Legislative Analyst's Office

Analysis of the 2000-01 Budget Bill

---

Teacher Quality and Supply

The 2000-01 Governor's Budget devotes significant attention to (1) expanding existing programs; and (2) creating new programs aimed at increasing the supply of, and improving the professional preparation of, teachers for the state's K-12 schools. The budget approaches this area with an array of programs spread across many state entities--including the University of California (UC), California State University (CSU), Commission on Teacher Credentialing (CTC), Student Aid Commission (SAC), State Department of Education (SDE), Secretary for Education, and even the Department of Housing and Community Development (HCD). These programs can be divided into three basic groups:

• Programs to recruit/retain credentialed teachers, targeted toward "low-performing" schools.
• Programs to recruit/retain credentialed teachers generally in the public schools.
• Staff development.

Figure 1 lists the Governor's proposals, categorized on the above three purposes. The figure indicates the degree of specificity with which resources are proposed for allocation--through 21 state categorical programs, including 12 new ones. What the figure cannot convey is the degree of specificity in terms of programmatic detail and state control. Generally, the administration's approach leaves local school districts and school sites without meaningful discretion over the allocation or deployment of these significant fiscal resources. We believe the budget's approach will limit the effectiveness of the dollars provided. We consider this problem further in the discussions below, which are organized in the same three groupings used in Figure 1.

Recruitment/Retention of Teachers in Low-Performing Schools

As shown above in Figure 1, the budget proposes five new programs--and expands another program--in a significant effort to recruit and retain teachers in low-performing schools. Below, we provide background on the problem that this major budgetary initiative attempts to address.

The Nature of the Problem

Over the last few years the number of teachers in California's public schools lacking a teaching credential has risen dramatically. In 1995-96 approximately 16,000, or about 7 percent, of the "teachers of record" in California's public K-12 schools (that is, assigned as the regular teacher for a classroom) lacked a teaching credential. By 1998-99, that number approximately doubled to about 32,000, or 11 percent, of the teaching force. This number includes 28,500 teaching under an "emergency permit" issued by the CTC, and the rest teaching under various waivers issued by the CTC.

Schools that face extra challenges in attracting and retaining qualified staff--which tend to be schools in poor neighborhoods and, to some extent, rural areas--have been especially hard hit. A recent definitive study of this problem prepared by SRI International (SRI) found extraordinarily high percentages of noncredentialed teachers in such schools. For example, SRI found that, on average, 16 percent of the teaching staff lacked credentials in schools where 75 percent or more of the student body received free or reduced-price lunches (an indicator of low family income). In contrast, 4 percent of the teaching staff lacked credentials in schools where 25 percent or less of the student body received free or reduced-price lunches. At many schools the percentage of teachers without credentials was significantly higher, in extreme cases exceeding 50 percent.

Many factors contribute to the extra challenges that schools serving poor--and in some cases--rural populations have faced in teacher recruitment and retention. Probably the single greatest factor in the recent increase in noncredentialed teachers in these schools, however, has been K-3 class size reduction (CSR)--or, to be precise, the rapidity and near-universality of CSR implementation across the state. The CSR Research Consortium in a multiyear study funded by state, federal, and private funds found that in 1995-96, just before implementation of K-3 CSR, there was little difference between the percentage of noncredentialed K-3 teachers in schools serving the poorest quartile of pupils (2 percent) and schools serving the most affluent quartile (0.5 percent). Two years later, the share of K-3 teachers lacking credentials at schools serving the poorest pupils jumped to almost 20 percent, while the share of K-3 teachers lacking credentials at schools serving the most affluent pupils rose to less than 5 percent. (Please see Figure 2.) The state's rapid implementation of CSR not only stressed the supply of credentialed teachers on a systemwide basis, it created a strong differential dynamic. Schools in affluent areas were able to meet the challenge not only by recruiting new credential holders, but by drawing credentialed teachers away from schools in poor areas with the offer of attractive classroom settings (only 20 pupils) in attractive schools. In contrast, schools in many poor areas saw their recruiting efforts essentially collapse, in some cases unable to hire any credentialed teachers.

We note that teaching credentials are not necessarily synonymous with teaching quality. The ability to inspire children and teens to learn rests on a mix of personal qualities, knowledge, and energy that are not guaranteed by a document issued by the state. There are individuals teaching on emergency permits who have these attributes, just as there are individuals with credentials who don't. What the credential does certify is that the individual has received at least a basic level of preparation in subject matter and pedagogy. The above observations are relevant to an important finding in the SRI study--that many relatively "well off" districts and schools hire teachers on an emergency permit basis for strategic reasons, while the districts or schools with the most serious recruitment challenges hire emergency permit holders out of desperation.

When districts or schools hire emergency permit holders strategically, they are selecting highly promising candidates, often moving them through an internship program that will soon result in the teacher becoming credentialed. In these cases, the "problem" of noncredentialed teachers isn't really a problem at all. At the other extreme of desperation hiring, however, the new teachers are less promising candidates. (Some teachers in these schools have failed the state's basic skills exam for teachers [C-BEST] and are teaching under a waiver.) Thus, as outlined above, the SRI study found both quantitative and qualitative differences in the situations faced by schools that are "advantaged" in terms of attracting and retaining teachers and schools that are "disadvantaged" in this regard.

