transfer of a student who would state no longer publishes the Open Enrollment List. cost more to educate than the additional funding Due to these changes, transfers under this law are no generated for the district. The law also allows the longer available. districts from which students are leaving to prohibit 8 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT FINDINGS In this section of the report, we share our findings, in elementary or middle school and one-third are organizing them around the areas of district and in high school (similar to the state as a whole). As student characteristics, districts finances, academic Figure 5 (see page 11) shows, these students outcomes, and program oversight. Within each are concentrated among five large districts. One section, our analysis responds to the statutory district, the Walnut Valley Unified School District requirements for the evaluation as well as questions in eastern Los Angeles County, enrolls more posed by the Legislature and legislative staff. Most of than 3,000 participating students—nearly one-third our findings reflect data for 2018-19—the first year of the total. The five largest districts combined enroll in which the state collected complete data for the nearly 7,000 students (more than 70 percent of the program. Where possible, we make comparisons to total). As we discuss in the box on page 12, most the results in our previous evaluation, which reflected Districts of Choice also accept students through the 2014-15 data. We also draw upon several district interdistrict permit system. (On a statewide basis, the interviews we conducted during the fall of 2019 and a interdistrict permit system has far more participating larger set of interviews we conducted for our previous districts and students.) evaluation. Figure 3 (see next page) summarizes our Apart From Five Largest Districts, Most findings, which we discuss in detail below. Districts of Choice Are Small and Rural. The five districts with the greatest participation range from DISTRICT AND STUDENT medium to large in size. All five are located in urban or CHARACTERISTICS suburban areas, and all are unified (enrolling students in grades K through 12). These characteristics differ Below, we provide information about the districts notably from the other 40 districts participating in the participating in the program and the characteristics program. Among these other districts, 78 percent of students who transfer to them. We also compare have fewer than 1,000 total students, 78 percent are these students to their home districts and explore located in rural parts of the state, and 83 percent are how the program affects racial balance. Finally, we elementary districts (enrolling students in grades K analyze differences in participation based on family through 8 only). income and a few other student characteristics. Fewer Participating Districts Compared With 2014-15. We estimate that 49 districts were Overview of Participation participating in the program in 2014-15. This total State Has 45 Districts of Choice. The state includes the 47 districts identified in our previous had 45 registered Districts of Choice during evaluation and two small districts we identified later. the 2018-19 school year, representing 5 percent Between 2014-15 and 2018-19, 13 of these districts of the 944 districts in the state. As Figure 4 (see left the program and 9 new districts joined the page 11) shows, participating districts are located program. This turnover occurred mainly among small throughout the state. The highest concentrations are districts. In both 2014-15 and 2018-19, the same five located in Los Angeles and surrounding counties, the large districts accounted for most of the participating North Bay, and the southern part of the Central Valley. students. Districts of Choice Enroll About 9,600 Transfer Most Districts Leaving the Program Cited Students. We identified 9,568 students using the Reductions in State Funding or Increases District of Choice program as of the 2018-19 school in Resident Enrollment. We analyzed the year, representing less than 0.2 percent of students characteristics of the 13 districts leaving the program in the state. This total reflects our estimate of all and surveyed their superintendents. We found transfer students actively enrolled in 2018-19, that four of these districts had no transfer students regardless of the year in which they first transferred. enrolled as of 2014-15, making their departure Approximately two-thirds of these students are from the program largely a formality. The remaining www.lao.ca.gov 9 analysis full gutter AN LAO REPORT Figure 3 Summary of Findings District and Student Characteristics Program Scope. As of 2018-19, the state has 45 Districts of Choice enrolling nearly 9,600 transfer students. Participation by Race. Latino students recently overtook white students as the largest users of the program. As of 2018-19, participating students are 40 percent Latino, 28 percent Asian, and 26 percent white (less than 7 percent belong to other groups). Relative to their share of home district enrollment, Asian students and white students transfer at relatively high levels, and Latino students transfer at relatively low levels. Effects on Racial Balance. The program appears to increase racial balance for some districts and reduce it for others, although these changes typically are small. Compared to a hypothetical alternative in which all students return to their home districts, the overall effect of the program on racial balance is a minor decrease. Compared to a more likely alternative in which only some students return to their home districts, the overall effect appears neutral. Participation by Income Level. Low-income students account for 32 percent of program participants, an increase of 5 percentage points relative to 2014-15. Low-income students transfer at low rates compared with their share of home district enrollment. The early application deadline for the program (January 1) can be a challenge for these students. District Finances Effects on Enrollment. The median District of Choice generates 22 percent of its total enrollment from students transferring through the program. These districts draw students from 196 different home districts. Enrollment decreases among home districts tend to be small (usually less than 1 percent and rarely more than 5 percent). Basic Aid Districts. Basic aid school districts have reduced the number of students they are willing to accept through the program in response to the lower funding rate the state implemented for these districts in 2017-18. Academic Outcomes Educational Options. The program allows students to access educational options that are not offered by their home districts, such as schools with specialized themes or instructional models. In the specific academic areas we examined, students gained access to an average of five to seven courses not offered by their home districts. Nearly all students transfer to districts with higher test scores than their home districts. Home Districts. Home districts often respond to the program by taking action to retain students, such as adding new programs or educational options. The home districts most affected by the program have made above-average gains in student achievement over the past several years, although the role of the program in these gains is uncertain. Program Oversight Audit Results. Auditors did not find any districts improperly denying transfer applications. Three small districts, however, did not complete all of the paperwork required to register for the program in 2018-19. Transfer Denials by Districts of Choice. Districts of Choice approved nearly 90 percent of the transfer applications they received in 2018-19. Almost all denials involved districts reaching their locally determined transfer limits. Transfer Denials by Home Districts. At least four home districts have prohibited all future transfers using the cumulative cap. Most of the drop in student participation between 2014-15 to 2018-19 appears related to this cap. 10 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT Figure 4 DDiissttrriiccttss ooff CChhooiiccee iinn CCaalliiffoorrnniiaa As of 2018-19 Transfer Students Enrolled More Than 500 Fewer Than 500 Figure 5 Participation in the District of Choice Program 2018-19 Share of Enrollment District of Choice Share of All District Total District From District of Choice District(s) Students of Choice Students Enrollmenta Program Walnut Valley Unified 3,025 32% 13,879 22% Oak Park Unified 1,424 15 4,579 31 Riverside Unified 992 10 40,708 2 West Covina Unified 790 8 8,754 9 Glendora Unified 734 8 7,198 10 Basic aid districts (18) 969 10 12,440 8 All other districts (22)b 1,634 17 30,954 5 Totals 9,568 100% 118,512 8%c a Excludes students attending independently managed charter schools. b Includes five districts with no students currently participating in the program. c This average is affected by a handful of relatively large districts with relatively few participating students. The median is 22 percent. www.lao.ca.gov 11 analysis full gutter AN LAO REPORT Comparing the District of Choice Program With Interdistrict Permits District of Choice Program Is Small Compared With the Interdistrict Permit System. Whereas 9,568 students were using the District of Choice program in 2018-19 (less than 0.2 percent of all students statewide), 146,109 students were using an interdistrict permit (2.4 percent of students). Similarly, the 45 districts participating in the District of Choice program compare with the 635 districts accepting students through interdistrict permits. Although