Donate
Independent, objective, nonpartisan research

R 1110MWR

Authors

R 1110MWR

Tagged with:

Publication PDFs

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3742 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(13) "R_1110MWR.pdf" ["wpmf_size"]=> string(6) "561187" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(108380) "Pathways for School Finance in California November 2010 Heather Rose, Jon Sonstelie, and Margaret Weston with contributions from Hans Johnson Supported with funding from The William and Flora Hewlett Foundation http://www.ppic.org/main/home.asp Pathways for School Finance in California 2 Summary This report demonstrates how California can improve its school finance system steadily over time as economic and demographic conditions permit. The improvements we suggest here are derived from our analysis of California’s current system using the following five principles :  Meet resource needs : Schools should have the resources necessary for their studen ts to meet state academic standards, and the cost of those resources may vary from school to school for a variety of reasons.  Structure incentives properly: The formulas allocating revenue to schools should not give schools incentives to deviate from actio ns in the best interest of students and taxpayers .  Allocate f unds transparently : The formulas for allocating revenue to schools should be clear and relatively simple.  Treat s imilar districts equitably : When the state has chosen the factors that determine t he revenue a school district receives, school districts with the same values for those factors sho uld receive the same revenue.  Balance s tate and local authority : Restrictions on the use of funds must properly balance the state objectives with the realities that schools differ widely across the state and that school administrators have unique knowledge about local conditions. California’s school finance system violates th ese principles in many ways. Under the current system, different district s are funded at different rates, a clear violation of horizontal equity. Unlike school finance systems in other large states, California does not adjust revenue to school districts based on regional differences in the cost of hiring employees, failing to recogn ize a large and obvious cost difference among districts. Because of the many state categorical programs directing revenue to public schools, the allocation of revenue to districts is not transparent, and the many restrictions on the use of funds in those p rograms unduly constrain local school administrators. Moreover, although California does provide additional funds for school districts with many economically disadvantaged students, the additional funds are not large enough to compensate for the differences in student need correlated with poverty. Our analysis also reveals several other areas in which California’s system could be improved. Making improvements without making some districts worse off would require additional revenue, which is now in short sup ply. However, as the economy improves, state tax revenue will rise, and the state can afford to invest again in its schools. At the same time, school enrollments are projected to rise relatively slowly, allowing an increase in revenue per pupil. This increase will not be dramatic, but it promises to be relatively steady, permitting the state to make slow and steady progress over time. To illustrat e the possibilities, this report simulates this process for a variety of potential improvements. One scenario eq ualizes funding rates for the main programs in the current system. In another scenario, funding is increased in districts with many economically disadvantaged students. A final scenario demonstrates the consequences of adjusting funding rates for regional differences in the cost of hiring personnel. Contents Summary 2 Figures 4 Tables 5 Introduction 6 School Finance Principles 7 Meet Resource Needs 7 Structure Incentives Properly 9 Allocate Funds Transparently 10 Treat Similar Districts Equitably 10 Balance State and Local Authority 10 Assessing California’s System 12 Revenue Limit Funding 13 Special Education 15 Economic Impact Aid 16 Other Categorical Programs 18 Simulating Pathways 21 Parameters 21 Scenarios 25 Conclusions 36 References 38 About the Authors 39 Acknowledgments 39 All technical appendices to this paper are available on the PPIC website: http://www.ppic.org/content/pubs/other/1110MWR_appendix.pdf http://www.ppic.org/main/home.asp Pathways for School Finance in California 4 Figures Figure 1 . Revenue limit base rates, 2009 –2010 .................................................................... 14 Figure 2 . Special education base rates, 2009 –2010 ............................................................ 16 Figure 3 . Economic Impact Aid f unding rates, 2009 –2010 ................................................... 18 Figure 4. Flex item funding rates, 2009 –2010 ..................................................................... 20 Figure 5 . Public school expenditures as a percent of state personal income, 1970 –2007 ......... 22 Figure 6 . Real personal income per t axpayer, 1970 –2008 .................................................. 23 Figure 7 . Taxpayers per s tudent, 1970–2007 ...................................................................... 24 Figure 8 . Status quo baseline ............................................................................................. 27 Figure 9 . Revenue limit focus scenario ............................................................................... 