Budget

K­12 Education

K-12 Priorities

The budget contains a number of major issues that raise the question of the appropriate role of the state in the design and operation of California's K-12 education system.

The 1998-99 Governor's Budget proposes to spend $30.8 billion in state and local Proposition 98 support for K-12 education. This represents an increase of $2 billion, or 6.9 percent, from the amount included in the 1997-98 Budget Act. The budget proposes to (1) provide growth and cost-of-living-adjustment (COLA) funds to virtually all programs and (2) spend $800 million to support state initiatives to improve the K-12 system. Figure 8 (see next page) illustrates the budget's proposed allocation of K-12 funds for 1998-99.

Budget Poses Major K-12 Issues

• The Budget Does Not Provide Any "Real" Increase in District General Purpose Funding.Except for the 2.2 percent COLA, districts will have no new funds to meet locally identified needs. Instead, the budget proposes to use more than $800 million in

Proposition 98 funds to expand 21 existing programs and create four new programs. The administration also proposes to direct districts to use new lottery funds for a specific purpose (textbooks).

• The Budget Proposes Several Initiatives That Could Result in Significant Unfunded Cost Increases for School Districts. For instance, the Governor proposes to lengthen the school day by excluding the time between class periods (known as "passing time") from district calculations of mandated instructional minutes. The budget includes no funding to districts to meet the potential additional costs that this new state policy could create. Thus, in a sense, the Governor proposes to spend district funds to meet administration goals.
• The Governor Proposes to Override Local Control Over Standards for Adequate Student Progress. The Governor, for instance, proposes to end "social promotion"--the practice of advancing a student to the next grade even though the student is not proficient at the skills and knowledge expected of students at that grade. This proposal would inject the state into a policy area that, currently, is controlled by school district governing boards, teachers and parents.
• The Budget Would Allow Some Parents to Opt Out of the Public System Altogether as a Means of Finding a Better Education for Their Children. The budget proposes to use vouchers to allow up to 15,000 students who attend low-performing schools to attend another public or private school. By ceding to parents the ability to direct state funds to any private school, the budget proposes a major change in the governance of K-12 schools.

State's Role in K-12 Education Has Increased Over Time

The budget proposals continue a trend of increasing state activity in K-12 education. Under the Constitution, responsibility for K-12 education rests with the state. Until the 1970s, however, governance and funding of schools was primarily local. The state's role was limited primarily to providing additional funding through K-12 categorical programs. In the 1970s, school governance was altered with the enactment of state collective bargaining laws, and school finance became a state issue with the Serrano decision (funding equalization) and Proposition 13 (property tax limitation).

Thus, as the 1980s began, school districts had lost a significant amount of control over important aspects of school operations. As schools lost control over finances, the state gained control. And, with the financial control came accountability--state decision-makers needed to justify to voters that funds were used wisely. This need for accountability tends to bias state discussions over the K-12 budget toward specific categorical approaches and away from providing general purpose funding increases to districts to address their own priorities.

The 1990s have seen the state continue to expand its jurisdiction over the K-12 system. Proposition 98 has furthered this trend. By requiring the state to spend a minimum amount on education, Proposition 98 has moved the state budget process away from discussing how muchmoney schools need and, instead, has become focused on how to spend the "required" amount. As a result, Proposition 98 has further encouraged the budget process to concentrate on categorical program proposals.

Data on the composition of school spending show that spending on categorical programs has increased faster than general purpose support for schools. In 1982-83, 77 percent of state and local spending on K-12 was provided through revenue limits--the general purpose funding schools receive for each student. The other 23 percent of K-12 funding was spent on categorical and other K-12 programs. The 1998-99 budget proposes to spend 70 percent of K-12 funding on "base" revenue limits (excluding certain categorical programs that are funded through revenue limits). This 7 percent difference may seem small, but it represents more than $2.1 billion that would be in general purpose funding in 1998-99 had the state continued to spend 77 percent of state and local spending on revenue limits.

State Needs Long-Term K-12 Framework

This long-term trend toward increasing state intervention in K-12 finance and policy has a significant impact on the effectiveness of the overall system. Yet the proper roles of the state and the local school district have never been identified. We question, for instance, whether state intervention in detailed matters of local administration and policy result in the most effective education system. The categorical approach to school finance assumes that the state can "fine tune" educational practices through its funding system in order to increase student achievement. The evidence from educational research, however, suggests the opposite is true--that "one size" state policies do not "fit all" district needs. In fact, the categorical approach to educational funding risks reducing the impact of additional funds. When the state substitutes its judgment of what K-12 funds should buy, it short-circuits the local process for determining how additional funds best translate into improved student achievement.

Similarly, state intervention creates problems of governance and accountability. Who should be held accountable for the successes and failures of the K-12 system when several levels of government are responsible for managing K-12 resources? A locally governed, locally financed system provides clear accountability for outcomes and a direct way for voters to hold local decision-makers accountable. Undue state intervention blurs responsibility at the local level. Additionally, voters in any particular school district have no direct way to hold state decision-makers accountable for K-12 outcomes.

We think the state should develop a framework for K-12 education that establishes a guide for its decisions on school finance and other policy matters. This framework would identify the appropriate state and local role in the design and operation of the K-12 system. In our view, the framework would place the state in control of determining the broad design features of the K-12 system. Local school districts would have substantial flexibility within this system, maintaining control over major policy and budget decisions. This would focus the state's attention on the following areas.

Governance. Effective governance at the local level is crucial to a well-functioning locally controlled education system. The state would assure that there is an appropriate balance of interests at the local level. For instance, the state would need to ensure that the system's clients--parents and the business community--have sufficient voice in decision-making. Other governance issues that would be monitored by the state include (1) the impact of the collective bargaining process on the ability of districts to allocate funds in a way that best balances the educational needs of students with teacher compensation and working conditions and (2) whether districts may be too large or too small to provide efficient financial operations and an accessible governing process.

Funding. Funding is clearly a state concern because of past court decisions and voter-approved initiatives. That does not mean that the state must make most funding decisions, however. The state needs to ensure that funding levels are adequate and distributed fairly and that schools and school districts have enough funding flexibility to meet high-priority local needs. In addition, creating new local funding mechanisms (such as the ability to increase local property taxes) would provide school boards with greater financial flexibility and accountability.

Minimum Educational Standards. The state needs to establish an accountability system that provides a clear sense of what schools are producing. This is essential for both local voters as well as state decision-makers. The state, however, needs to focus standards on critical system inputs and outputs. Inputs could include student-teacher ratios, measures of school safety and facility adequacy, and the amount of time students actually spend learning (as opposed to time spent in recesses, assemblies, and other non-academic activities). Outputs could include test scores, dropout rates, college attendance rates, and employment rates and wages of high school graduates. These standards should create an incentive for the school districts to improve. Therefore, as part of the framework, the state would need to create consequences for schools that cannot meet minimum state standards, including state receivership.

Oversight and Information. The state can exert significant influence over the K-12 system through monitoring its operations and by providing useful data to school districts. State monitoring of the success of school districts in achieving the desired inputs and outputs is an important part of making state standards meaningful. In addition, the state as an information broker can encourage districts to adopt policies that are most effective in improving student achievement.

Creating this long-term framework would not be a simple task. However, it would be well worth the effort. The Master Plan for Higher Education has been guiding the development of higher education in California for more than 30 years. The value of the plan is that it provides a reference point for understanding the critical design features of the higher education system, such as the roles of the three segments (University of California, California State University and community colleges). The plan is not static--the Legislature periodically reviews and refines the Master Plan, reshaping elements as needs and resources change. We think a similar long-term framework for K-12 would provide similar benefits.