we are uncertain what factors account for this difference, we did ask district administrators about potential explanations. Several noted that in cases where home districts were willing to sign interdistrict permits, participation in the District of Choice program often seemed unnecessary. Others believed districts preferred the interdistrict permit system because it provides districts with more control (such as the ability to revoke transfers by students with poor behavior or grades). Most Districts of Choice Accept Students Through Interdistrict Permits. We found that 31 of the 45 participating districts accept students through both the District of Choice program and the interdistrict permit system. These 31 districts had 4,870 students enrolled through interdistrict permits in 2018-19—about half as many as they received through the District of Choice program. (The state generally does not publish statewide or district-level data for the interdistrict permit system. Our analysis relies upon a special data extraction prepared by the California Department of Education.) nine districts also had relatively low participation, 1,304 students among districts that participated averaging around 20 transfer students each as in the program in both 2014-15 and 2018-19, of 2014-15. Of the six districts responding to our (2) a decrease of 180 students associated with survey, three were basic aid districts that cited the the 13 districts that left the program after 2014-15, lower funding rate the state adopted in 2017-18. and (3) an increase of 713 students associated with According to these districts, the funding generated the 9 districts that recently joined the program. The by their transfer students no longer justified the cumulative cap on transfer activity appears to be a costs of the program. Two districts attributed their major factor in the overall decrease over the period. decision to increases in their residential enrollment. We provide more detail on these trends later in this These districts indicated that due to space and report. other constraints, they no longer were interested Participation by Race in accepting students from other districts. The last district indicated that permissive interdistrict permit Latino Students Account for the Largest Share policies adopted by other districts in its county of Participating Students. Data for the District of had made its participation in the District of Choice Choice program show that participating students program unnecessary. were 40 percent Latino, 28 percent Asian, and Fewer Participating Students Compared 26 percent white as of 2018-19. All other subgroups With 2014-15. We estimate that 10,339 students account for a much smaller share of program were using the program in 2014-15. (This total participants. Students indicating they belong to reflects the estimate from our previous evaluation, two or more races accounted for 3.5 percent of with an adjustment for students attending the program participants. Black students accounted two districts we identified later.) Compared for 2.4 percent, and American Indian and Pacific with 2014-15, program participation in 2018-19 is Islander students each accounted for less down 771 students (7.5 percent). This decrease than 0.5 percent. is the net change resulting from (1) a decrease of 12 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT Share of Latino Students Figure 6 Rising Over Time. Figure 6 shows Latino Students Have Become the how the program has changed Largest Users of the Program from 2014-15 to 2018-19. Most notably, Latino students overtook Share of Total District of Choice Students white students as the largest 40% users of the program over this period. This increase is due to a 35 majority of participating districts 2014-15 30 (about 60 percent) increasing 2018-19 their share of Latino students, 25 including a significant increase in Riverside Unified (one of the five 20 large districts). In addition, two 15 smaller districts with especially high shares of Latino transfer students 10 (72 percent for one district and 97 percent for the other district) 5 joined the program after 2014-15. Asian Students Participate Latino Asian White Blacka All Othera, b at the Highest Levels Relative to Their Share of Home District a Due to limitations in the 2014-15 data—as well as the very small number of students involved—changes for these groups should be interpreted with caution. The differences from 2014-15 to 2018-19 may reflect Enrollment. After analyzing better data rather than changes in underlying levels of participation. trends in overall participation, we b Consists of students who identified as American Indian, Pacific Islander, or two or more races. examined relative participation levels for the three largest Figure 7 groups of students using the program. Figure 7 compares the Comparing Shares of Students by Race share of students using the program Share of Total, 2018-19 with the average among home districts. Although Latino students 70% account for the largest users of Participating Transfer Students the program, they participate at 60 Home District Averagea relatively low levels compared District of Choice Averagea 50 with their overall share of home Students on Interdistrict Permitsb district enrollment. Specifically, they 40 Statewide Average average 68 percent of home district enrollment but only 40 percent of 30 participating transfer students. By contrast, Asian students 20 have relatively high participation, 10 accounting for 9 percent of home district enrollment and 28 percent of participating transfer students. Latino Asian White White students also have relatively high participation, accounting a Reflects district averages calculated prior to District of Choice transfers. Excludes students attending charter schools in these districts. Percentages are weighted according to the number of students transferring in or out for 17 percent of home district through the District of Choice program. enrollment and 26 percent of b Reflects weighted average for students transferring to a District of Choice using the interdistrict permit system. www.lao.ca.gov 13 analysis full gutter AN LAO REPORT participating students. As the figure shows, these developed by the Ed-Data Partnership in the 1990s. trends are not unique to the District of Choice The EDI measures the extent to which a district’s program. Similar differences in participation are students are distributed evenly among the racial evident in the interdistrict permit system. categories tracked by the state. It can theoretically Differences Between Participating Students range from 0 (for a district in which all students and Home Districts Have Narrowed Somewhat belong to a single race) to 100 (for a district with Over Time. The differences in participation by race equal numbers of students from every category). mirror the findings from our previous evaluation of For our analysis, we adjusted the EDI to focus on the program. Compared with our previous findings, Latino, Asian, and white students (the largest users however, the differences between participating of the program). For an individual district, incoming transfer students and their home districts have or outgoing transfer students increase the EDI when narrowed somewhat (Figure 8). In 2014-15, the they reduce the differences in the share of enrollment difference between the share of Latino students using attributable to each group. Conversely, they decrease the program (32 percent) and their share of home the EDI when they magnify these differences. (Our district enrollment (66 percent) was 34 percentage analysis focuses specifically on changes at the district points. In 2018-19, this difference had decreased level. School level transfer data are not available.) to 28 percentage points. Differences in the share …And Two Counterfactual Scenarios. To of white students using the program also have analyze changes in racial balance, we compared the decreased to some extent. current distribution of students with the distribution Effects on Racial Balance Analyzed Using that would exist in the absence of the program. Our the Ethnic Diversity Index (EDI)... Overall student first counterfactual assumed all students would characteristics and relative participation levels attend their home districts. Although this assumption do not necessarily indicate how the program is provides a baseline, we do not think it reflects how affecting racial balance among districts. To address students would react to the end of the program. the Legislature’s questions about changes in racial District administrators often told us students and balance, we rely upon a version of the EDI, an index parents had strong feelings about transferring. In a survey involving a similar program in Wisconsin, only about one-third Figure 8 of parents indicated they were Differences in Participation Levels “very likely” to send their children Have Decreased Somewhat to their home districts if the Difference Between Home Districts and Transfer Students, interdistrict transfer option did not In Percentage Pointsa exist. Another one-third said they 30 were “not at all likely” to do so. Our second counterfactual simulates 2014-15 20 a more realistic alternative by 2018-19 assuming about one-third of 10 Above average students using the program would participation transfer to the same districts using Asian White Below average participation the interdistrict permit system, -10 another one-third would attend -20 their home districts, and the remainder would pursue some -30 other option (such as a charter -40 Latino school). Our second counterfactual