One of the key findings of the study was that there is a "tipping point," beyond which the daily situation faced by schools becomes essentially unmanageable. The following passage from the study describes this essential point (The Status of the Teaching Profession, pages 31 and 32):

As the percentage of underqualified teachers in a school increases, serious problems begin to arise. In schools with few underqualified teachers, these teachers tend to be hired for some desirable characteristics, and they are supported by a large portion of veteran teachers. . . At some point, as the percentage of underqualified teachers grows, [a] school's overall functioning is impaired. The exact point at which this occurs depends on the strength of professional development and induction systems in the district and at the school. . .However, we observed in case studies that schools with more than 20 percent underqualified teachers were hard pressed to provide adequate professional support to their entire faculty. These are the hard-to-staff schools in the state, those where a child's opportunities to receive the kind of instruction needed to meet the state standards are severely compromised.

The SRI study concluded that one in five public schools in the state falls across this tipping point where a manageable situation crosses into dysfunction. Thus, for approximately 80 percent of the state's schools, the problem of attracting and retaining teachers is being managed more or less successfully. Approximately 20 percent of the state's schools face a severe problem that calls for state attention.

Concerns With Governor's Proposals

The Governor's proposal targets schools scoring in the lower half of the state's Academic Performance Index (API). Although this approach recognizes that there is a relationship between low test scores and schools facing extraordinary recruitment challenges, the chosen target is far too broad for the problem at hand. The administration's approach also concentrates almost all control over the resources and operating details of these programs with state agencies rather than school district and school site administrators. We discuss each of these concerns below, as well as recommend an alternative approach to the problem.

Misplaced Target. As noted above, the available research indicates that the real problem of too few credentialed teachers is concentrated in about 20 percent of the state's public schools. These are the schools where the systems for providing professional mentoring and support have been overwhelmed by the imbalance between veteran and novice teachers. These are also the schools that face the most serious problems in terms of poor academic performance.

Although the administration's various proposed teacher recruitment incentives are targeted to low-performing schools, the administration's definition of performance is overly broad. Under the Governor's various proposals, any school scoring below the 50^th percentile on the state's API is considered low performing. Thus, by definition, half of the state's schools are "low-performing" and would qualify for the targeted recruitment/retention incentives. (Because the API is a California-only measure, this definition of performance means that 50 percent of the state's schools would always be low-performing no matter how much the state's schools improve over time.) This definition is overly broad and has no relationship to the problem that the proposals seek to address. This misplacing of the target not only would dissipate large amounts of state funds, it could leave the fundamental problem largely unaddressed. By offering the same incentives for teaching at a relatively attractive school as for teaching at a school in distress, the incentives would divert qualified teachers away from the schools that need them the most.

Misplaced Discretion. As noted above, the administration's various incentive programs leave the people who must manage teacher recruitment and retention problems on a daily basis--school district and school site administrators--without meaningful discretion over either the allocation of resources or their application. For example, the administration's $50 million housing down-payment program for teachers, to be managed by HCD, would provide payments to individuals on a first-come, first- served basis. The proposal completely bypasses local school officials, who would have no way of assuring that incentives reach the individuals that they need to hire. In fact, under the program's design, the state could grant a $10,000 payment to a teacher for committing to teach for the next five years at a school that does not wish to retain that teacher. (For a more detailed discussion of the administration's housing incentive proposal, please see our analysis of the HCD budget [Item 2240] in the General Government chapter of this analysis.)

This bypassing of the people who must make the day-to-day decisions for operating local schools is a problem running through several of the administration's proposals in this area, including proposed awards for individuals attaining National Board certification and the new Credentialed Teacher Recruitment program to be administered by the CTC. The regional recruitment centers proposal ($9.4 million under Item 6110-229-0001) has additional problems, including (1) being poorly conceived and defined; and (2) potentially duplicating other state and local efforts already underway, including those of the California Center for Teaching Careers (Cal Teach) established by the Governor and Legislature last year.

Analyst's Recommendation

In order to maximize the effectiveness of state efforts to improve teacher recruitment at the state's hard-to-staff schools, we recommend that the Legislature redirect the monies for four programs for low-performing school teacher recruitment--a proposed total of $122.3 million--into a block grant to school districts for teacher recruitment, retention, and support targeted to the schools most in need of this help. We recommend targeting schools where at least 20 percent of the teaching staff lack credentials. For these targeted schools, we recommend that annual decreases in the percentage of teachers without credentials be a condition of ongoing receipt of the block grant.

We have noted above the two fundamental problems in the Governor's proposals to recruit teachers to low-performing schools:
(1) the overly broad set of schools to be targeted and (2) the lack of local discretion. In our analysis of the CSU budget request, we recommend moving the $3.5 million Teaching Fellowship program to SAC, and recommend adjusting the target set of schools for which fellowships apply to schools where at least 20 percent of the teaching staff lack credentials. In our analysis of the SAC budget, we recommend adjusting the target set of schools for the Assumption Program of Loans for Education to the same schools.