Develop Information on Future Funding Options

No matter which level of government makes the major decisions about the best use of K-12 resources, there is a pressing need for better information about the effectiveness of different approaches to improving schools. The state--either to improve its own budget decision making or in the role of supporting school districts with data and research--should play a central role in developing this information.

At the current time, there are little good data on the impact of different services and programs on student achievement. Because of the important role the state currently plays in determining how K-12 funds are used, this lack of data requires the Legislature to make decisions with little information about the consequences of its actions and the relative effectiveness of the available options.

As we discuss later in this chapter, for instance, there is little solid evidence to support the budget's proposal for a longer school year. To be fair, there is very little evidence at all on the question. The research we could find on the issue suggests that it's unlikely the proposal will have a measurable impact on student achievement.

Similarly, there are very little data on the effectiveness of other K-12 programs the Legislature could fund as an alternative to the longer year proposal. We considered two major possibilities: additional class-size reduction (grades 4 through 6) and a longer school day in middle and high school.

We also reviewed the research on the impact of school vouchers on student achievement. The results were the same as our other research reviews--very little solid data. It is possible that the Governor's proposal for Opportunity Scholarships would make an important difference to students attending low-performing schools. Having an option to "exit" the public system may allow students who leave to get a better education and spur improvements in the quality of education for those who stay behind. On the other hand, it is also possible that markets would not function effectively in this case--private schools may be unwilling to locate in poor neighborhoods, for instance, which would limit the choices available to parents.

Legislature Should Create Demonstration Programs

For the most part, this is unlikely to happen without government providing the impetus and funding for such research. Demonstrations are costly undertakings--usually more than most districts could afford. The benefits of research, however, are available to all school districts. As a result, this type of research usually is supported by a higher level of government, primarily states and the federal government.

The class-size reduction demonstration, conducted in Tennessee, illustrates the benefits of such programs. Tennessee sponsored a long-term demonstration program testing the impact of smaller class sizes in grades K-3. The program provided high quality data on the impact of smaller classes for different types of students. Indeed, if the State of Tennessee had not invested substantial resources into its evaluation, there would have been little good data to support California's program.

To begin the process of evaluating the state's future options regarding K-12 programs, we recommend the Legislature establish four demonstration programs. These demonstration programs would allow the state to measure the impact of major programs designed to improve the quality of K-12 education. Each demonstration would be relatively small, with perhaps 10 to 20 schools participating in the program. Control groups would be created so that the impact of the new services could be easily measured. We suggest the Legislature consider the following four programs.

Class Size Reduction--Grades 4 Through 6. As schools complete the implementation of class size reduction in grades K-3, parents and teachers will begin to ask whether the state will create a program for smaller classes in the intermediate grades. Smaller classes are intuitively appealing to parents. Smaller classes in grades 4 through 6 also make sense from a teacher perspective. Intermediate grade teachers teach up to 32 or 33 students, while the K-3 teacher next door works with only 20 students. Existing data on the impact of smaller classes in grades 4 through 6 suggest little impact, however. This also was true of research findings on smaller classes in grades K-3 before the Tennessee demonstration program. A good evaluation, similar to the Tennessee demonstration, would yield valuable data on the question. We suggest trying two minimum class sizes--20 and 24 students.

Longer School Day--Grades 7 Through 12. A six- or seven-period day is standard in most middle and high schools in California. Unfortunately, this permits students to take little more than the core curriculum--math, language, science, history, and gym use up five of the six/seven class periods. As a result, students have little chance to explore other subjects, such as foreign language, computer science, music, art, geography, and vocational education. These types of "enrichment" courses get squeezed out of the school day. For some students, these enrichment courses are the main attraction of school.

In addition, the short day means students get out of school early, sometimes with little home or school supervision. If the school day ends at 2:15 in the afternoon, students have at least three hours before working parents begin returning home. Therefore, extending the school day would permit these older students a broader curriculum and reduce the amount of unsupervised time after school for students whose parents are not at home.

The demonstration program would measure the cost and impact of longer school days in middle and high schools. Measuring costs is pretty straight-forward. Benefits would include the impact of additional courses on student achievement, dropout rates, and other educational indicators. The effect of the longer day on juvenile crime in the area would be another potential impact.

Longer School Year. As we discussed above, there is not much good data on the impact of a longer school year. While we are currently funding a study of a longer school year in high school (Oxnard), we do not think this demonstration will provide a good indication of the impact of a statewide initiative to increase the length of a longer year. A larger demonstration program would gauge the extent to which districts of different sizes and types would use the additional time productively. For this reason, we suggest a separate demonstration program to test the impact of a longer school year in all grades.

Vouchers for Students Attending Low-Performing Schools. We think the idea of vouchers has sufficient merit that the Legislature should sponsor a demonstration program as a way to understand the costs and benefits of the concept. A demonstration program would select a small number of schools where students would be given the opportunity to apply for a voucher. As part of the program, researchers could address a number of important questions:

• What choices do voucher-holding students have?
• Do they fare better than similar students who remain at the school?
• How does the school respond to this type of competition?
• Do special education students seek and use vouchers?

There also are a number of legal questions that would have to be addressed before a demonstration program of the impact of vouchers could begin. At the current time, these questions include (1) whether public funds could be used for the vouchers? and (2) whether religious schools could receive funding through vouchers?

Provide Start-Up Funding. The demonstration programs would require a year of development--time to design the programs (with evaluations as part of that design), address legal and financing questions, and select schools to participate. Therefore, we recommend the Legislature appropriate $2 million in federal Goals 2000 funds to begin the design, planning, and initial implementation of the four demonstrations. This would provide support for state and local costs in the budget year so that the demonstration programs could begin operation in 1999-00.

LAO Outlook Indicates Fewer Proposition 98 Funds

Figure 9 details the Proposition 98 amounts available for 1997-98 through 1999-00 under the Governor's budget forecast of the Proposition 98 minimum funding guarantee and under the forecast developed by the Legislative Analyst's Office. In 1997-98, we forecast no difference in Proposition 98 funding. While we project higher General Fund revenues than the administration for 1997-98, this increase does not translate into any increase for K-14 education. Since Proposition 98 is in "test 2" during 1997-98 (which does not use General Fund revenues to determine the minimum guarantee), the formula does not generate any additional Proposition 98 spending requirements.

Figure 9
LAO and Governor's Budget Proposition 98 Forecastsa
(In Millions)
Forecast	1997-98	1998-99	1999-00
Governor's budget	$32,693	$34,682	$36,529 b
LAO	32,693	34,486	36,320
Difference with budget	$0	-$196	-$209
aAssumes funding at the minimum required under Proposition 98. All estimates assume passage of the Opportunity Scholarship program, which reduces K-12 attendance by 15,000 students.
bLAO long-term extrapolation of economic and revenue projection underlying the 1998-99 Governor's Budget proposal.

As Figure 9 shows, we estimate that Proposition 98 funding will be $196 million lower than indicated in the Governor's budget for 1998-99. Even with our lower estimate of the guarantee, Proposition 98 funding is expected to increase $1.8 billion from the level provided in the 1997-98 Budget Act.

Why our Proposition 98 estimate is slightly lower is a complicated story. Proposition 98 is based on "test 3" in 1998-99, where additional revenues can increase the minimum guarantee. Initially, however, it does not, because our estimate of the increase in 1997-98 General Fund revenues is greater than our projected increase in 1998-99 General Fund revenues. This difference in the amount of additional revenues in the two years creates a smaller net change in General Fund revenues. Because "test 3" uses the change in General Fund revenues to determine the guarantee, the formula generates a smaller Proposition 98 guarantee.