accounts for differences in the a For each group, reflects the average percentage of students transferring through the willingness of home districts to District of Choice program minus the average percentage of home district students. grant interdistrict permits. It also 14 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT assumes low-income students would be much more in EDI for the subset of districts most affected by likely to attend their home districts than high-income the program. (The figure contains all districts with students. (District administrators told us that at least 300 total students and at least 2 percent higher-income students were more likely to have of their enrollment transferring in or out through explored other options before settling on the District the program.) The top portion shows the effect of of Choice program.) the program relative to the alternative in which all Districts of Choice Becoming Slightly students return to their home districts. Relative to this More Balanced, Home Districts Slightly Less. baseline, Districts of Choice are becoming slightly Figure 9 contains our estimates of the change more balanced overall, with their EDI increasing Figure 9 Effects of the District of Choice Program on Racial Balance Change in Ethnic Diversity Indexa Program Effects Relative to the Alternative Where All Students Return to Their Home Districts Less Balance More Balance Districts of Choice -5.0 -3.0 -1.0 1.0 3.0 5.0 7.0 Home Districts Less Balance More Balance District Group Average Overall Average Program Effects Relative to the Alternative Where Students Pursue a Mix of Options Less Balance More Balance Districts of Choice -5.0 -3.0 -1.0 1.0 3.0 5.0 7.0 Home Districts Less Balance More Balance a Reflects districts with at least 300 students and at least 2 percent of their enrollment transferring in or out through the District of Choice program. Ethnic Diversity Index can range from 0 to 100. www.lao.ca.gov 15 analysis full gutter AN LAO REPORT by an average of 0.7. Home districts, on average, to qualify for the federal free or reduced-price meals are becoming slightly less balanced, with their program. For the 2018-19 school year, a family of EDI decreasing by an average of 1.0. The bottom four qualified for this program if its annual income portion shows the effect of the program relative did not exceed $46,435. As a point of comparison, to our second counterfactual, which accounts for 59 percent of all students statewide qualify for free or the likelihood that students would pursue a mix of reduced-price meals. (The program data technically options. Relative to this baseline, the increase in the include all students—regardless of income—who are EDI for Districts of Choice still averages 0.7. The homeless, migrant, or foster youth, or who come from average decrease for home districts, however, drops families in which neither parent received a high school to 0.4. The effect on home districts is smaller in the diploma. Virtually all of these students qualify for free second counterfactual because of the assumption or reduced-price meals.) that some of the students who live in those districts Share of Low-Income Students in the Program would not attend those districts even if the program Rising Over Time. The share of low-income did not exist. In practical terms, the effect of the students using the program has risen 5 percentage program on racial balance in most districts is small. points since 2014-15, when low-income students For example, an EDI decrease of 1.0 would be the accounted for 27 percent of program participants. equivalent of a district that is 70 percent Latino and The most notable development explaining this 30 percent white becoming 71 percent Latino and increase is that all five of the large Districts of Choice 29 percent white. Although Figure 9 shows a few increased their share of low-income students over the districts are experiencing somewhat more notable period. In addition, two smaller districts with very high increases or decreases in their EDI, these districts are shares of low-income transfer students joined the exceptions. program between 2014-15 and 2018-19. Under the More Likely Counterfactual Low-Income Students Have Relatively Low Scenario, the Statewide Effect Is Neutral. Under Participation. Figure 10 compares the share the counterfactual where all students return to of low-income students transferring through the their home districts, the average change in the EDI program with the corresponding share among all across all Districts of Choice and home districts students who attend a District of Choice or home in Figure 9 is a decrease of 0.4. This decrease corresponds to Figure 10 a minor net decrease in racial Comparing Shares of Students That Are Low Income balance overall. Under the more Share of Total, 2018-19 likely counterfactual, however, the average change across those 80% districts is zero. (We also analyzed 70 changes in the EDI based on larger 60 Statewide Average or smaller subsets of affected 50 districts. These analyses produced results similar to those discussed in 40 this section.) 30 20 Participation by Income 10 Level Participating Home District District of Choice Students on Approximately One-in-Three Transfer Students Averagea Averagea Interdistrict Permitsb Participating Students Is Low a Reflects district averages calculated prior to District of Choice transfers. Excludes students attending charter Income. The available data show schools in these districts. Percentages are weighted according to the number of students transferring in or out through the District of Choice program. that 32 percent of participating b Reflects weighted average for students transferring to a District of Choice using the transfer students came from interdistrict permit system. families with incomes low enough 16 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT district. Overall, the share of participating students especially likely to describe the January 1 deadline as who are low income is slightly lower than the a barrier to participation, whereas districts receiving average for Districts of Choice (38 percent) and fewer applications from these students saw this much lower than the average for home districts deadline as less of an issue. (67 percent). Compared with their higher-income Other Student Characteristics peers, low-income students are about half as likely to participate in the program. This difference, however, English Learners Participate at Relatively Low has shrunk somewhat over time. In 2014-15, Rates. State law requires all districts to identify low-income students transferred at only 40 percent English learners using a home language survey of the rate of their higher-income peers. As a further and an assessment of their ability to read and write comparison, we examined the share of low-income English. Students maintain their designation as students transferring through the interdistrict permit English learners until they pass an examination system. We found that 41 percent of these students demonstrating proficiency. Available data show that are low-income, which is somewhat higher than the as of 2018-19, 6 percent of the students using the share for the District of Choice program but still low District of Choice program were English learners. compared with the average home district. (For all of As Figure 11 shows, this share is notably lower these comparisons, we weighted districts based on than the average for home districts (20 percent) and the number of students transferring in or out.) somewhat lower than the average for Districts of Early Application Deadline Could Be an Issue Choice (12 percent). Although we are not certain of for Low-Income Students. We asked districts if all the factors explaining these differences, some any particular requirements of the program could portion could be attributable to differences in student be limiting participation for interested students. progress. Available data show that students attending The most common responses revolved around the Districts of Choice have above-average scores on January 1 application deadline. Several districts measures of progress toward English proficiency. To described this deadline as hindering participation in two ways. First, Figure 11 it requires students and their parents to begin thinking about Program Participation for transferring for the upcoming English Learners and Students With Disabilities school year in the fall of the previous Share of Total, 2018-19 school year. For a district that Participating Transfer Students begins school in late August, the 25% Home District Averagea January 1 deadline comes nearly eight months before the start 20 District of Choice Averagea of the school year. Interviewees Students on Interdistrict Permitsb indicated some students have 15 Statewide Average not yet begun considering their options by this point. Second, 10 the January 1 deadline usually falls during the middle of winter break. Districts indicated that 5 communicating with students and providing reminders about English Learners Students With Disabilitiesc the deadline during the break is more difficult compared to when a Reflects district averages calculated prior to District of Choice transfers. Percentages are weighted according to school is in session. Administrators the number of students transferring in or out through the District of Choice program. from districts receiving relatively b Reflects weighted average for students transferring to a District of Choice using the interdistrict permit system. high numbers of applications c District averages include