29 Figure 10 . Flex item focus scenario .................................................................................... 30 Figure 11. Economic Impact Aid focus scenario ................................................................... 31 Figure 12 . Program rates adjusted by regional wage index, 2009 –2010 ............................. 33 Figure 13 . Simulated pathway for revenue limit rates adjusted by regional wage index ............. 34 http://www.ppic.org/main/home.asp Pathways for School Finance in California 5 Tables Table 1 . School districts by type and size, 2009 –2010 ......................................................... 13 Table 2 . Projections of economic and demographic trends ............................................... 25 Table 3 . Summary of scenarios .......................................................................................... 32 Table 4. Average gains in revenue limit rates between 2009 and 2013 ($) ......................... 35 http://www.ppic.org/main/home.asp Pathways for School Finance in California 6 Introduction California’s bu dget crisis has diminish ed educational resources for the state’s current cohort of public school students . Because school districts have less revenue, class sizes are larg er and struggling students receiv e less assistance. Under these circumstances , it seems beside the point to suggest that California should begin plan ning for t he next cohort of students . Yet, history demonstrates that a failure to plan now will leave the state unprepared for what will surely follow . Although the current recession is deep, economic recovery will come , offering increas ing tax revenue and an opportunity for the state legislature to be more generous . How wi ll the state take advantage of this opportunity ? It may respond as it has in the past by allocating new revenue to schools for specific purposes . By 2005 –06, the state had more than 6 0 pro grams targeting a variety of purposes such as reducing class sizes, hiring counselors, purchasing textboo ks, and involving parents (Weston, Sonstelie, and Rose 2009 ). Altern atively, the state might use additional revenue to address underlying weaknesses in its school finance system . Our paper explores this alternative. We begin by discussing five broad principles for assessing California’s school finance system . Th ese principles do not lead to a single superior system, but they do suggest sev eral ways in which California could improve its current system . T hrough numerical s imulations, we illustrate the effects of pursuing some of th ese improvements steadily over time. The simulatio ns demonstrate that corrective long-term polic ies could significant ly strengthen California’s school finance system. http://www.ppic.org/main/home.asp Pathways for School Finance in California 7 School Finance Principles California’s school finance system is fundamentally different from the systems of most other states. In most state s, school districts have the power to set tax rates on real property. They have a robust source of discretionary local revenue. In California, school districts have limited taxing authority . They receive property tax revenue, but the state determines the amount they receive. 1 Because the instit utions governing schools are not the institutions financing them, conflict between the two is inevitable. Aligning these institutions should be a high priority . Our goal is this paper is more modest , however. We take as a given California’s current mixture of state finance and local governance and ask how California’s school finance system might be improved , given that mixture. We believe that five principles can be useful in guid ing this improvement . Yet, school districts are not agencies of the state. Each district has an elected school board that determines how its revenue is spent. Meet Resource N eeds We expect many things from our schools. Chief among th ese expectations is that students graduate from high school with a sound education. Over the p ast several years, California has spent considerable effort defining that education. The result is the state’s Academic Content Standards. The state has also implemented a battery of tests to determine whether students meet those standards. Although the tests are imperfect measures of knowledge and the s tandards do not include everything we expect students to l earn, a fundamental goal for any school finance system is to ensure that schools have the resources (teachers, textbooks, aides, counselors, and so on) necessary for their students to meet the state’s standards. B ecause the relationship between resources and academic achievement has not been firmly established, it is difficult to determine th ese resources with certainty . 2 Although these two highly regarded experts disagree on the conclus ions to be drawn from existing research, both agree that better research on the rel ationship between resources and achievement is needed. In particular, researchers need longitudinal data that tracks the academic improvement of individual students over tim e. Using such data , Rivkin, Hanushek, and Kain (2005 ) found that students in small classes did improve more rapidly than students in large classes. In that study, however, class sizes for individual students were determ ined through a process that the researchers did not explicitly account for in their analysis , raising concerns that class size might be related to unmeasured characteristics of students that also affect their achievement . The best response to th ese concerns would be an experiment in which students are randomly assigned to classes of different sizes. Using data from the onl y large experiment with class sizes, For example, several studies have focused on the rela tionship between class size and student achievement. After reviewing 59 of these studies, Hanushek (1997) judged the research inc onclusive: Most studies failed to find a statistically significant effect of class size on achievement, and the positive findings were offset by an equivalent number o f negative findings. Reviewing th e same set of studies, however, Krueger (2002) concluded that the research supports the belief that s maller class size lead s to higher student achie vement. 