This is another demonstration of the unpredictability of the Proposition 98 formula and how the formula does not adequately balance school funding requirements with the state funding situation. At the time the 1997-98 budget was adopted, the state General Fund condition was fairly "tight," yet the Proposition 98 formula required large additional education expenditures (which further complicated the state's budget picture). Then in 1998-99, when the state's fiscal position is much better, the Proposition 98 formula results in moderate increases.

Legislature Faces Major Trade-Offs

As we discussed above, once funds are allocated for growth and COLA, the budget proposes to spend about $800 million to support new or expanded state activities. Before the Legislature can make any decisions on allocating Proposition 98 funds for K-12 education in 1998-99, it must first resolve three major issues that will significantly affect the amount of "available" funds remaining. These issues involve such large amounts of funds that the rest of the Governor's K-12 spending proposals--and any legislative proposals that are contemplated--hinge on the outcome of the three issues. These three issues are:

• The Proposition 98 Minimum Guarantee. Based on the current split of Proposition 98 funds between K-12 education and community colleges, we estimate the Governor's budget overstates by $176 million the amount of Proposition 98 funds for K-12 in 1998-99. This is due to our lower estimate of the Proposition 98 guarantee in the budget year.
• The Appropriate K-12 COLA for 1998-99. Although the budget proposes a 2.22 percent COLA, statute calls for a 4 percent COLA in the budget year. Providing the higher amount would require an additional $530 million in Proposition 98 funding. We discuss this issue later in this section of the Analysis.
• The Governor's Longer School Year Proposal. This one proposal accounts for $350 million in new funds in the budget year. This issue also is discussed in detail later in this section of the Analysis.

With the $176 million drop in the amount available to K-12 schools in 1998-99, funds for new or expanded programs fall from about $800 million to $625 million--still a substantial sum. One central implication of the lower funding level is that the amount needed to fund the statutory COLA or the longer year program proposed by the Governor constitutes a very large proportion of available funds. In fact, if the Legislature chooses to provide the statutorily required COLA (4 percent), there would be only $100 million in funds left with which to support other programs in K-12. Funding the longer year program would also require the Legislature to reject many of the other proposals made by the Governor.

Balance State and Local Funding Needs

Despite our long-term concerns discussed above, the Legislature faces the immediate question of how best to appropriate funds in the 1998-99 Budget Bill. We have recommended in the past that the Legislature ensure that its budgetary decisions balance the needs of state and local decision-makers as well as the state's long-term goals for K-12 education. We think this balance continues to be the central issue in the state K-12 budget.

Figure 10 displays our recommended guidelines for K-12 education funding priorities. The first two guidelines focus on the existing K-12 budget. Current program costs--including funding for growth in the student population, COLAs, state mandates, and other funding commitments--should take top priority for funding. These "base" programs, however, also should be assessed and refined as needed to make sure they (1) are as effective as possible and (2) further the state's K-12 goals.

After meeting basic program costs, any additional Proposition 98 funds should be used to meet both state and local funding needs. In California's divided system of K-12 school governance, the Legislature, the Governor, and local school boards all play a role in identifying and funding school priorities. Under this system, it is our view that the state and local school boards each should determine the priorities over about half of available new K-12 funds.

For that reason, our guidelines call for using about half of available funds to increase and equalize district revenue limits. With the remaining funds, the state should pursue programs for which there is a strong case for state intervention. We discuss our specific recommendations below.

Figure 10
LAO Guidelines for Establishing K-12 Education Funding Priorities
	Fund the Continuing Costs of the Current Program.The state should ensure that funding for growth, COLAs, and other financial commitments is provided.
	Review and Reform Current Programs Consistent With Long-Term Goals. The Legislature should periodically review whether existing programs further the long-term goals of the state.
	Balance State and Local Funding Needs. Once existing programs are adequately funded, the allocation of additional funds should reflect the shared state and local governance of the K-12 system.
	Increase and Equalize Local Revenue Limits. About half of any additional funds should increase and equalize local revenue limits, thereby allowing school districts to address their priorities.
	Support State Improvement Efforts. The remaining funds should support state-directed uses based on a long-term plan of reform.

Options for 1998-99 Funds

Figure 11 (see next page) displays the Governor's proposal and the LAO's recommendations for the 1998-99 K-12 budget. The Governor's budget request is based on the amount of Proposition 98 funds identified in the budget. Our recommendations are based on the LAO projection of the minimum guarantee, which reduces available funding by $176 million.

In addition, our recommendations reflect an increase of $91 million in Proposition 98 funding that is connected to the Governor's proposal to create Opportunity Scholarships (or vouchers) for 15,000 students. The budget assumes that all 15,000 students would use the scholarships to attend private schools. Because Proposition 98 funds may only go to public schools, the budget removes the 15,000 students from the Proposition 98 calculation. This reduces the minimum guarantee by $91 million in the budget year. Because we recommend the Legislature first test the impact of vouchers before implementing a program as large as proposed

Figure 11
1998-99 K-12 Proposition 98 Increases LAO and Governor's Budget Proposals
(In Millions)
	Governor's Proposal	Legislative Analyst's Proposal
Fund the continuing costs of the current program
Base programs and adjustments	$19,597.8	$19,597.8
Enrollment growth	540.2	540.2
Cost-of-living increases	623.4	623.4
Class size reduction "step" increase	--	77.3
Pay program deficiencies and other commitments
Mandates	$6.6	$7.8
Child care	44.0	44.0
Voluntary desegregation	18.9	18.9
Digital high school	26.0	26.0
Review current program consistent with long-term goals
Increase revenue limits	--	$150.2
Child care augmentations	$138.1	138.1
PERS rate adjustment	-59.0	--
Longer day/longer year incentive	--	0.2
Support state improvement efforts
Deferred maintenance	$135.0	$135.0
High-risk youth education	20.0	10.0
Assessment	27.8	7.6
Beginning teacher support and assessment program	16.1	10.0
Reject other proposed augmentations
Staff development day buyout	$400.0	--
Opportunity scholarships	-58.0	--
Deficit factor buyout	52.9	--
New grade 3-6 summer school	10.0	--
Zero tolerance	6.2	--
Proposition 98 Reserve	--	$75.0
Totals	$21,546.0	$21,461.5
LAO adjustments to Proposition 98
LAO forecast	--	-175.9
Opportunity scholarships	--	91.4
Total Adjustments	--	-$84.5

in the budget (see our discussion below), our alternative budget "adds back" this amount. The net effect on Proposition 98 funding is that we are only $85 million lower than the Governor's budget.

The amounts in Figure 11 represent most of the proposed K-12 increases for 1998-99. Not shown are several relatively small proposals with which we take no issue. Since we also make a number of recommendations later in this section, we include these changes in Figure 11.

The LAO recommended plan departs from the Governor's plan more than in past years. As Figure 11 displays, we recommend approval of all the budget's proposed COLA increases and those increases requested to keep previous state commitments. We also propose to spend $77 million that is not included in the budget to maintain K-3 Class Size Reduction funding at the average statewide cost and $1.2 million for a mandated cost we expect will come due in the budget year. We also recommend approval of the proposed $135 million increase in the deferred maintenance program.

In the area of new or expanded school improvement programs, our alternative uses available funds quite differently. In part, this was caused by our lower estimate of Proposition 98 funding for 1998-99. In addition, our alternative does not include funding for the longer school year. The Governor's budget earmarks $400 million for this program (a $350 million augmentation).

Our alternative plan also includes proposals to: (1) increase local revenue limits by $150 million, (2) eliminate the revenue limit adjustment that results from changes in Public Employees' Retirement System rates (at a cost of $59 million), and (3) eliminate the "deficit" in county office of education and other school district programs without providing new funding, for a savings of $53 million.

Below, we discuss in more detail several of the elements of our alternative plan. Other differences are discussed in separate sections later in this chapter.