school-age children only. from low-income students were www.lao.ca.gov 17 analysis full gutter AN LAO REPORT the extent transfer students experience similar gains, DISTRICT FINANCES they would transition out of English learner status In this section, we examine how the District of more quickly. Choice Program affects district enrollment levels and Students With Disabilities Participate at rates of fiscal distress. We also analyze the reduction Relatively Low Rates. State and federal laws require in funding for basic aid districts. all districts to identify students with disabilities that affect their ability to learn. The available data show Effects on Enrollment that 8 percent of the students using the District of Choice program in 2018-19 had a disability. This Some Districts of Choice Generate a Notable share compares to the district average of 9 percent Share of Their Enrollment Through the Program. for Districts of Choice and 12 percent for home Changes in enrollment are an important fiscal issue districts. (These percentages exclude students with for districts because the state allocates most funding disabilities who are outside the traditional school based on student attendance. As Figure 12 shows, age.) A small portion of this difference likely reflects participating districts generate varying shares of the fact that approximately 1-in-20 students with their total enrollment through the program. For the disabilities receives instruction in a non-district median district, 22 percent of enrollment consists setting. The most common non-district setting is a of transfer students. Very small districts (those special day class operated by a COE. (Special day with fewer than 300 students) are especially reliant classes are designed for students with relatively on the program—the median share among these severe disabilities and often employ specialized districts is 35 percent of enrollment. Many of these instructional techniques.) A few students with severe small districts indicated they joined the program to disabilities receive instruction in nonpublic schools gain economies of scale and keep their enrollment or residential facilities. For students in non-district from dropping to a level where they would not be settings, transferring from one district to another fiscally viable. For larger districts (those with more likely provides less benefit than other forms of choice than 300 students), the median share is 9 percent (such as being able to choose their specific special of enrollment. Many of these districts described day class). As Figure 11 shows, the share of students declining enrollment as the key fiscal factor prompting with disabilities transferring through them to join the program. Some districts also noted the interdistrict permit system is similar to the share for the District Figure 12 of Choice program. Participating Transfer Students Represent Equal Participation by Gender. Varying Shares of Total District Enrollment The data show that show male and Number of Districts, 2018-19 female students each accounted 16 for 50 percent of the students using the program in 2018-19. The 14 variation among individual districts Median Share = 22 Percent 12 also is relatively small. When we examined individual districts with 10 at least 50 transfer students, the 8 share of female students ranged 6 from 40 percent to 60 percent. These results are similar to 4 our previous findings based 2 on 2014-15 data. Less than 10% 10 to 20% 20 to 30% 30 to 40% 40 to 50% More than 50% Share of Enrollment Attributable to the Program 18 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT that neighboring districts had Figure 13 become less willing to approve Relatively Small Share of Students interdistrict permits than they were Transferring Out From Most Home Districtsa in the past. Number of Districts, 2018-19 Most Home Districts Experience Relatively Small 16 Changes in Their Enrollment. 14 The 45 districts participating in the 12 program receive transfer students from 196 different home districts. 10 We found that 147 of these 8 districts have less than 1 percent 6 of their students attending a 4 District of Choice. Among the other 49 districts, the share of 2 students transferring out also 1 to 2% 2 to 3% 3 to 4% 4 to 5% 5 to 6% More than 6% tends to be small (Figure 13). We identified seven districts with more Share of Students Transferring Outb than 6 percent of their students a For display purposes, excludes 147 home districts for which the share transferring out is less than 1 percent. transferring out. All of these districts b Share of total home district students attending a District of Choice as of 2018-19. Percentages are based on pre-transfer enrollment levels. had extremely low enrollments (averaging 82 students), such that a small number of students their enrollment levels and saw the District of Choice leaving accounted for a relatively large percentage program as contributing to these declines. Districts of their enrollment. Although the share of students noted that declining enrollment could require them transferring out tends to be small, home districts to close schools, lay off staff, or take other difficult typically cited the corresponding reduction in state actions to balance their budgets. Many of these funding as their chief concern about the program. districts also are affected by the interdistrict permit These districts often reported long-term declines in system, as we discuss in the box below. Districts Affected by District of Choice Program and Interdistrict Permits Interdistrict Permits Can Amplify or Reduce Changes Attributable to the District of Choice Program. We analyzed interdistrict permit transfers for the 25 home districts with the highest share of students transferring out through the District of Choice program. We found that interdistrict permit transfers often have larger effects on enrollment than the District of Choice program. Specifically, nearly half of these districts had more students transferring out through interdistrict permits than the District of Choice program. As a group, these 25 districts reported about 4,500 students transferring out through the interdistrict permit process, compared with 3,400 students transferring out through the District of Choice program. We also found that most of these districts had at least some students transferring in through interdistrict permits, in some cases offsetting a significant portion of the decrease in their enrollment. Six of these districts had net increases across both programs, meaning the total number of students transferring in from other districts exceeded the number of students transferring out. www.lao.ca.gov 19 analysis full gutter AN LAO REPORT Rates of Fiscal Distress Figure 14 Low Rates of Fiscal Distress Low Levels of Fiscal Distress for Districts of Choice and From 2014-15 Through 2019-20a Home Districts Over the Past Number of Districts in Distressb Five Years. Figure 14 shows the 5 number of districts in fiscal distress among the 25 Districts of Choice and 25 home districts with the 4 highest percentage of students Home districts transferring in or out through the 3 program. For this analysis, we identified a district as being in 2 distress if it received a qualified or negative budget rating. (State law requires COEs to review district 1 budgets at least twice per year Districts of Choice and assign qualified ratings when districts are at risk of not meeting 2014-15 2015-16 2016-17 2017-18 2018-19 2019-20 their financial obligations in the a Fiscal distress defined as a district receiving a qualified or negative budget rating from its current or upcoming two years and county office of education. b Reflects the number of districts in distress out of the 25 Districts of Choice and the 25 home districts negative ratings when districts face with highest shares of students transferring in or out, respectively. Excludes districts with fewer than immediate budget problems.) Few 100 students from both groups, as these districts are more likely to be funded with a special formula that is not tied directly to student attendance. districts in either group received qualified or negative ratings over the past five years, likely reflecting districts we interviewed indicated this reduction had the significant increases in per-pupil funding the caused them to become more cautious and reduce state provided over the period. As Figure 14 shows, the number of students they were willing to accept 4 of the 25 home districts reported distress in the through the program. As an example, one district first half of 2019-20, the highest number over the noted that if it needed to hire another teacher, the period. In their local budget documents, these cost would exceed the funding generated by its districts attributed their distress to multiple factors, transfer students. In addition, six of the nine districts including projections for slower funding growth in the that were enrolling transfer students in 2014-15 but future, rising pension costs, higher costs for special subsequently left the program were basic aid education, and declining enrollment. districts. Basic Aid Districts ACADEMIC OUTCOMES Fewer Students Transferring to Basic Below, we describe how the program can benefit Aid Districts Since State Reduced Funding. participating transfer students. After that, we review Figure 15 (see next page) shows the number some of the changes home districts have made to of students attending basic aid school districts retain and attract students. through the District of Choice program. (Due to data limitations, the numbers in Figure 15 exclude most Transfer Students students transferring from other basic aid districts.) Between 2016-17 and 2019-20, the number of Nearly All Students Transfer to Districts With transfer students attending basic aid districts Higher Test Scores. The state’s assessment system decreased by 227 students (24 percent). This decline tests students for proficiency in mathematics and began in 2017-18, corresponding to the year in English Language Arts in grades 3 through 8 and 10. which the state reduced funding. Several basic aid To compare test scores across districts, we analyzed 20 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT Figure 15 Fewer Students Attending Basic Aid Districts Through the Program Attendance Attributable to District of Choice Studentsa Per-pupil funding rate for basic aid districts reduced from 70 percent to 25 percent. 