We recommend that the Legislature redirect the monies for the administration's other four programs for low-performing school teacher recruitment--a proposed total of $122.3 million in the budget year--into a block grant to school districts for teacher recruitment, retention, and support. Again, we recommend targeting to schools where at least 20 percent of the teaching staff lack credentials. For these targeted schools, we recommend that annual decreases in the percentage of teachers without credentials be a condition of ongoing receipt of the block grant. Under our recommended approach, school districts and school sites would have broad discretion to pick from an extensive menu of options for recruiting and retaining credentialed teachers, including the provision of mentoring and other means of supporting the development of new teachers.

Recruitment and Retention of Teachers Generally

The Governor's budget makes six proposals to address teacher recruitment and retention in a nontargeted fashion. Five of these are shown in Figure 1. The sixth concerns the State Teachers' Retirement System's (STRS) earnings cap for retirees. These proposals have a combined General Fund cost of $48.8 million in the budget year.

In three cases--the CalTeach, the Alternative Certification program expansions, and a new initiative to provide Financial Crisis and Management Assistance Team (FCMAT) services to school districts--we recommend approval of the changes as proposed:

• The budget provides $20.8 million from the General Fund to expand the Alternative Certification program to 5,400 additional participants and increase grant amounts. This program provides an alternative route for individuals entering teaching as a second career and, as such, provides an additional source of qualified teachers.
• The budget proposes $9 million in ongoing funds to expand CalTeach, a teacher recruitment program using print, television, radio, and Internet media. Early evidence suggests that CalTeach's initial efforts have been effective and further investment appears warranted.
• The budget proposes $1 million to provide FCMAT technical assistance to school districts to streamline teacher hiring practices. We think this is a worthwhile investment that should help to lower barriers that keep some candidates from joining the teaching ranks in public schools.

Our analysis finds that the proposed elimination of the Cal Grant T program, administered by SAC, would have negative consequences on teacher recruitment, especially among low-income teaching candidates. More discussion of the Cal Grant T program and the likely consequences of its elimination for statewide recruitment efforts is provided in our SAC analysis, later in this chapter.

In this section, we examine the two remaining proposals. We find that the budget's proposal to provide $15 million in awards to teachers earning National Board certification lacks sufficient justification. Instead we recommend providing these resources to school districts through block grants to allow greater local discretion over the allocation of resources. While we support the proposal to lift the STRS earnings cap for retirees returning to teach in public schools, we think one modification would increase the program's effectiveness. We discuss both the National Board Certification and STRS proposals below.

National Board Certification

We recommend the Legislature delete $15 million proposed for awards for National Board certified teachers because it is unclear that these awards would lead to improved teacher quality. Instead, we recommend that the $15 million be included in staff development block grants for districts.

The National Board for Professional Teaching Standards certificates teachers who demonstrate substantial achievement in subject matter knowledge, teaching skill, school leadership, and community participation. Applicants for certification undergo a rigorous year-long review based on videotaped classroom sessions, written exercises, oral interviews, and student work samples.

The budget proposes $10 million to provide $20,000 awards to National Board certified teachers who commit to teaching for four years in California public schools ranked below the 50^th percentile on the API. The one-time award would be given in a single payment at the time a teacher commits to teaching in one of these schools for at least four years. In addition, the budget proposes $5 million to provide one-time $10,000 awards for teachers employed in a California public school who, though recipients of these awards, would not be required to make prospective teaching commitments.

Enforcement Costs Could Be High. The program would require recipients of the $20,000 awards to repay their awards if they failed to honor the four-year teaching commitment. In the event recipients fail to repay their awards, the program would require the CTC to take steps to recover the funds, including initiating administrative and civil procedures. Depending on the number of teachers failing to complete their teaching commitments, these procedures could prove burdensome and costly.

Effect on Teacher Quality Is Debatable. While National Board certification is certainly an honor, it is less clear that the certification process improves an individual's teaching abilities. It is likely that many of the individuals who attain certification would have been superior teachers even in the absence of the certification process. However, it is possible that rewarding teaching excellence may work to recruit and retain highly qualified teachers.

Let Districts Decide if Rewarding Certification Improves Their Teaching Corps. Districts are best able to determine if providing awards for National Board certification is a cost-effective way to improve teacher quality in their local schools or whether other approaches would be better. For this reason, we recommend that $10 million be redirected to our recommended Hard-to-Staff Schools block grant and $5 million be redirected to our recommended Nontargeted Staff Development block grant. (We discuss these block grants in detail elsewhere in this section.)

Lift STRS Earnings Cap

We recommend that the Legislature enact legislation adopting the Governor's proposal to remove the statutory earnings limitation for current retirees of the State Teachers' Retirement System in order to induce retirees to reenter the teaching workforce. We further recommend that the Legislature expand coverage of this proposal to include retirees returning to employment as principals and vice principals.

Current law requires a dollar-for-dollar reduction in pension for each dollar earned by a STRS retiree from employment in a public school, above an annual cap of $19,050. The budget proposes that the Legislature enact legislation to remove this earnings cap for STRS retirees who retired on or before January 1, 2000, and who return to work in public K-12 classrooms as teachers or as support providers for teachers in the Beginning Teacher Support and Assessment, alternative certification or Paraprofessional Teacher Training programs. Because the retirement date cutoff is in the past, there is no threat that this change would create an incentive for current employees to retire in order to take advantage of the cap removal.