Create a Proposition 98 Reserve

In the years after Proposition 98 was passed by voters, the Legislature created a reserve account to protect the state in the event the projections of General Fund revenues assumed in the budget were too optimistic. During the early 1990s, a reserve was not created and, as a result, the state over-appropriated the minimum funding guarantee several times.

The Proposition 98 guarantee is hard to predict. Changes in the growth of state revenues, property taxes, and population can dramatically affect the General Fund share of the guarantee. For instance, our forecast for the current and budget year shows higher growth in General Fund revenue compared to the Governor's budget, but the calculated guarantee for 1998-99 is less than the Governor's projection.

A Proposition 98 reserve provides the Legislature with some measure of protection against over-appropriating the minimum guarantee without reducing amounts already provided to schools. When the reserve is not needed, the funds would be available as "settle-up" funds in the following year to support one-time programs, such as block grants or deferred maintenance.

Therefore, to create a buffer from an unanticipated reduction in the Proposition 98 minimum guarantee, we recommend the Legislature set aside $75 million as a Proposition 98 reserve.

Approve the 2.2 Percent COLA

Current law requires the Department of Finance (DOF) to use the inflation index for state and local government purchases as the annual COLA for K-12 schools. Each year, the budgeted COLA is calculated, according to Education Code Section 42238.1, as the annual change in that index over the prior two years. In most years, this is a straight-forward calculation.

It becomes more complicated when the inflation index is revised. The Education Code requires DOF to calculate the COLA by dividing the prior-year revised index by the unrevised index of a year earlier. This is the case for 1998-99. Due to revisions in the index, the statutory formula calls for a 4 percent COLA.

The budget proposes trailer bill language that would delete the requirement to compare revised data with prior-year unrevised data. Instead, the trailer bill language would use only revised data (when the inflation index is revised) so that COLAs reflect the best estimate of inflation that occurred during the past year. This change would result in a 1998-99 COLA of 2.22 percent. (Providing the full 4 percent COLA in 1998-99 for most K-12 programs would cost an additional $530 million.)

Change Is Appropriate. We recommend the Legislature approve the proposed trailer bill language because the proposal more accurately reflects inflation that has occurred over the past year. In fact, using the current statutory formula to calculate K-12 COLAs would compensate schools for inflation that has not actually occurred.

The counter-arguments to the proposed change appear to be: (1) a higher-than-necessary adjustment in 1998-99 helps to compensate for past-year "shortfalls" and (2) there is some type of bias in using "revised-to-revised" numbers. We could not find any evidence to support these arguments.

• Schools Have Been Adequately Compensated for Inflation in the Past. Figure 12 displays, for 1992-93 through 1997-98, the COLAs as they were calculated each year for the budget and actual inflation. While the figures differ marginally from year to year, the net change over the six years is almost identical. Consequently, there is no need to provide extra COLA compensation in 1998-99.
• The Proposed COLA Calculation Method Seems Fair. The proposed COLA calculation would use a formula that is almost identical to the current community college inflation adjustment. We would be concerned if we detected a systematic bias in this calculation--that is, one that resulted in COLAs that were too high or too low over the long-run. We did not, however, find any evidence of such a bias.

Based on these findings, we conclude that the trailer bill language proposes a COLA that more accurately reflects actual inflation than the current statutory formula. Therefore, we recommend the Legislature approve the budget proposal. This would provide a 2.22 percent COLA in 1998-99.

Provide Step Adjustment for Class Size Reduction

The Governor's budget proposes $1.5 billion for the class size reduction (CSR) program, an increase of $57.6 million over the revised 1997-98 level. The increase includes $24 million for growth and a $33.6 million COLA. This program was introduced by the Governor and Legislature in 1996-97 to reduce average class sizes in kindergarten through grade 3 from 28.6 students to no more than 20.

Almost all districts participate in the program, implementing half or full day programs in one or more of the K-3 grades. Schools with half day programs receive $400 per participating student and those with full day programs receive $800 per student. The amount provided for class size reduction roughly equals the average cost of the program.

In our 1997-98 Analysis we commented that the long-term cost of class size reduction would rise over time since many of the teachers hired for class size reduction were new teachers. New teachers generally start at the bottom of the salary schedule, rather than the statewide average. Teachers hired at the beginning of the program have moved up the salary schedule, increasing teacher salary costs for districts. The proposed class size reduction augmentations in the budget do not provide sufficient funding for these salary increases.

We feel the state needs to maintain funding at a level sufficient to meet the average district's costs. Failure to do so will shift to districts the higher cost of the program, requiring districts to redirect funds from other programs. Therefore, to fund salary step adjustments that are built-in to most district pay schedules, we recommend increasing CSR funding by 5 percent, or $77.3 million.

Additional Mandate Costs Likely

Current law requires school districts to notify law enforcement agencies of two types of pupil behavior: (1) the possession or sale of narcotics and other controlled substances and (2) the possession of weapons (for example, guns and knives) on school grounds. The Commission on State Mandates determined the mandated local costs of these notifications total $1.2 million in the budget year. In addition, $4 million is needed to satisfy claims for this program for 1994-95 through 1997-98.

The Governor's budget does not include funding for this mandate. Because the state is obligated to pay these claims, we recommend the Legislature provide $1.2 million to fund the costs of the mandate in 1998-99. We also recommend that the Legislature provide $4 million in one-time funds to pay for the prior-year claims.

Opportunity Scholarships

The Governor's budget proposes $52.2 million from non-Proposition 98 General Funds, to provide educational vouchers to 15,000 students attending low-performing public schools. These students would be allowed to use their voucher towards attending a public or private school. The proposal sets the amount of the voucher at $3,500, or equivalent to 90 percent of the statewide average per student revenue limit of $3,890.

The proposal requires a set of changes to the budget.

• First, the budget removes 15,000 students from the Proposition 98 calculation, for a General Fund savings of $91 million. (As shown in Figure 11, the budget also removes the 15,000 students from the calculation of school district revenue limits, which reduces Proposition 98 spending in the budget year by an estimated $58 million.)
• Second, the budget proposes to spend $52 million in General Fund monies for the private school vouchers. As a result of these transactions, Proposition 98 is $91 million lower, and the General Fund saves $39 million.

Although the details of the scholarship program were not available, we have identified the following concerns with the general outlines of the proposal:

• Transfers Are Already Permitted. Students participating in the program would be allowed to attend any public school or a private school. Students who choose to attend an alternative public school are already able to do so without a voucher, as the state currently allows districts to claim funding for students from other districts.
• Selection Criteria Unclear. According to the Governor's proposal, students attending schools with achievement tests in the bottom 5 percent across the state would qualify for vouchers. We were unable to obtain information on how the bottom 5 percent would be determined.
• Is Voucher Amount Sufficient? The budget sets the voucher amount at $3,500 per student. The Department of Finance, however, was unable to provide any information about whether this amount would be sufficient to allow students to attend a private school. This feature is critical to the success of any voucher program.

As we discuss above, the value of vouchers in improving educational choices has not been established. For this reason, we believe a large voucher program, such as proposed in the budget, is premature. Instead, we propose the Legislature initiate a smaller demonstration program that would provide good data on the impact of a voucher program.

Until better data are available on the impact of school vouchers, we recommend the Legislature reject the Governor's proposal. This would require a net $39 million General Fund increase.

Delete Summer School Augmentation

The budget proposes $10 million to establish grants to schools interested in offering remedial summer school classes to improve reading skills of students in grades 3-6. Students would attend summer classes to bring them up to grade level or to a higher level of reading proficiency.