1,000 800 600 400 200 2012-13 2013-14 2014-15 2015-16 2016-17 2017-18 2018-19 2019-20 a Due to data limitations, excludes attendance generated by most students transferring from other basic aid districts. Figure 16 the share of students meeting or exceeding state performance Most Students Transferring to standards in 8th grade mathematics. Districts With Higher Test Scores (Focusing on 8th grade allows us Share of Students Meeting State Standards at to include elementary districts.) District of Choice Relative to Home District a In 2018-19, 37 percent of all students in the state met this Lower standard. The average among Districts of Choice was well above Somewhat higher the state average (55 percent), Much higher (up to 20 percentage points) (more than 40 whereas the average among home percentage points) districts was slightly below average (33 percent). (In calculating these percentages, we weighted each district to account for the number of students transferring in or out.) As Figure 16 shows, nearly all students (89 percent) transfer to districts with higher test scores than their home districts. We found Higher (20 to 40 percentage points) similar results when we analyzed transfer patterns using scores in a As measured by student performance on eighth grade mathematics assessments other grades and scores for English in the spring of 2019. Language Arts. www.lao.ca.gov 21 analysis full gutter AN LAO REPORT Nearly All Students Transfer to Districts With Career Technical Education (CTE) pathways. (A CTE Higher College-Going Rates. Many participating pathway is a broad grouping of courses organized districts believed one factor attracting students around 1 of 15 industry sectors.) First, we compared was the high proportion of their graduates who the total number of courses offered in each student’s attend college. For our analysis, we analyzed District of Choice with the number of courses offered transfer patterns using the college-going rate in the home district. As the top portion of Figure 17 for each district. The state defines this rate as (see next page) shows, the average student the percentage of students enrolling in a public transferred to a district with a few more Advanced or private postsecondary institution (including a Placement, International Baccalaureate, and foreign community college) the year after completing high language courses and slightly fewer arts and music school. For 2017-18 (the latest year available), courses and CTE pathways. Next, we examined the the average college-going rate across all districts specific subjects taught within each area. For this was 64 percent. The average among Districts of analysis, we tallied the number of unique courses Choice was well above average (79 percent), whereas taught in each student’s District of Choice but not the average among home districts was slightly above taught in the home district. For example, two districts average (67 percent). We also found that nearly all might offer the same number of foreign language students (95 percent) transfer to districts with higher courses, but one district might teach Mandarin while college-going rates than their home districts. the other teaches German. As the bottom portion of Program Provides Participating Students Figure 17 shows, in each area the average student With Additional Educational Options… During gained access to five to seven courses not offered in our interviews, we asked participating districts to the home district. We also found that the reverse is describe the academic programs they believed true—a home district generally offers some courses were attracting students. Districts frequently not available in the District of Choice. mentioned college preparatory courses (including Surveys in Other States Find Participating courses affiliated with the Advanced Placement Families Are Highly Satisfied. Although we did and International Baccalaureate programs), foreign not have the ability to survey families participating languages, and arts and music as the options most in the District of Choice program, we did review frequently sought out by their transfer students. In a surveys conducted in three other states with similar few cases, districts described students transferring to interdistrict transfer options. These surveys all found attend schools with a specific theme or instructional that more than 90 percent of families using the model. For example, one district indicated it operated program were satisfied or very satisfied with their the only high school in the area with a “community new schools. In addition, fewer than 10 percent schools” model. (This model focuses on linking of families were thinking about changing schools schools with other local resources to improve student again or returning to their home districts. A survey achievement and family engagement.) Another conducted in Minnesota asked parents to elaborate district indicated it received transfer applications on the behavioral changes they observed in their from students seeking to attend an arts and media children during the first year in their new schools. academy that emphasized hands-on training and Parents most frequently described improvements career preparation for students interested in those in self-confidence, satisfaction with learning, and fields. motivation, with about 60 percent of parents saying …Including a Different Variety of Courses their children were doing better in these areas after Compared With Home Districts. After our transferring and less than 3 percent saying their interviews, we analyzed data on the courses available children were doing worse. Administrators for the to students who transfer through the District of Districts of Choice we interviewed consistently Choice program. We focused on the types of courses reported similar results, characterizing their transfer districts mentioned most—Advanced Placement, students as happy with their new schools and International Baccalaureate, foreign languages, and generally doing at least as well academically as arts and music. We also analyzed the availability of resident students. Most administrators of these 22 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT Figure 17 Analyzing Changes in Course Availability for Participating Transfer Students Districts of Choice Offer More Courses in Some Areas, Fewer in Other Areas Average Change in Total Courses Offered From Home District to District of Choicea 2 1 Arts and Music CTE Pathwaysb Foreign Advanced International Languages Placement Baccaleurate -1 -2 -3 Districts of Choice Offer Students a Different Mix of Courses Than Their Home Districts Average Number of Unique Courses Offered by the District of Choicea, c 8 7 6 5 4 3 2 1 Foreign Advanced International Arts and Music CTE Pathwaysb Languages Placement Baccaleurate a Reflects high school-level courses. Excludes courses with fewer than ten students. b Based on the number of pathways for which the district offers at least one course. c “Unique courses” means courses offered by a student’s District of Choice but not available in the home district. CTE = Career Technical Education. www.lao.ca.gov 23 analysis full gutter AN LAO REPORT programs also cited high retention rates, with improved their performance at a faster pace, with relatively few students returning to their home the share of their students meeting standards districts. increasing by 7.6 percent over the same period. (For this analysis, we focused on the 20 home districts Home Districts with the highest shares of students transferring Home Districts Taking Action to Attract and out and at least 200 total students. Districts with Retain Students. Although home districts expressed fewer students often experience more notable reservations about the program, several responded fluctuations in their annual test results.) Although to it by taking steps to attract and retain students. we are uncertain about the effect of the District of As we described in our previous evaluation, some Choice program on these gains, the improvement districts convened community meetings to ask among home districts exceeded the improvement families about their concerns and the potential among a comparison group consisting of districts changes that could increase their satisfaction and with similar 2014-15 achievement levels and similar attract students back. In other cases, they identified shares of low-income students as the typical home the districts their students were leaving to attend and district, but no students transferring out through the studied programs that were popular in those districts. program. Districts of Choice also improved, but at a In some instances, districts discovered concerns somewhat slower rate than home districts (potentially related to their academic programs, such as a desire because their achievement levels in 2014-15 already for more college preparatory courses or greater focus were relatively high). Improvements in test scores for on science and math. In other cases, districts realized English Language Arts followed a similar pattern. We their communities wanted other changes, such also found that the gains for both home districts and as more opportunities to transfer to other schools Districts of Choice were relatively consistent across within the district. Districts made the implementation student subgroups. of changes to address these concerns a priority. Although the Figure 18 success of these efforts is difficult Home District Performance Improving Faster Than Average to measure, we found that some Median Increase in Share of Students home districts have experienced Meeting State Standards (in Percentage Points)a significant drops in the number Home Districtsb of students transferring out. For 10 Home District Comparison Groupc this analysis, we focused on Districts of Choiceb 9 the 26 home districts with the Statewide Average 8 greatest number of students transferring out in 2014-15. 