Difficult to Estimate Participation Rate. Due to the many factors involved, it is difficult to estimate how many additional teachers would be enticed from retirement based on the proposed changes to the STRS earnings cap. In an action very similar to that being proposed, Chapter 965, Statutes of 1998 (AB 2765, Honda), lifted the STRS earnings cap for retirees returning to fill teaching vacancies caused by grades K-3 CSR. Under that statute, between 1,000 and 1,500 teachers returned to public school classrooms during the 1998-99 school year. While this gives an indication of the possible magnitude of effect, differences between the two cases, including the categories of eligible teaching positions, prevent that earlier experience from being an accurate estimator of the number of retirees who would return to teaching under the current proposal.

Program Should Be Extended to Cover Principals. In addition to shortages of qualified teachers and teacher support providers, many districts across the state face a shortage of fully qualified applicants for principal positions. The importance of principals to school success has always been critical. The state's new accountability framework places even greater expectations on principals. Thus, the state's public schools need qualified, experienced individuals to serve as principals and vice principals. Accordingly, we recommend that the removal of the STRS earnings cap be extended to include principals and vice principals.

Teacher Professional Development

The 2000-01 Governor's Budget proposes $156.3 million in General Fund spending for new programs and program expansions for professional development for teachers and school site administrators, as shown in Figure 3. This is an increase of $119.3 million over current-year funding.

These programs would be administered by UC in coordination with CSU, CCC, independent colleges and universities, and SDE. Eligibility criteria vary among the programs, but most would target teachers from low-performing schools as defined by pupils' Standardized Testing and Reporting (STAR) scores and number of beginning teachers (see Figures 4 and 5 for program details).

All of these programs, with the exception of the New Teacher Center, are professional development courses based on a model developed by the California Subject Matter Projects (CSMP), a UC-based teaching professional development network. This model involves training school site teams comprised of new teachers, experienced teachers, and administrators. The training consists of an intensive one-week session, typically located on a college or university campus, followed by shorter sessions held each month at the school site during the academic year.

The Governor proposes funding the institutes in both the UC and SDE budgets. The UC budget (Item 6440-001-0001) contains $97.3 million (including $71.3 million in new funding) for the various programs. The SDE budget (Item 6110-135-0001) contains $59 million in Proposition 98 funds (including $48 million in new funding) to pay teachers stipends of $1,000 to $1,500 each for attending the institutes.

Concerns With the Proposal

These proposals represent a significant expansion of state programs for K-12 staff development. While we recognize that school districts need more staff development resources, we outline below our concerns with the Governor's proposals and then offer an alternative approach.

Proposals Micromanage K-12 Staff Development. Staff development needs vary among schools and among local school districts. Accordingly, school districts need flexibility to choose the types of staff development programs that best meet their particular needs. Under the Governor's proposal, new professional development funding would be available only for training provided by the UC-administered institutes, only in the specific subject areas and grade levels offered, and only in the funding portions allocated "from the top down." Figures 4 and 5 illustrate the degree of specificity proposed. There are many other staff development opportunities that school districts may view as high priorities, but the new funding would not be available for these purposes. In our view, this degree of state-determined specificity will reduce the programmatic effectiveness of the expenditures.

Inefficient Allocation of Resources. The Governor's proposal allocates new staff development resources across eight specific training programs. To the extent that the Governor's allocations fail to reflect the needs of individual districts, the new resources are not efficiently allocated. This problem would be avoided by providing funding directly to school districts in a block grant, allowing them to purchase staff development services based on local priorities.

Monopoly Provider. Under the Governor's proposal, the UC-administered institutes would operate as a virtual monopoly provider of these staff development programs to school districts. This arrangement could reduce pressure on the UC institutes to improve program quality or service delivery because they would not compete for school districts' business. In addition, by keeping other staff development providers out of the market, the proposal would limit the training options available to school districts.

Expansion Is Too Rapid. The Governor proposes increasing the number of educators trained at the institutes from 22,500 in 1999-00 to 93,500 in the budget year, more than quadrupling their output. The administration's proposal does not make clear how UC would accomplish a one-year expansion of this magnitude without undermining training quality. Our concerns are amplified by the fact that all the institutes rely, in part, on the intellectual and administrative capacity of the CSMP. In our view, such a rapid expansion risks overtaxing the CSMP infrastructure and undermining program quality.

Low-Performing Indicator Rewards Low Performance. The Governor proposes targeting most of this investment in staff development to low-performing schools, defined as schools whose pupils on average score at or below the 40^th percentile on the Stanford-9 exam. Using a STAR-based mechanism to allocate funding rewards schools for low performance, running counter to the principles of a well-designed accountability system. A more appropriate targeting would be toward disadvantaged schools--those facing additional challenges due to demographic factors. To do this, we suggest using a socioeconomic targeting mechanism, such as the number of pupils eligible for free and reduced-price lunches. While the two indicators are highly correlated, our targeting method would compensate schools for actual challenges faced at the school site level rather than reward low performance itself.

For the above reasons, we recommend deleting $119.3 million proposed in new funding for the UC-administered institutes. Instead, we recommend providing these funds directly to districts through staff development block grants. (Our detailed recommendations for two staff development block grants are outlined below.)