This program would be in addition to regular summer school programs already offered by districts. The budget includes $122.5 million for regular summer school programs in the budget year. Schools are permitted to provide a wide range of instruction, including remedial classes such as those contemplated in the Governor's budget, as part of regular summer school.

As a result, we see no need to augment summer school funding for this purpose. Therefore, we recommend the Legislature delete the $10 million proposed for the new program.

Preserve Local Autonomy Over Expulsions

The budget proposes trailer bill language to mandate student expulsions when a student is caught with unlawful drugs (except for small amounts of marijuana) at a school or at an off-campus school event. To provide an alternate educational placement for expelled students, the budget proposes $6.2 million for the higher costs associated with those placements. Under current law, schools may expel a student for drug-related offense.

In general, we think these types of policies are best left to school districts. Our system of local control allows districts to craft expulsion policies that best meet local needs and preferences. The administration offered no data or other information to justify a statewide policy in this area. We know of no reason why local expulsion policies are not appropriate in this case. Therefore, we recommend the Legislature reject the proposed trailer bill language and delete the $6.2 million for alternative placements.

Options for Prior-Year Funds

As we discussed above, the budget also proposes to spend a significant amount of Proposition 98 funds that are available in the current year. This includes "settle-up" monies from 1996-97 and 1997-98 as well as unspent Proposition 98 funds from previous budgets. Figure 13 (see next page) displays the Governor's budget proposal and our recommendations for the use of these additional funds.

Generally, we recommend the Legislature use the same principles discussed above to determine the expenditure of these one-time funds. We recommend approval of the increases proposed to pay program deficiencies and other commitments. We also recommend the Legislature approve the proposed increase in funding school district revenue limits caused by a reduction in estimated property taxes in 1996-97. We discuss our other recommendations below or in later sections of this chapter.

Figure 13
Prior-Year Proposition 98 Increases LAO and Governor's Budget Proposals
(In Millions)
	Governor's Proposal	Legislative Analyst's Proposal
Pay program deficiencies and other commitments
Revenue limit increases	$136.1	$136.1
Year-round schools deficiency	6.0	6.0
Deferred maintenance	-5.5	-5.5
Oxnard extended year pilot	4.1	4.1
Long Beach USD settlement	4.1	4.1
Mandates	--	4.1
Other	0.2	0.2
Review current programs consistent with long-term goals
PERS rate adjustment	--	$53.7
Adult education CalWORKs	$12.5	12.5
Support state improvement efforts
School site block grants	$180.0	--
School district block grants	--	$188.2
Digital high school	60.0	60.0
Test development	11.0	--
Standard account code structure	5.5	5.5
Single gender schools	3.0	3.0
Reject other proposed augmentations
Staff development--mathematics	$40.0	--
Community policing	10.0	--
Low performing schools	3.0	--
Teacher National Board Certification	1.0	--
Salary schedule pilot	1.0	--
Totals	$472.0	$472.0

Provide Block Grants to Districts

The budget proposes providing $180 million for school site block grants. Schools would be able to use this money for high-priority activities they identify. We support block grants as a means of providing one-time funds in the most flexible manner to schools.

We think, however, that the state should provide block grants to districts, not individual schools. School district officials have a broader view of the district's needs than those at the school sites and can better set funding priorities based on the district's overall needs. Therefore, we think the block grants could be used more effectively if they were distributed to districts.

In addition, data on the use of past district block grants show that a significant portion of the funds were passed on to school sites by many districts. (These data also show that districts used the block grant funds for deferred maintenance, computers, textbooks, and other one-time costs.) Therefore, it is likely that some portion of any future district block grants also would be distributed to school sites.

For these reasons, we recommend the Legislature change the proposed school site block grants into a district block grants. Because our other recommendations would free-up $8.2 million in one-time funds, we recommend the Legislature increase the block grants to $188.2 million.

Reject Other Policy Initiatives

The Department of Finance (DOF) and the Governor's Office of Child Development and Education (OCDE) were unable to provide any details on several of the policy initiatives included in the budget. As a result, we were unable to assess the costs and benefits for the Legislature of these proposals. Therefore, we recommend the Legislature delete the following funds:

Community Policing ($10 Million). These funds would provide one-time grants to local school districts and county offices of education that adopt a community policing approach to school safety. The administration advises that a legislative proposal is under development.

Low-Performing Schools ($3 million). The budgets states that these funds would be used to assist low-performing schools. As above, a legislative proposal is being developed by the administration.

The Legislature can consider funding such requests at such time as a complete proposal is before them.

Current law permits schools to include up to eight staff development days within the 180-day instruction year. Staff development days provide teachers time away from the classroom for planning and training activities designed to improve the quality of education at the school site. Figure 14 shows the number of instructional days provided by California's school districts in 1996-97 (not including staff development days). On average, schools include four to five staff development days within their 180-day school year.

The Governor's budget would substantially alter current law by (1) deleting the Education Code statute authorizing the use of up to eight days of instruction time for staff development, thereby requiring districts to provide at least 180 days of instruction time; and (2) using $400 million for an incentive program giving districts $220 per day for each teacher that participates in staff development activities outside the 180-day school year.

The budget proposal would also eliminate the $50 million staff development buy out program included in the 1997-98 Budget Act (the budget uses the $50 million to reduce the cost of the Governor's longer-year program). This program gives school districts $220 per teacher for one staff development day if they also add one day to the school year. The current-year program is not mandated. In contrast, the Governor's proposed plan would require that all districts offer at least 180 days.

Little Evidence to Support a Longer Year

Our review of the research on instructional time found little evidence to support increasing the school year as a means of increasing student achievement. Research suggests that, to maximize the impact of a longer school year on student achievement, schools also must introduce other changes so that the additional time increases the focus on learning. Even with additional changes, longer school years do not always improve student achievement. For example, Indiana adopted several school reform provisions in 1988, including a longer school year. After several years of evaluating student test scores and finding no benefit from the longer year, it rescinded its longer-year program.

The Oxnard Union High School District is conducting a pilot project that includes a longer school year. All students in the district attend school for 195 days. Besides adding days to the school year, the school introduced policies that maximize the instructional focus during the school day by decreasing disruptions to class time and reducing the number of days students miss class to participate in school functions (such as field trips, athletic events, and school productions). The budget proposes providing $4.2 million during 1998-99, or around $500 per teacher for each additional day of instruction, for this project. The project began in 1996-97 and will conclude in 1998-99. Administrators from the district report that an evaluation of the program will be available at the conclusion of the project.

Because we cannot find evidence to support the assertion that lengthening the school year will improve educational outcomes, we recommend the Legislature reject the Governor's longer-year proposal. Instead of focusing on time as the solution to improving schools, we think the Legislature should focus on quality. There are several proposals in the budget that already do this, including new and expanded programs to improve teachers through training and other support services.

We do think it would be useful to understand how different approaches to lengthening the school year affect student achievement. As discussed above, we recommend the Legislature begin demonstration projects in various areas including longer year and longer day (please see the K-12 Priorities discussion). Information from a demonstration project that lengthens the school year would help future decisionmakers to assess the potential benefits of such a proposal.

Legislature Has Other Options

• The Additional Staff Development Days May Require Some School Districts to Supplement State Funding With Other Local Funds.Since the proposed incentive grant amount is based on the average cost of teacher salaries, about half of the districts will receive more than they need (assuming districts pay teachers for these extra days at their average daily salary) and the other half will have less than they need to fund the additional days. As a result, some school districts may have to redirect local funds from other educational activities to pay for the actual district costs of adding staff development days to the school year.
• The Proposal Would Encourage Districts to Add Staff Development Days for Relatively Low-Priority Training. On average, districts currently use four or five of the eight staff development days allowed by statute. By providing eight days of funding for training, the proposal encourages districts--especially those districts with costs below $220 per teacher--to provide additional training that districts might otherwise consider unnecessary or low priority. As a result, we question whether the added staff development days would result in the most effective use of funds at the local level. Alternatively, giving districts flexibility over how they use the funds would allow them to focus resources on their highest-priority improvement activities.