1 School districts may enact a parcel tax if it is approved by tw o-thirds of the voters. In 2005 –06, 98 districts had a parcel tax, representing 0.4 percent of school dist rict revenue. 2 Recent studies exploring the resource needs of California schools include Chambers, Levin, and DeLancey(2006); Imazeki (2006) ; and Sonstelie (2007). http://www.ppic.org/main/home.asp Pathways for School Finance in California 8 Krueger (1999) found that students in smaller classes did progress faster. Taking advantage of a natural experiment in whi ch class sizes were determined through a well-understood process th at was unlikely to be affected by unmeasured student characteristics , Angrist and Lavy (1999) reached the same conclusion. Examining data from a similar natural experiment, however, Hoxby ( 2000) found no significant effect of class size on achievement. In reviewing recent research, Angrist and Pischke conclude that reductions in class size do increase student achievement and that the estimated effects are consistent across studies. Reasonabl y well-identified studies from a number of advanced countries, at different grade levels and subjects, and for class sizes ranging anywhere from a few students to about 40, have produc ed estimates within a remarkably narrow band. 3 While w e agree that the best recent research tends to find a statistically significant relationship betwee n class size and student achievement , we do not believe this research is sufficient to give precise guidelines about the class sizes sufficient to achieve the state’ s academic standards . R educing class size is also very costly. A more efficient use of resources might be to focus on struggling students through interventions such as after -school tutoring or summer school ( Betts, Zau, and Koedel 2010 Recent research has also confirmed the importance of effective teaching (Rivkin, Hanushek, and Kain 2005) . This research suggests t hat effective teachers may have a more important influence on student achievement than reduction s in class size . Accordingly, i dentifying, recruiting, developing , and retaining s uch teachers should be a high priority for schools. From th is perspective , the most efficient use of a school’s revenue may be in providing the compensation and support that will attract and retain excellent teachers. Considering all the possible uses of school revenue, w e conclude that although the best research is consistent wi th a positive relationship between resources an d achievement, the parameters of th is relationship are not yet well understood. ). Howeve r, we are not aware of research on these inter ventions with the scale and statistical sophistication of the best recent research on class size. On the other hand, it is well understood that achievement varies dramatically among students provided with the same level of educ ational resources. Learning disabilities hinder the progress of some students. O thers may lack English language skills. P reparation, motivation, and aptitude may also be issues. To achieve gra de-level proficiency, some students may need additional attentio n from their teachers or after- school tutor ing. Because these needs are not uniformly distributed across sc hools, some schools will require more resources than others to meet the state’s standards. The cost of resources also varies across school districts in the state. More than half of a district’s budget consists of the salaries and benefits of its teachers. For the services of these and other employees, school districts must compete with other employers in local labor market s. As Rose and Sengupta (2007) show , the compensation offered by these employers differs significantly across regions of California, and thus the compensation of public school teachers also varies by region. In regions where other employers are offering relatively high salaries and benefits , school districts must do so also . To offer similar levels of educational services to their students, districts in high compensation regions must have higher revenue than similar districts in other regions. 