Staff Development Block Grants

We recommend that the Legislature provide $149.3 million from the General Fund to school districts for staff development training to address teacher quality training needs. Of this amount, we recommend that the Legislature target $100 million to school districts facing demographically based challenges and provide $49.3 million in untargeted funding available to all districts.

Given our concerns, noted above, we cannot recommend that the Legislature approve the Governor's proposal to expand the UC-administered institutes for K-12 staff development. Below, we present an alternative approach to funding increased staff development.

Provide Flexible Funds Through Block Grants. To provide resources for districts to address their unmet staff development needs, we recommend that the Legislature provide $149.3 million for two staff development block grants. Figure 6 shows the sources for these funds and our recommended allocation among the two block grants.

The grants would be funded from three reductions we are recommending to budget proposals:

• $119.3 million for UC Professional Development Institutes (as described above).
• $25 million in one-time funds for CSU information technology staff development (as described later in this section).
• $5 million for National Board certification awards (as described earlier in this section).

Districts would have broad discretion in using funds in the following two block grants to address their staff development needs, including contracting with the UC institutes, or whichever provider can best meet those needs.

• Needs-Based Staff Development Block Grant ($100 million). This block grant would provide additional staff development resources to schools facing serious challenges associated with educating poor and limited-English-proficient children. We recommend that, of the funds provided in this block grant, $90 million be allocated to school districts based on a socioeconomic poverty indicator and $10 million be allocated to school districts based on the number of English-language learners. This split of resources between disadvantaged schools ("low-performing"in the Governor's proposal) and English-language learners approximately mirrors the distributions proposed in the Governor's budget. However, in our proposal, districts would have flexibility and appropriate operational control over how funds are used. Also, our targeting mechanism for disadvantaged schools would recognize district needs rather than reward low test scores.
• Nontargeted Staff Development Block Grant ($49.3 million). The remainder of the funds made available by our recommendations would be provided in a block grant to provide flexible resources to all districts for staff development needs. This grant would be distributed based on average daily attendance. Of this amount, $25 million would be one-time funding (to match it with the $25 million one-time funding source).

Governor's Scholars Program

The Governor's budget provides $111 million from the General Fund (non-Proposition 98) for the "Governor's Merit Scholarship program." The new program and its funding would be authorized under legislation separate from the budget bill. The Governor proposes to reward public school 9^th, 10^th, and 11^th grade students demonstrating high achievement in the state's Standardized Testing and Reporting (STAR) program with $1,000 higher education Merit Scholarships. The awards would be made to students scoring (1) in the top 10 percent statewide on the content standard-aligned portion of the STAR exam (STAR augmentation), or (2) in the top 5 percent of their school on the nationally norm-referenced Stanford Achievement Test version-9 (SAT-9) portion of the STAR program. The $1,000 scholarship would be placed into a Scholarshare Trust Fund account in the name of the student (see box below). During high school, a student could receive up to $3,000 in Merit Scholarships and--for some of these students--an additional $2,500 for the Governor's Distinguished Math and Science Scholarship Program (discussed later in this chapter), for a possible total of $5,500.

Merit Scholarships Address Two Policy Objectives

The Governor's Merit Scholarship program addresses two policy concerns by: (1) creating an incentive for students to improve academic performance and (2) providing early college outreach to some students who otherwise might not consider going to college.

Student Incentives. Student scores on the STAR (including the SAT-9 and the STAR standards-based augmentation) are not used currently to determine high school graduation or college admittance for individual students. As a result, students do not have a strong incentive to try their hardest on these tests. However, the state is holding high schools accountable for their students' performance on the STAR test. The Governor's Merit Scholarship program provides an incentive for at least some students to try their best, thereby increasing the appropriateness of using the STAR results as the yardstick for the accountability system's rewards and sanctions for schools.

On the other hand, virtually all the students who would receive scholarships at the high-achieving high schools would be college-bound in the absence of this program. In addition, these college-bound students take many other high-stakes tests that already encourage them to work hard during high school, as shown in Figure 1. For these students, it is questionable what further incentive for strong academic work the scholarships would provide.

Early College Outreach. Since the Governor's proposal would provide scholarships to students as early as 9^th grade, students at low-performing high schools may be provided with early exposure to the idea of higher education as a realistic option in their lives. Providing $1,000 to students at low-achieving high schools could create a strong incentive for some of these students to attend college or find other post-secondary training (such as trade schools). If students see going to college as a possibility in their future, they may also take a more rigorous course load in high school to qualify them for the California State University or University of California. In view of the above, the proposal has an element of college outreach. The same outreach effect is not likely to occur at the state's higher-achieving high schools, where virtually all potential recipients of the proposed scholarships are likely to attend college anyway.

Restructure the Distribution of Scholarship Eligibility

We recommend providing scholarships to the top 10 percent of students at each high school in order to (1) provide incentives for a more diverse set of students across the state and (2) increase the program's college outreach effect.