In short, the proposed staff development program exhibits many of the characteristics of a categorical program that may reduce the effectiveness of spending at the local level. For this reason, if the Legislature wishes to lengthen the school year, we suggest it base a proposal on the following guidelines:

• Reduce or Eliminate Existing Staff Development Days. The Legislature could reduce the number of days allowed by statute or eliminate them entirely, as proposed by the Governor. By reducing the days, the Legislature can balance the goals of increasing the length of the school year, while also making some allowance for staff development.
• Provide Additional Funds for Staff Development Through Revenue Limit Increases. By providing general purpose funds to districts, the Legislature can maximize local flexibility over the use of available funds. This would allow districts to choose between staff development days and other school improvement activities. The flexibility also reduces the likelihood that districts would be required to spend local funds to supplement incentive funding provided by the state.
• Legislature's Policy Should Enable Schools to Provide a Minimum Level of Staff Development. The Legislature could eliminate the statutes that authorize the current eight staff development days without providing any new funding. In that case, however, it is likely that some schools would provide little or no resources for staff development days outside the 180-day year. For this reason, we think any longer-year proposal should provide the resources to enable schools to provide a minimum level of staff development. There is no reason, however, that the Legislature needs to ensure schools have eight days available for this purpose.

Building a longer-year program based on these guidelines would achieve the same goals as the Governor's proposal--extending the school year and providing funding for staff development. For instance, the Legislature could (1) allow two days of staff development within the 180-day school year (in effect, increasing the school year by up to six days); and (2) add to revenue limits the equivalent of two days of funding for staff development outside the school year. This approach would provide up to four days of staff development along with flexibility to use additional funds for each district's highest-priority improvement programs. We encourage the Legislature to consider these options if it decides to move forward with a longer school year proposal.

Three Agency Report on Revenue Limits

In our Analysis of the 1996-97 Budget Bill, we commented on a number of complexities with the state's revenue limit system. Based upon this analysis, the Legislature adopted language in the Supplemental Report of the 1996 Budget Act directing the State Department of Education (SDE), the Department of Finance (DOF), and Legislative Analyst's Office (LAO) to jointly review the revenue limit apportionment process and make recommendations to simplify it. The stated intent of the review is to (1) make the process more understandable and (2) reduce unnecessary workload at the state and local level.

The report, which was due to the Legislature by November 1, 1997, has taken longer than anticipated. However, we expect that a report containing recommendations based on the joint review of SDE, DOF, and LAO will be available during the spring. We will update the budget subcommittees on the status of the report during the hearings.

Include Public Employees' Retirement System Costs in Base Revenue Limits

The statutory revenue limit formula requires districts to complete many steps to determine what they will ultimately receive. One step in the calculation of revenue limits is the Public Employees' Retirement System (PERS) adjustment. The PERS adjustment reduces revenue limits to account for lower district PERS costs. The budget proposes to reduce revenue limits by $53 million in the current year and $59 million in the budget year due to a recent downward revision in district PERS costs.

How the Calculation Works. The PERS adjustment began in 1981-82. In that year the PERS rate paid by districts for their nonteacher employees was 13.02 percent. In 1982-83, when the rate declined to 12.045 percent, the Legislature recaptured the savings through the revenue limit process.

Over the years, the state continued to recapture savings from lower PERS rates. As the PERS rate and the number of covered employees within each district changes, the PERS adjustment changes. As a result, each year, districts make this adjustment to revenue limits. The PERS rate in 1997-98 is 6.2 percent.

We have identified the following problems with the PERS adjustment:

• It Complicates District Financial and Accounting Practices. From an accounting standpoint, the PERS adjustment is a nightmare. The district receives PERS funding from the state (based on the 1982-83 rate of 13.02 percent) and then returns a portion of the funds back to the state (the difference between 13.02 percent and 6.2 percent). This complex accounting makes it difficult for school boards and others that are interested in school finance to understand the PERS reduction--all of these complex transactions occur over revenue that is never seen by the district.
• PERS Cost Should Be Treated as a Regular District Cost. The PERS costs are not really different from other district costs. The state, however, does not adjust district funding levels to account for price changes for other specific educational inputs (such as maintenance, utilities, or payroll).

Include PERS Funding in Base Revenue Limits. During our three-agency discussions on the revenue limit process, our office has proposed to eliminate the PERS adjustment. In the future, districts would pay for any increases or benefit from any decreases in PERS costs. Our proposal would result in districts paying for PERS as they would any other district expense.

The PERS rate has decreased from 7.7 percent in 1996-97 to 6.2 percent in the current year, resulting in a PERS adjustment of $53 million. Due to rate reductions in 1998-99, the budget would reclaim an additional $6 million. Our proposal would allow districts, instead of the state, to reclaim the reduced costs of PERS since 1996-97. Our proposal allows districts to benefit immediately from decreased PERS rates and simplifies revenue limits. By doing this the state sends a message to districts that it is serious about making districts responsible for PERS no matter whether costs rise or fall.

For these reasons, we recommend the Legislature eliminate the PERS adjustment calculation by providing districts with their PERS rate as part of their base revenue limit. We further recommend an augmentation to revenue limits of $53 million in 1997-98 and $59 million in 1998-99 to fund these changes.

How Does Senate Bill 727 Affect Revenue Limits?

Currently, the statewide excused absence rate is around 4.3 percent of total ADA. During the first year of implementation, SB 727 holds districts harmless by increasing each district's revenue limit by a percentage equivalent to its excused absence rate. This calculation keeps the total amount of funds provided through revenue limits the same, but increases the per-student amount.

For instance, a school district with a per student revenue limit of $4,000 in 1996-97 that reports an excused absence rate of 5 percent will have (1) its revenue limit increased by 5 percent to $4,200 and (2) its ADA reduced by the same percentage. The district's total revenue limit--total ADA times the per-student revenue limit--remains unchanged.

In the budget year, schools that improve attendance will be able to increase their total revenue limit funding. If the above district reduces its absence rate from 5 percent to 4 percent in the future, it would receive an extra $4,200 for each additional student. This additional money could create a strong financial incentive for districts to ensure that students are actually attending school.

Senate Bill 727's Impact On the Deficit. Senate Bill 727 also eliminated other parts of the statutory revenue limit formula. For instance, it appears that SB 727 eliminates the revenue limit deficit factor. The DOF has interpreted SB 727 as eliminating the revenue limit deficit factor.

The deficit represents the amount that the state reduced statutory revenue limits by not providing cost-of-living adjustments during the recession years of the early 1990s. The deficit factor reduces revenue limits for school districts and county offices of education by a percentage that is approved as part of the annual budget process. The size of the deficit has decreased over the last few years because the Legislature has approved "deficit reduction" funding as part of past annual budget acts. Since 1994-95, the state has reduced the deficit factor from 11 percent to 8.8 percent as a result of additional funding.

In our 1995-96 Analysis of the Budget Bill, we recommended elimination of the deficit as a way to improve revenue limits for the following reasons:

• The Deficit Factor Has Been a Cumbersome and Confusing Part of Revenue Limits. The deficit factor adds an additional revenue limit--the "deficited revenue limit for each school district." Because not all funds provided through the revenue limit are subject to the deficit factor, keeping track of the deficit results in an overly complicated revenue limit calculation.
• Reporting the Deficit Does Not Serve Any Useful Purpose. In past years, the revenue limit deficit has served (1) to remind policymakers that the K-12 revenue limit falls short of the statutory standard and (2) as an argument for revenue limit increases. In our view, however, the reporting of this gap no longer serves any useful purpose as a budgetary benchmark. During our visits to school districts, they reported that, in the past, the deficit was a useful way to explain to school boards why revenue limits were not increasing. More recently, however, as revenue limit funding has increased, districts have viewed the deficit factor as a technical component of revenue limits.