7 By 2018-19, nine of these districts 6 had seen the number of students 5 transferring out decrease by at 4 least 20 percent. (We excluded 3 districts in which the reduction 2 was attributable to one of the statutory transfer limits.) 1 Home Districts Improving Mathematics English Language Arts Test Scores Over Time. Between 2014-15 and 2018-19, a Change from 2014-15 to 2018-19. the share of students meeting b Each group consists of the 20 districts with the highest share of students transferring in (Districts of Choice) or out (home districts) through the program and an overall enrollment of at least 200 students. state standards in math increased by 5.4 percent on a statewide c Consists of districts with similar 2014-15 achievement levels and similar shares of low-income students as the median home district, but no students transferring out through the program. basis (Figure 18). Home districts 24 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT PROGRAM OVERSIGHT Transfer Applications In this section, we begin by analyzing the local Districts of Choice Report Accepting Nearly audits of the District of Choice program. Next, All Transfer Applications. Based on the available we describe the reasons districts deny transfer district data, we estimate that 1,853 students applications and assess the provision of school applied to attend a District of Choice during transportation. Finally, we examine a few data the 2018-19 application cycle. We estimate districts collection issues. approved 1,650 of these applications, for an overall acceptance rate of 89 percent. (These totals exclude Audit Results approximately 600 applications withdrawn by the students who submitted them.) Disaggregating the Auditors Found No Major Issues, but Flagged data by district, we found that 25 districts approved Three Districts for Missing Paperwork. Local all applications they received, 14 districts denied auditors flagged three small districts for issues related some applications, and 6 districts received no to the District of Choice program during the 2018-19 applications. audit cycle. Specifically, two districts received audit Most Denials by Districts of Choice Related to findings because they had not registered for the Local Transfer Limits. Of the 14 districts denying program in a timely manner. One of these districts transfers, 10 indicated they had reached their locally indicated it had been unaware of the new requirement determined transfer limit. These districts ranged in to register with the state. The other district cited size, but most of them were basic aid school districts. technical difficulties, indicating it had attempted to The other four districts denying transfers cited other complete the online registration form and had been factors. Two districts issued denials because one unaware CDE did not receive its information. The of their home districts had invoked the cumulative third district received an audit finding because it did cap on transfers. (These districts explained that not adopt a board resolution establishing its local although their application materials indicated that transfer limit. CDE resolved all three findings with students from those home districts were ineligible to the districts agreeing to implement new procedures apply, they had still received a few applications from to ensure compliance moving forward. When we students residing in those districts.) Another district examined these districts in 2020-21, we found attributed its denials to applications it received after all three had submitted timely registrations and the deadline. (We understand that most districts stop adopted the necessary board resolutions. As a point processing applications after the deadline instead of comparison, auditors identified 825 instances of of reporting them as denied.) Only one very small noncompliance across all of the school districts and district indicated it had denied any transfers because program areas subject to audit in 2018-19. Even students were requesting a program the district did accounting for the relatively small number of districts not offer, and no districts denied transfers using the participating in the District of Choice program, the rule about displacement of existing students. total number of audit findings issued for the program was relatively low. Denials by Home Districts Generally Related to the Cumulative Cap. In addition to examining No Complaints Lodged With CDE. In addition the reasons Districts of Choice deny transfers, we to the audit process, we examined the new examined denials by home districts. Although the complaint procedure. CDE indicated it received state does not collect systematic data, we identified no complaints about districts failing to register or four home districts that have invoked the 10 percent refusing to submit data. Department staff noted that cumulative cap. Students who transferred before the most of the inquiries they received were for general cap applied are not affected, but additional transfers information about the program, including questions from those districts are no longer allowed. We are from superintendents exploring whether the program uncertain how many transfers would have occurred would be a good fit for their districts and questions without the cap, but historical trends suggest the from auditors seeking information about the new number could be significant. Whereas 2,500 students requirements. from these four districts were attending a District of www.lao.ca.gov 25 analysis full gutter AN LAO REPORT Choice in 2014-15, only 1,900 students were doing found only minor differences. Transfer students so as of 2018-19. The decline among students receiving transportation were 59 percent white and transferring from these districts appears to be a key 33 percent Latino, whereas students not receiving reason for the overall drop in program participation transportation were 55 percent white and 36 percent over the period. We also found that home districts Latino. Finally, we examined whether districts were rarely invoke any of the other statutory reasons for more likely to transport students from particular home limiting transfers. districts. We found that only two of the seven districts had a significant number of transfer students from School Transportation multiple home districts. In both districts, rates of Relatively Few Districts of Choice Provide transportation were similar for students from each of School Transportation. Available data show the home districts. that 11 of the 45 Districts of Choice provided Data Reporting transportation for their transfer students. All of these districts are small and rural, with an average Some Smaller Districts Interpreted State enrollment of about 300 students each. Together, Reporting Instructions Differently. To obtain they provided transportation for 290 (3 percent) of information on the socioeconomic and demographic all students participating in the program in 2018-19. characteristics of transfer students, CDE asked Although very few transfer students receive districts to flag those students in CalPADS. Most transportation, the average for other students also districts flagged all transfer students enrolled as is low. According to a federal survey conducted of 2018-19, regardless of the year in which those in 2017, only 9 percent of students in the state students first transferred. Several districts flagged receive district-provided transportation. The share is only a subset of these students—generally those even lower in urban areas, where the larger Districts who transferred for the first time in 2018-19. of Choice are located. Consistent with these findings, Regarding transfer approvals and denials, most most districts we interviewed indicated they did not districts provided information for the 2018-19 provide transportation for resident or transfer students. Figure 19 Disadvantaged Students Somewhat More Likely to Disadvantaged Students Somewhat Receive Transportation. One More Likely to Receive School Transportation reason the Legislature required Share of Total, 2018-19 districts to report transportation 50% data was to determine whether Low-Income Students districts were providing this service English Learners on an equitable basis. Overall, we 40 Students With Disabilities found that seven districts provide transportation for some of their 30 transfer students and four districts provide transportation for all transfer students. Focusing on the 20 districts providing transportation only to some transfer students, 10 we found that low-income students and English learners were somewhat more likely to Receiving Transportation Not Receiving Transportation receive transportation than other students (Figure 19). When we a Most Districts of Choice do not provide school transportation to any transfer students. The percentages in this figure reflect the average for the subset of students transferring to districts analyzed the data by race, we that transport some of their students. 