Under the Governor's proposal a predominant share of the scholarship funds would go to students in high-achieving high schools. Figure 2 shows our estimate of the number of students who would qualify for Governor's Merit Scholarships at 50 high-achieving high schools and 50 low-achieving ones. At these high-achieving schools, we estimate that 5,900 of 16,600 students taking the SAT-9 (36 percent) would receive scholarships by placing in the top 10 percent statewide. In contrast, an estimated 1,520 of 30,400 students taking the SAT-9 at the low-performing schools would receive scholarships based on the requirement that a minimum of 5 percent from each school qualify under the terms of the Governor's proposal.

Few Disadvantaged Students Would Qualify. As Figure 2 indicates, scholarship recipients under the administration's proposal would be concentrated at the higher-performing schools. From available STAR data, we can also predict that relatively few of the scholarships would go to
(1) students living in poverty (as measured by free and reduced priced lunches) or (2) English language learners. By changing the distribution, a more diverse group of students, both across socioeconomic groups and on a racial/ethnic basis, would have access to the Merit Scholarships. This type of distribution would also increase the scholarship program's college outreach effect, by reaching more students who otherwise might not be college bound.

The Governor's budget provides 111,000 students with scholarships. We estimate that the proposed $111 million funding level could instead provide the same number of scholarships to the top 10 percent of the pupils enrolled at each school with a Merit Scholarship. In order to (1) reach a more diverse set of students and (2) increase the college outreach effect of the program, we recommend that the Legislature provide Merit Scholarships to the top 10 percent of students at each high school.

Use One Test to Reinforce Academic Content Standards

We recommend using only the Standardized Testing and Reporting (STAR) program augmentation test scores to determine which students receive scholarships because (1) using the two tests causes unnecessary confusion, and (2) using the standards-aligned STAR augmentation test reinforces the central role the state's academic content standards must play in the state's accountability framework.

As mentioned above, the Governor proposes using both the STAR augmentation and the SAT-9 test results to determine eligibility for the Merit Scholarships. We believe that using both tests (1) creates unnecessary confusion, and (2) is a missed opportunity to reinforce the central role that the state's academic content standards ultimately must play in the state's accountability system. (Please see nearby box on the state academic content standards for more detail.)

The state has invested heavily--in every sense of the word--in the development of the academic content standards. These investments have taken many forms, including aligning tests to the standards, providing instructional materials, and providing staff development for teachers. The state should structure the merit scholarship incentives in a way that builds on these important investments. In view of the above, we recommend using only the STAR augmentation test scores be used to determine Merit Scholarship recipients.

Defer Implementation One Year

We recommend that the Legislature appropriate funds for the Merit Scholarship program for expenditure in the 2001-02 fiscal year instead of 2000-01 because the program needs to be in effect prior to students taking the test in order to affect their effort. Adoption of this recommendation would create $111 million of one-time non-Proposition 98 savings and $1 million in one-time Proposition 98 savings for other legislative priorities.

To create the incentive effect desired of the Merit Scholarship program, students first must be made aware of the program and then must be given time to adjust their behavior in reaction to it. Assuming this proposal is adopted as part of the 2000-01 budget, the program would not be established and funded until July 2000 at the earliest. Nonetheless, under the administration's proposal, students would receive scholarships rewarding performance on the spring 2000 test. Because students will have taken the test before the scholarship program is in effect, it is hard to conceive how the program could create a retroactive incentive for students to try harder.

We recommend that the Legislature appropriate the $111 million, but redesignate it for expenditure in the 2001-02 fiscal year. This would give students full opportunity to become aware of the new program approved by the Legislature and to prepare thoroughly for the test offered in the spring of 2001. Under this recommended change, students would first receive Merit Scholarships in the summer of 2001--that is, in the 2001-02 fiscal year. This recommendation, if adopted, would create $111 million of one-time non-Proposition 98 savings and $1 million in one-time Proposition 98 savings for other legislative priorities in the budget year.

Governor's Distinguished Math and Science Scholars Program

We recommend deleting $6 million provided for the Governor's Distinguished Math and Science Scholars Program because the funding would not create an additional incentive for California's highest achieving students.

The Governor proposes providing $2,500 scholarships to high school students scoring at the highest achievement level on advance placement (AP) tests in both calculus and science. Alternative qualifications are proposed for students at schools with no or minimal AP course offerings and for students taking advanced calculus or advanced physics. Based upon historical data, the Department of Finance (DOF) estimates that 2,400 students would score a five (the highest score) on both tests. On this basis, the budget allocates $6 million from the General Fund (non-Proposition 98) for this new program. The program and its funding would be authorized under legislation separate from the budget bill. The Governor's budget states the intent of the program is to encourage high school students to pursue rigorous course work leading to careers in scientific fields. We question whether this program would have much of an effect in this regard, since the students targeted by the scholarships are already highly motivated.

Students who receive a score of five on both AP calculus and AP science tests generally now receive a semester's credit for the college-equivalent class. The college credits can accelerate a college student's graduation date, which can result in significant savings in the cost of college. Thus, high school students already have a strong fiscal incentive to try hard on AP tests.

These elite students, representing approximately the top seven-tenths of 1 percent of the state's high school students, are likely to receive other scholarships from public and private sources based upon their achievement. The proposed $2,500 scholarship under this program--targeted to students who already are bound to the best universities and in line for numerous other scholarships--is a low pay-off investment of General Fund monies. In view of the above, we recommend that the Legislature reject the Distinguished Math and Science Scholars proposal and use the $6 million for other legislative priorities.