Senate Bill 727 has some significant drafting problems. The administration's proposed trailer bill language would correct many of these problems. We are continuing our review of the language and will inform the subcommittees of any concerns during budget hearings.

The Deficit Still Applies to Some Programs

Senate Bill 727 does not eliminate the deficit factor for all programs. Programs such as summer school, community schools, and unemployment insurance costs still calculate funding using a deficit factor. The proposed budget includes $30.9 million and $22 million to "buy-out" the revenue limit deficits of school district and county office programs, respectively. This would eliminate virtually all deficits in K-12 programs.

While we think it makes sense to "wipe the slate clean" and eliminate all program deficits at this time, there are several ways to rebench revenue limits. The budget proposes one way--provide sufficient new funding. The deficit factor also can be eliminated without additional funds. This is the route used by SB 727 to eliminate the deficit for the rest of the revenue limit. The DOF provided no justification for why these district and county office programs merited additional funding.

Eliminate the Deficit--Save the Money. Eliminating the deficit factor for all programs makes sense. While SB 727 rebenches school district revenue limits, deficits in other programs remain. Eliminating the deficit factor for these programs will make the calculation of all revenue limit apportionments simpler and easier to administer. Accordingly, we recommend that the Legislature enact trailer bill legislation to eliminate the deficit factor used for county offices of education and various K-12 revenue limit calculations. Since no funding is needed to accomplish this, we also recommend deletion of $52.9 million proposed in the budget to buy-out these deficit factors.

Reopen Longer Day and Longer Year Incentives

Since 1984-85, the state has provided school districts incentive funds to extend the school year and day. We estimate that the proposed budget includes around $600 million for the longer day and longer year incentives. Almost all districts receive these funds. To qualify for the incentives, districts had to apply to the Department of Education between 1984-85 and 1986-87 and meet minimum standards, as follows:

• Longer Year. To qualify for longer year incentive money, a school needs to provide 180 days of instruction (this includes up to eight staff development days). Districts that qualify receive around an additional $45 per student.
• Longer Day. To qualify for the longer day incentive, a district must provide a minimum number of minutes of instruction each day. The minimum varies according to grade, but students in fourth through eighth grade must receive an average of 5.2 hours per day and high school students must receive at least 6.3 hours per day. The state provides around an additional $25 per student for grades K-8 and around $50 per student for grades 9-12.

There are about 20 districts that do not qualify for one or both of the incentives. We surveyed these districts and found that the majority of them now meet the requirements, but do not receive the adjustment to their revenue limit because they did not meet the statutory application deadline. Of the 20 districts that do not meet the requirements, most reported they would add the necessary minutes and days if the incentive were available to them.

We think the Legislature should allow these districts to receive the longer day and longer year incentives, for the following reasons:

• Excluding Some Districts From the Incentives Treats Districts Unfairly. The longer day and longer year incentives should reward all districts that add the required instructional time. More than ten years has passed since the application "window" closed. At this point--excluding a few districts from receiving the incentive funds serves no policy purpose--it only creates unnecessary funding inequities.
• The Costs of Reopening the Incentives Are Minimal. We estimate the costs of reopening the incentive will be around $225,000. This would be funded through revenue limits.

Conclusion. This is a small problem that is easily fixed. We find no analytical basis to exclude districts from receiving the incentive money if they meet the longer day and longer year requirements. Accordingly, we recommend that the Legislature adopt trailer bill language to reopen the longer day and longer year incentives to all school districts. We also recommend the Legislature add $225,000 to school districts revenue limits for this purpose.

Test of Applied Academic Skills Behind Schedule

State law calls for two tests to measure the achievement of K-12 students.

• Standardized Testing and Reporting. The Standardized Testing and Reporting (STAR) test, given to all students in grades 2-11 each year, is designed to provide ongoing information about students' academic skills to parents, teachers, and students. The 1998-99 budget proposes $30.4 million for the STAR test. The STAR test will be given for the first time in May 1998.
• Applied Skills. The test of applied academic skills is designed to assess students' ability to use academic skills to solve problems. Test items may include, for example, writing an essay in response to a question, conducting an experiment, or drawing a diagram to explain a concept. State law mandates that this test (1) be given statewide each year in grades 4, 5, 8, and 10, and (2) produce reliable school-level and district-level scores. The budget proposes $32.2 million for the applied test. Of this amount, $12 million is for test development and $20.2 million is for test administration. In addition, the 1997-98 Budget Act allows the department to carry over up to $11 million in unspent development funds into 1998-99.

Chapter 69, Statutes of 1996 (SB 430, Greene), postponed the development of the applied test until after SBE adopted content and performance standards. Content standards describe the knowledge and skills students should acquire in a given grade. Performance standards describe what students need to do to demonstrate they are "proficient" in the knowledge and skills outlined in the content standards. Chapter 69 required the standards commission to submit to SBE proposed content and performance standards in language arts and mathematics by October 1, 1997, and SBE to adopt a version of these standards by January 1, 1998. Assuming that standards would be adopted by this date, SDE projected that it could produce a statewide test of academic skills in language arts and mathematics by spring 1999.

Lack of Performance Standards Further Delays Test. The standards commission and SBE have approved content standards in language arts and mathematics. The commission, however, has not started work on performance standards. At the time this analysis was prepared, neither the commission nor the board could advise us of a date by which performance standards would be adopted.

Because the performance standards are not available, the statewide test of applied academic skills will not be ready by spring 1999. The department's time line for producing a test by that time requires SDE to award a contract for test development to a publisher by March 1, 1998. If a contract is not awarded by this date, there will be insufficient time to develop and field test enough items for a statewide test.

Allow Test Development to Go Forward. As matters now stand, we cannot advise the Legislature of a date by which it can reasonably expect the test to be ready. We see two options for the Legislature:

• Require the standards commission to turn its attention to developing performance standards in language arts and mathematics. This may delay development of content standards in science and social science on which the commission is currently working.
• Allow development of the performance standards and test questions to occur simultaneously.

We believe that allowing test development along with performance standards makes sense from both a timeliness and a quality perspective, for two reasons. First, it is unclear that the performance standards would be ready this time next year, even if the commission devoted full-time to their development. The standards commission and SBE have not established a time line or process for adopting performance standards. It is not known, therefore, when performance standards will finally be adopted. Holding up test development until they are adopted may push the eventual start date for the test past spring 2000.

Second, some experts in the field of assessment believe that performance standards are best developed along with a test of applied skills. In this scenario: (1) test questions are developed based on the content standards, (2) a representative sample of students answer these questions, and (3) examples of students' work are used to determine categories of performance (for instance "proficient" and "exemplary.") The SDE advises that it is confident that if joint development of test questions and performance standards begins by mid-1998, it can produce a test of applied academic skills by spring 2000.

We recommend, therefore, that the Legislature adopt trailer bill language allowing SDE to move forward with the development of test questions before performance standards are adopted. We also recommend that SDE, SBE, and the standards commission provide the Legislature, during budget hearings, a joint work plan describing how they will work together to adopt performance standards and develop a test by spring 2000. Finally, we recommend the Legislature delete $20.2 million for test administration to reflect the fact that the test will not be ready in the budget year. In addition, test development appears to be double-budgeted. The department advises us that it needs $12 million for development. Because this amount is in the budget and $11 million is available in carryover funds for this same purpose, we recommend the Legislature delete $11 million for test development.Identifying Low-Performing Title I Schools

We recommend the Legislature adopt budget bill language directing the State Department of Education to use data from the Standardized Testing and Reporting program test to identify the lowest-performing Title I schools, beginning with the spring 1998 test.