26 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT application cycle. Several districts, however, provided same topics that districts are required to include information on applications and denials over several in their local reports. In both cases, districts are years. Although these differences introduce some providing information about the number and potential error in our analysis, the districts reporting characteristics of students who transfer through the data differently account for a relatively small share of program. We also found that the format and content students using the program. In addition, we adjusted of these local reports can vary widely. For example, district data to provide for consistent comparisons in some districts provide extensive detail in their local this report wherever possible. reports, whereas others provide highly summarized State and Local Data Reporting Requirements information. Due to these differences, the local Overlap. We found that the program information reports are much less useful for making comparisons collected and published by CDE largely covers the among districts than the data collected and published by CDE. ASSESSMENT In this section of the report, we review the main reductions in the number of students transferring out strengths of the program. We then highlight a few over the past several years (excluding districts with policies limiting access for students and discuss drops related to the cumulative cap) suggests that differences in participation by student subgroups. students can respond positively to these efforts. In addition, we found that students who remain in their Strengths of the Program home districts continue to make gains, with home Provides Students With Additional Educational districts improving their test scores faster than most Options. Students can benefit from the District of other districts. Choice program in several ways, including being able Program Has Met or Is Making Progress to take classes not offered in their home districts Toward the Legislature’s Goals. The Legislature and attend programs that align with their personal made several notable changes as part of interests and educational goals. Overall student the 2017 reauthorization of the program. Our satisfaction with the program seems high, with understanding is that these changes broadly were districts describing their transfer students as doing intended to (1) add transparency, by making program well academically. Although the program is relatively data readily available; (2) enhance accountability, small, it offers students a few advantages compared by creating a mechanism to monitor compliance with the interdistrict permit system. Most notably, with core program requirements; and (3) increase it allows students to transfer even if their home access, by promoting the participation of students districts would not approve a permit and ensures who transfer at lower rates. Our findings suggest students can remain in the program until graduation. the program has met the first two objectives and is Finally, the program seems to have relatively broad making progress toward the third, as we discuss appeal, attracting students from a range of racial below: backgrounds, income levels, and geographic regions • Transparency. The data posted by CDE of the state. provides detailed information about the number Encourages Home Districts to Improve Their of participating districts, the characteristics of Programs. Despite their reservations about the transfer students, and the outcome of transfer program, some home districts took steps to address applications. Moving forward, this information the reasons students were leaving their districts. will allow the Legislature and the public to These steps included learning more about the monitor and assess changes in the program priorities of their communities and adjusting their more easily. instructional options accordingly. Our finding that about one-third of home districts saw significant www.lao.ca.gov 27 analysis full gutter AN LAO REPORT • Accountability. The new audit procedure Early Application Deadline. The found no systemic problems or districts January 1 deadline limits the program to students discriminating against interested students. The who plan to transfer well in advance of the school few issues identified by the auditors seem to year. Some districts believed low-income students reflect unintentional oversights or districts being were particularly affected by this deadline because unfamiliar with the new requirements. The new they had lower levels of awareness about the complaint process administered by CDE also program. Given that many English learners also are did not result in any investigations or penalties. low-income students, we think the January 1 deadline • Access. Our report reflects data for 2018-19— might also contribute to lower participation for these the first school year after the state implemented students. We are aware of no other educational the 2017 changes. Although the full effects of program in which the law requires students to apply these changes will not become clear for several as early as January 1. As a point of comparison, more years, recent trends seem promising. we understand that most charter schools set their Most notably, the share of low-income students application deadlines during late winter or spring. using the program is up 5 percentage points Funding Reduction for Basic Aid Districts. Our compared with 2014-15. The share of Latino findings indicate that the lower funding rate for basic students is up nearly 8 percentage points. aid districts has reduced the number of students these districts are willing to accept through the Policies Limiting Access program. Although basic aid districts account for only about 10 percent of students in the program, Cumulative Cap. The data show that their participation is notable for at least two reasons. nearly 1,900 (20 percent) of the students using the First, students transferring to basic aid districts are program in 2018-19 came from districts that have somewhat more likely to be disadvantaged than reached the 10 percent cumulative cap. These other students. For example, 11 percent of students students are 40 percent Latino, 40 percent Asian, and transferring to basic aid districts are English learners, 13 percent white—roughly resembling other students compared with 5 percent of students transferring using the program. As these students graduate to other districts. Students transferring to basic aid and are not replaced by future cohorts, overall districts also are slightly more likely to be low-income participation in the program is likely to decrease. This or have a disability. Second, nearly all of the basic decrease could accelerate over the coming years if aid districts participating in the program are located additional districts reach the cap. We are concerned in rural areas, where fewer educational options are that the cumulative cap has no connection to readily available. To the extent these districts leave student interest or the quality of local transfer options the program or adopt lower transfer limits, access for available. Once reached, the cap disallows all future these students would diminish. participation, including transfers by students whose participation the Legislature has tried to encourage. Program Participation Levels In addition, the original justification for the cumulative Available Data Does Not Reveal Why Usage cap has weakened over time. Our understanding is of the Program Varies Among Student Groups… that the Legislature included the cumulative cap in the The information collected by the state provides a first version of the program to prevent home districts detailed picture of who uses the program, but it does from experiencing sizeable drops in enrollment over not indicate why students from different subgroups several consecutive years (and losing the associated are more or less likely to use the program. Many per-student funding). The cumulative number of explanations are possible. For example, these transfers since 1993, however, is not a good indicator differences could reflect varying levels of awareness of a district’s current fiscal condition. In addition, about the program. Even though the Legislature the Legislature added a separate provision in 2009 added some new communication requirements allowing districts to limit transfers that would worsen in 2017, the effects of these changes might not be their fiscal distress. evident for several years. Alternatively, students 28 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT might differ in their satisfaction with their home …But Districts Typically Can Accommodate districts. Many home districts offer unique courses All Interested Students. A key finding from our or academic options, and some students might evaluation is that most districts participating in the prefer those options. Our evaluation, however, found program are able to accept all students interested similar subgroup differences within the interdistrict in transferring. In other words, high usage of the permit system. Specifically, white students and Asian program by some students generally does not students use interdistrict permits relatively frequently, translate into fewer transfer opportunities for other whereas Latino students use them less frequently. interested students. In cases where a district receives Low-income students, English learners, and students more transfer applications than it can accept, the with disabilities also use interdistrict permits less program provides priority for low-income students. frequently. These findings suggest that the underlying (The cumulative cap is a notable exception—because causes of these participation differences probably are the cap establishes a maximum number of transfers not unique to the District of Choice program. over the life of the program, high usage by some students can result in fewer transfer opportunities for other students.) RECOMMENDATIONS Below, we provide our recommendations regarding we think the coronavirus disease 2019 pandemic and the extension of the program. We also make several associated school closures could have anomalous recommendations intended to promote access to the effects on student participation in 2020-21 and program. 