Advanced Placement Courses

The Governor's budget includes $16.5 million in new funds for initiatives aimed at increasing the availability of, and student access to, Advanced Placement (AP) courses by providing:

• $8 million (Proposition 98) in grants of up to $20,000 to at least 400 high schools for AP expansion activities.
• $5 million in federal funds for tutorial assistance and teacher support under the Advancement Via Individual Determination (AVID) program.
• $500,000 for the Office of the Secretary for Education (OSE) to contract for services to districts without AP courses and to assess the current availability of AP courses.
• $3 million (General Fund, non-Proposition 98) for the University of California (UC) to expand its on-line AP course development efforts. The budget also includes a one-time appropriation of $175 million in Proposition 98 funds for OSE to help schools purchase computers, including an unspecified amount for high schools without AP courses so that students can access such courses on-line. (We discuss the $175 million in further detail under the Education Technology write up later in this chapter.)

Background

What Are Advanced Placement Courses?

The AP program, overseen by the College Board (a national nonprofit organization), provides high school students the opportunity to take--and receive college credits for--college-level courses. High schools may offer up to 32 AP courses in 21 different subject areas, including biology, calculus, computer science, economics, English, music theory, psychology, and U.S. history. Near the end of a course, students can take a standardized exam administered by the Educational Testing Service called the AP test. In 1998, about 90,000 public high school students (or about 8 percent of enrollments in grades 10 through 12) in California took AP tests.

Students who complete AP courses and take the national AP tests can receive three benefits. First, many colleges and universities credit students who complete an AP course with an extra bonus point in the calculation of their grade-point-average (GPA). For example, a student receiving a B in an AP course gets four grade points instead of three, making the grade equivalent to an A in a non-AP course. Both UC and the California State University (CSU) grant bonus points for AP courses.

Second, UC campuses consider the number of AP courses students complete in their admission policy. If a student's scores on AP tests are available, UC campuses will consider the scores in their assessment of the applicant's academic achievement. Third, the UC and CSU also reward high school students who have scored three or higher (on a scale of one to five) on an AP exam with college credit for that course, thus shortening their stay at the university.

Availability of AP Courses

The Governor proposes that every high school in California offer at least four AP courses by fall 2001. Data on the number of AP courses at each California high school is somewhat unreliable. The California Basic Educational Data System (CBEDS), maintained by the State Department of Education (SDE), provides the number of AP courses available at each high school as reported by the school. However, the CBEDS data is subject to error, because the state does not verify what schools report.

A recent study by the CSU Institute for Education Reform reported on the availability of AP courses based on 1997-98 CBEDS data that was cross-checked with other sources. This report represents the most comprehensive review on the distribution of AP courses in California. The report noted that out of 868 regular public high schools (as of October 1997):

• 64 offered no AP courses.
• 216 offered between one and four AP courses.
• 311 offered between five and ten courses.
• 231 high schools offered more than 11 courses.
• 46 had self-reported AP data that could not be reconciled.

The OSE attempted to update the CBEDS data and the CSU Institute for Education Reform report in order to verify how many schools currently offer no AP courses. Based on the data we received from the OSE, we estimate that there are 73 regular public high schools in California that do not currently offer AP courses. Figure 1 presents a profile of these schools.

As Figure 1 indicates, most high schools without AP classes have small student enrollments. Of the 73 schools, 84 percent serve 600 or fewer students. In terms of location, 63 percent of the schools lacking AP courses are in rural areas. As we discuss below, these demographic characteristics provide some indication why schools may not offer AP courses.

Why Some Schools Do Not Offer AP Courses

The OSE surveyed the 73 high schools in order to identify why they offer no AP courses. The most common reason reported was that the school was too small to provide AP courses. For instance, small schools may not offer an AP course due to (1) lack of student demand, (2) difficulty of hiring AP teachers, or (3) excessive costs. Some of the small high schools stated that only a handful of students would sign-up for an AP course if it was offered, and it would be prohibitively expensive to offer a class to so few students.

Many schools reported that they provide other opportunities for students who want to take challenging and rigorous coursework. For example, 16 of the high schools offer four or more UC-approved honors courses. Both UC and CSU grant an extra GPA point for students completing an honors, international baccalaureate (IB), or college-level course. These schools generally provide honors courses rather than AP courses because (1) the curricular restrictions from the College Board give teachers less flexibility in AP courses, (2) AP courses require additional teacher training, and (3) there may be less demand for AP courses. When available, schools encourage their students to take courses at their local community college. Finally, many of the schools located in rural areas plan to provide their students with access to on-line AP courses because their students may not have access to courses at a community college. The Governor's budget proposes to help these schools access AP courses on-line (see below).

The Governor has established a goal that each high school offer at least four AP courses (the number of courses recommended by the College Board for a standard AP program) on-site by fall 2001. The budget includes initiatives to address the unequal distribution of AP courses across California public high schools. Last July, the American Civil Liberties Union filed a class action lawsuit (Daniel v. California) against the state challenging this unequal distribution. The plaintiffs, four students from Inglewood High School, claim they were denied equal and adequate access to AP courses by the state and their local school district because their school offered only three AP courses.