What Is Title I?

What Does Title I Require From California?

• Standards. The state must adopt content and performance standards in language arts and mathematics by fall 1997. Content standards describe the knowledge and skills students should acquire in a given grade. Performance standards describe what a student needs to do to demonstrate he or she is "proficient" in the knowledge and skills outlined in the content standards.

• Assessment and Accountability. The state must develop a statewide assessment and accountability system by fall 2000 that (1) is aligned with the state standards, (2) uses more than one measure to assess student performance (for example, standardized tests, writing samples, teacher evaluations), and (3) defines "adequate yearly progress" for schools. Schools not making adequate yearly progress for two years in a row are to be identified as "program improvement" schools and targeted for technical assistance from districts and the state.

• Low-Performing Schools. The state must establish alternative criteria to identify low-performing schools for program improvement until the statewide assessment system is in place. Title I calls these criteria a "transitional assessment system." States have great flexibility in designing transitional assessment systems--they do not, for instance, need to include multiple measures. Title I guidelines state, however, that the transitional assessment system must rely on accurate information about the academic progress of Title I schools.

Meeting Title I Requirements

State Out of Compliance on Standards. California is currently in the process of adopting academic standards. Chapter 828, Statutes of 1997 (SB 376, Alpert), requires SBE to adopt content and performance standards for language arts and mathematics, as recommended by the Commission for the Establishment of Academic Content and Performance Standards, by January 1, 1998. The board has adopted content standards for these subjects, but the standards commission has not yet developed performance standards for the board to review. As of this writing, neither the board nor the standards commission could identify a date by which performance standards would be adopted.

As a result, California is out of compliance with Title I requirements. Federal law says that states without content and performance standards by fall of 1997 may be required to adopt the standards of another state in order to continue receiving Title I money. The U.S. Department of Education has not yet decided to what extent it will enforce this provision.

The SDE Transitional System Uses Local Measures. During the transitional assessment period, the department has designed a system that uses data from district assessments of student performance. Specifically, the department requires districts to: (1) use multiple measures that are aligned with local standards to assess student performance; (2) combine the results of multiple measures in order to determine whether individual students are performing below, at, or above standards; and (3) report to SDE the percentages of students that are performing at or above standards. The department has decided that any school in which 60 percent or more of the students are below standard will be considered low-performing.

In order to monitor districts' assessment systems, the department required roughly 25 percent of districts (around 250) to submit in November 1997 descriptions of the "methods and procedures" they used to determine if students were meeting standards. The department is reviewing these descriptions and plans to use examples from them to guide districts in revising their assessment systems. The department has not yet prepared a descriptive summary of the 250 assessment systems it received from districts.

System for Identifying Low-Performing Schools Not Working

The department's approach to the transitional assessment system has some strong points. For instance, we think it was a good decision to include multiple measures in the transitional assessment system and to encourage districts to align their assessments with their standards. Title I does not require these things until fall of 2000, and the department's decision to require them now wisely gets districts started on refining the complicated assessment systems that will eventually be needed.

The department's transitional system, however, has serious shortcomings. The information generated by the transitional assessment system has little value from a statewide perspective. In addition, the system does not effectively identify low-performing schools. We discuss three major problems with the system in more detail below.

System Is Excessively Subjective. Based on our review of districts' assessment systems, we believe that SDE gains little useful or objective information when a district reports that a student is "meeting the standard." Three aspects of the system make it especially vulnerable to subjectivity. First, there is no consistency regarding the quality of local standards. For instance, the test score districts define as meeting the standard varies significantly. Four districts define scores above the 50 percent level as meeting the standard, four districts use 40 percent, three districts use 30 percent, and one district uses 20 percent.

Second, districts can combine scores from multiple measures in a way that allows relatively low-performing students to be considered as meeting standards. For instance, one district's method of combining scores results in classifying students as meeting the standard who (1) score in the bottom 5 percent on a nationally standardized test and (2) earn a "C" in their mathematics class. Another district's method weights a "proficient" score on the "classroom evidence" measure so heavily that a student can score a zero on the second measure and still be classified as meeting standards.

Third, the way class grades are used introduces subjectivity into the system. For instance, many districts defined a class grade of "C" as meeting the standard. The SDE did not require districts to explain the explicit connection between teachers' grading systems and districts' standards. As a consequence, it is not clear whether students earning a "C" grade have actually met local performance standards. Without this connection, grades can reflect the particular grading policies of teachers, schools, and districts as much as they reflect students' ability to meet a defined standard.

Low-Performing Schools Are Not Being Identified. Our review of district reports indicates that many districts are avoiding identifying low-performing schools by setting low standards. Six districts had assessment systems that defined what we considered to be poor performance--such as earning a class grade of "D" or scoring in the bottom 25 percent on a standardized test--as meeting standards. Not surprisingly, these six districts and four others did not identify a single low-performing school out of 169 total schools. These districts' assessments stand in contrast to one district that identified 42 "potential" low-performing schools out of 92 schools.

Problems Could Continue for Several Years. The department's plans regarding improving the transitional assessment system lead us to believe that the system may remain ineffective in identifying low-performing schools until at least the 2000-01 school year. Two points are most relevant. First, the department plans to take three years to review the assessment systems of the remaining 750 districts that have not submitted descriptions of their assessment systems (250 each year). This means that some districts may continue to "look the other way"--and some of the state's lowest-performing schools may remain unidentified--until the 2000-01 school year.

Second, SDE does not have a strategy for quickly and effectively correcting the inadequate assessment plans that have already been submitted. The department plans to (1) require districts that have submitted inadequate plans to submit revised plans in November 1998 and (2) issue guidelines to all districts regarding setting up assessment systems. This is the same strategy that produced the current system that does not work. The requirement for a transitional assessment system has been in place for three years, and the department has already issued guidelines to districts regarding Title I assessment systems. We have doubts, therefore, about how effective continuing this same strategy will be in improving the state's system for identifying low-performing Title I schools.

Long-Term, Short-Term Needs of Title I Assessment System

In the short term, we believe that the state must strengthen its Title I assessment system and get an immediate start on identifying the state's weakest schools. The department's system--with its subjectivity and the freedom it grants districts to define poor performance as adequate--makes it impossible for SDE to discern which schools are the lowest-performing and, therefore, which schools have the greatest need for assistance and monitoring from the department.

Use Data From STAR To Identify Low-Performing Schools

In addition, because local measures are an important component of a final assessment system, we believe the department should continue requiring districts to use multiple measures aligned with standards to measure the performance of schools. Schools, then, could be identified as low-performing either by STAR test scores or by district assessments of student performance.

Therefore, we recommend the adoption of budget bill language requiring SDE to use STAR results as one part of the state's Title 1 transitional assessment system as follows:

The State Department of Education (SDE), for the purposes of the transitional assessment system required by Title I of the Improving America's Schools Act of 1994, shall define a "program improvement school" as a school that meets at least one of the following conditions: (1) the school's score on the Standardized Testing and Reporting (STAR) program test ranks among the lowest in the state, or (2) 60 percent or more of the school's students are performing, as determined by the school's or its districts' assessment system, below the standards adopted by the district.

In addition, schools could be ranked using the STAR test in a variety of ways. For instance, schools could be ranked by their average student score or by the percentage of their students that score below a certain threshold. To assist the Legislature with its deliberations, we also recommend the budget subcommittees direct SDE to provide a recommended method of ranking schools based on STAR scores.