2021-22. The data for these years might not be Reauthorize the Program. We recommend the particularly helpful for assessing program trends.) Legislature reauthorize the program, potentially on a Repeal Cumulative Cap. Related to our permanent basis. Among other considerations, the recommendation to reauthorize the program, we program provides quality educational options for a recommend repealing the cumulative cap. Repealing range of students, the new oversight mechanisms the cap would allow students to use the program uncovered no notable concerns, and the share of regardless of the previous transfer activity in their disadvantaged students using the program has risen. districts. Leaving the cap in place, by contrast, seems Ending the program, by contrast, would be disruptive likely to reduce program participation over the next for students who currently use the program and several years. Even without the cap, districts could would deny future transfer students the opportunities limit transfers that exceed 3 percent of their annual that have benefited previous users of the program. enrollment or notably worsen fiscal distress. If the Regarding the length of reauthorization, the program Legislature remained concerned about the immediate currently is operating on its sixth temporary extension effect on home districts, it could phase out the cap since 1993. A permanent reauthorization would gradually. provide greater predictability for participating Allow Later Application Deadline. We think a students and districts. Ongoing oversight could later application deadline would make the program still take place—the state could continue collecting more accessible to students. Although the Legislature program data, and the Legislature could have future could consider many possible dates, one option oversight hearings to ensure the program continues would be moving the deadline from January 1 to to meet its goals. Alternatively, if the Legislature were March 1. This change would provide students with an to adopt another temporary extension, we would additional two months to learn about and consider recommend a minimum of five years. We think a their transfer options. To facilitate this change, the shorter extension would not provide the Legislature Legislature could repeal the requirement for districts with enough new data before it would be required to to provide students with a preliminary notice of their revisit the program. (Related to this recommendation, acceptance or denial by February 15. The state could www.lao.ca.gov 29 analysis full gutter AN LAO REPORT maintain the May 1 deadline for Districts of Choice to on the number and characteristics of students develop their final list of accepted students. (Although who transfer through the program. Ongoing data a later deadline would provide less advance notice collection would help the Legislature assess for home districts, the state could still structure the effects of the 2017 changes as well as our the program to provide home districts with a list of recommendations. We think CDE can address the transfer students in the spring, well in advance of differences in how districts reported their data without the school year. In addition, the policy of funding all additional legislation, but we recommend asking the districts according to the higher of their current- or department for an update on these efforts during prior-year attendance tends to mitigate the effects the next hearing on the program. The current data of unexpected changes in the number of students collection, however, does not help the Legislature transferring out each year.) better understand why some groups of students Increase Funding for Basic Aid Districts and transfer at higher rates than others, including Adjust the Rate for Disadvantaged Students. the differences by race. If the Legislature were We recommend setting a higher funding rate for interested in understanding how families make these basic aid districts and modifying this rate to mirror decisions, it could consider funding a survey. This the Local Control Funding Formula. Specifically, we survey could ask families about their (1) awareness recommend setting the rate at 60 percent of the of transfer options, (2) motivations for transferring formula’s base amount, increasing to 80 percent or not transferring, and (3) overall satisfaction with for low-income students and English learners. This their districts. We think this survey would be more approach would mirror the 20 percent differential valuable if it focused on a range of choice options for students attending other districts. More instead of the District of Choice program by itself. For importantly, it would reward districts for attracting this type of research, the state typically tasks CDE and retaining these students. On average, the new with overseeing a contract and soliciting bids from rate would generate similar amounts of funding as qualified organizations. The findings from this survey the 70 percent rate for all students that existed prior could inform future efforts to promote participation by to 2017. (We estimate that low-income students students who use the program less frequently. and English learners account for about half of all Repeal Redundant Reporting Requirements. students transferring to basic aid districts.) From a We recommend repealing the requirement for state fiscal perspective, this increase likely would districts to prepare their own annual report on the involve short-term costs and long-term savings. program. These reports have significant overlap The short-term costs—likely around $2.5 million per with the information published by CDE, which is a year for the next few years—would result from the more reliable source of data for the program. We state providing a higher rate for students who have also recommend repealing the requirement for already transferred and currently generate funding at participating districts to notify adjacent districts of the 25 percent rate. Long-term savings, by contrast, their intent to remain in the program every year. Under would emerge to the extent basic aid districts accept the registration requirements the state added in 2017, new students who otherwise would attend their home CDE maintains and publishes a list of participating districts and generate the regular rate. For example, districts. The law also requires participating districts if the higher rate caused these districts to accept an to register with their county boards of education, additional 500 students, the long-term state savings which can disseminate this information to nearby would be around $1.5 million per year. districts that might be affected. Eliminating these Continue Collecting Data, Consider Funding requirements could allow districts to spend more time Survey to Learn More About Transfer Decisions. on outreach and other activities promoting access to We recommend the state continue collecting data the program. 30 LEGISLATIVE ANALYST’S OFFICE analysis full gutter AN LAO REPORT CONCLUSION The Legislature created the District of Choice The state finally obtained comprehensive data program in the early 1990s during a period of in 2018-19, which provided evidence the program is experimentation with new forms of educational making progress toward the Legislature’s more recent choice. Our first evaluation of the program, published goals. In this report, we recommend reauthorizing in 2016, found evidence the program was providing the program and making changes to address some quality options for students but noted that limited of the factors limiting participation. We think our data and vague oversight made the program hard to recommendations would build upon existing progress assess. A year later, the Legislature made a number and promote access for all students. of changes to address these and other concerns. www.lao.ca.gov 31 analysis full gutter AN LAO REPORT LAO PUBLICATIONS This report was prepared by Kenneth Kapphahn, and reviewed by Edgar Cabral and Anthony Simbol. The Legislative Analyst’s Office (LAO) is a nonpartisan office that provides fiscal and policy information and advice to the Legislature. To request publications call (916) 445-4656. This report and others, as well as an e-mail subscription service, are available on the LAO’s website at www.lao.ca.gov. The LAO is located at 925 L Street, Suite 1000, Sacramento, CA 95814. 32 LEGISLATIVE ANALYST’S OFFICE