Legislative Analyst's Recommendations

Recommend Approval of Expansion in Two Ongoing Programs

In order to increase student access to AP courses, the budget proposes to expand two ongoing programs, as described below.

The AVID Program. The budget provides a total of $12 million ($1 million in General Fund (non-Proposition 98) and $11 million in federal Goals 2000 funds) for the AVID program. This represents an increase of $5 million (all federal funds) above estimated current-year expenditures. The AVID program is a precollege readiness program for underachieving students who are (1) from low-income families, (2) from groups historically under-represented in colleges, or (3) will be the first in their family to attend college. The program currently serves 36,000 students from about 600 middle and high schools. The Governor's budget proposes expanding AVID to include support for AP teachers and tutorial assistance for students in AP courses.

The UC's On-Line AP Program. The budget provides a total of $7 million for UC (General Fund, under Item 6440-001-0001) to deliver on-line AP courses to students who attend high schools with few or no AP courses. This total includes a proposed augmentation of $3 million for UC to develop and provide five on-line AP courses to about 1,000 students per semester (for a total of 2,000 course enrollments). The UC initiated the on-line AP program as a pilot project in 1998-99 and began receiving funding from the state for this program in the current year. For 1999-00, UC is using courses supplied by outside vendors and is currently delivering 12 courses to about 200 students from 34 schools. The budget provides funds for UC to develop its own AP courses, as well as provide a few courses supplied by outside vendors. The goal of developing UC's own courses is to offer a higher quality course than those available from outside vendors. The on-line courses are taught by experienced AP high school teachers or content-area college and university instructors.

We recommend approval of the proposed augmentations of $5 million for the AVID program and $3 million for the UC's on-line AP program.

Premature to Propose AP Grant Program

We recommend that the Legislature adopt budget bill language requiring the Office of the Secretary for Education to submit a report addressing the problem of access to Advanced Placement (AP) courses, including a comprehensive solution. We further recommend that the Legislature redirect $8 million for the AP grant program to other legislative priorities, because the grant funds are premature until the Secretary completes his study, identifies the problem, and develops a plan to address the issue.

Secretary to Develop Plan. The budget proposes $500,000 from the General Fund (non-Proposition 98, under Item 0650-011-0001) for the OSE to contract for services to assist school districts that do not currently offer AP courses in developing options to ensure access to such courses. The OSE would also use these funds to assess the current availability of AP courses, and determine a strategy to meet the Governor's goal of providing at least four AP courses at every high school. At the time of this analysis, OSE could not provide specific information on how it would complete these tasks.

The AP Grant Program. The budget includes $8 million under Item 6110-193-0001 to provide grants of up to $20,000 to at least 400 high schools to use for AP professional development, necessary instructional equipment and materials, and other course start-up costs. The Governor's proposal authorizes SDE to distribute the grants through a competitive evaluation process based on the following priorities: (1) schools offering three or fewer AP classes, (2) schools not offering AP classes in either math or science, (3) schools with low college participation rates, and (4) schools with a majority of students who qualify for free or reduced-price meals. Schools receiving grants would be expected to offer AP courses in a minimum of four subject areas, including math and science, and build an academic support system for low-income, minority students in AP courses. At the time of this analysis, the administration was unable to specify to what extent this $8 million program would advance the state toward the Governor's goal of at least four AP courses in each high school. It was also unclear from the budget proposal why four courses in each high school should be the state's policy goal.

We concur with OSE that further study is necessary to develop a comprehensive strategy to address the issue of access to AP courses. Accordingly, we recommend that the Legislature (1) provide OSE with the $500,000 requested for a comprehensive study; and (2) adopt budget bill language requiring the OSE to submit a report to the Legislature addressing the problem of access to AP courses, including a comprehensive solution. However, until such a report has been completed and the Legislature can review its strategy for equitable access to AP courses, we believe it is premature to start the $8 million grant program. Accordingly, we recommend that the Legislature delete the $8 million budget-year request. (Reduce Item 6110-193-0001 by $8 million.)

Some of the issues the study should address are :

• What Is the Current Distribution of AP Courses? The OSE needs reliable data on the number of AP courses available at each high school in California in order to develop a plan to increase student access to such courses.
• Why Do Schools Offer No or Few AP Courses? In order for OSE to develop a plan to increase student access to AP courses, it needs to first gather detailed information on why schools offer no or few AP courses. We believe this information is needed to determine the appropriate solution strategy. For example, based on the study's information, the plan might provide a different solution for small high schools versus large high schools.
• Access to AP or Access to UC? The UC grants extra GPA points for students completing AP courses, as well as IB, honors, and college-level courses. As a result, we believe that the study needs to fully examine to what extent access to UC is constrained by the lack of AP courses at high schools. At this point, available evidence is mostly anecdotal.
• Rural or Urban Problem? The OSE's preliminary findings suggest that many of the schools without AP courses are located in rural areas. However, the plan proposed by the administration does not propose different strategies for increasing access to AP courses at rural schools versus urban schools.
• What Are the Long-Term Costs? The OSE staff acknowledge that the Governor's goal of ensuring at least four AP courses in each high school would result in additional costs in the long run. However, OSE cannot provide an estimate of these costs. We believe that any plan to increase student access to AP courses should account for the long-term costs of implementing the plan.