3 Angrist and Pischke 2010, p. 24. http://www.ppic.org/main/home.asp Pathways for School Finance in California 9 Other costs may also vary across di stricts. As shown in Rose et al. ( 2008), the cost of transporting pupils to school is higher in rural areas . In the 100 districts with the lowest population density, transportation costs in 2003 –04 averaged more than $700 per pupil. In contrast, in the 300 districts with the highest population density, th e cost was about $100 per pupil. Utility costs also vary among districts, although not as widely as transportation cost s. Structure Incentives Properly In addressing cost differences, a school finance sys tem must not inadvertently reward districts for actions not in the best interests of students and taxpayers. For example, to account for differences in transportation costs, the state might reimburse school districts for the cost of transporting students t o school. Cost reimbursements would certainly neutralize cost differences across districts, but reimbursements would also remove any incentive school districts might have to control the costs of pupil transportation. This dilemma could be resolved, however, by using a measure of cost outside the control of school districts—for example, population density (see Rose et al. 2008), which is negatively correlated with transportation costs but independent of any action a district might take. The state could address the special needs of rural districts by allocating additional funds to districts with low population density.Districts would have flexibility in the use of these funds and thus have an incentive to use them wisely. This same concept applies to student achievement. Obviously, students who fall behind need additional instruction. However, if funds were allocated to school districts based on the share of students who fail to achieve proficiency on statewide tests, the districts with a lower share of profic ient students would receive more revenue than other districts, reducing funding for districts that were particularly successful in raising student achievement. As in the case of transportation, the resolution is to find a measure that is unaffected by dist rict actions but that is related to the likelihood that a student will fall behind. As many studies have shown, one such measure is the income of a student’s parents. Each year, the Census Bureau estimates the percentage of a district’s students living bel ow the federal poverty level. This measure is negatively correlated with student achievement, but it cannot be affected by any action taken by the district. Furthermore, this negative correlation exists within schools as well as across them, implying that the observed variation across schools cannot be solely due to a negative correlation between poverty and school effectiveness. Unfortunately, no external measure can precisely capture all of the differences i n cost across districts. For example, in the case of transportation, we might have two districts with the same population density; but in one, almost all students are concentrated in one town, while in the other, students are spread evenly throughout the d istrict. Transportation costs in the second district will be much higher than in the first, even though the population density is the same. Likewise, the percentage of students living in poverty is an imperfect indicator of average family income. Districts in which all of the families are just above the poverty line would be quite different from districts in which all of the families are well above the poverty line. These examples demonstrate that the principle of accounting for cost differences can conflic t with the principle of structuring incentives properly. A school finance system must find a balance between the two. And if a source of cost variation cannot be closely related to an external measure, it is hard to see how the school finance system can ta ke account of that cost without rewarding inefficiency. http://www.ppic.org/main/home.asp Pathways for School Finance in California 10 Allocate Funds Transparently Transparency is important in al l areas of government. The lack of transparency breeds distrust and undermines support for public institutions. Transparency is particularly important in the allocation of funds to school districts. Compared to other public services, the resources employed in public schools are clearly evident. Parents generally know the class sizes of their schools and the opportunities available to their children. At this level of the bureaucracy , public schools are relatively transparent , and parents rightly believe that they should be able to understand why resources differ across schools and districts , an understanding that ultimately requires them to k now why revenues var y across districts. Parents are more likely to understand why revenue s var y if the rules for allocating school funding are simple. Of course, t he simplest and thus most transparent rule is to a llocate revenue to districts in proportion to their enrollment , ignoring cost differences among districts. On the other hand , a set of rules for allocati ng revenue that account ed for every cost difference would be extremely complicated and thus not very transparent. These two extremes illustrate the tension between transparency and the recognition that costs are likely to differ across districts. This tension requires a pragmatic approach. If cost differences are small , they should be ignored. For example, a fter investigating the relationship between climate and utility costs, Rose et al. (2008) argue that variations in cost due to climate are not large enough to make climate a significant factor in all ocating revenue to schools. In contrast, regional salary difference s are large (Rose and Sengupta 2007) and should, in principle, be incorporated in a finance formula , although this would involve a number of complicated practical issues. Regardless of how regional boundaries are drawn, some adjacent districts would end up in different regions and thus receive different cost adjustments. With many regions, these differences would be small ; but nonetheless, a system involving many regions would be quite complicated, violating the principle of transparency . Other large stat es (Florida, New York, Texas) have been able to overcome this obstacle , however. Treat Similar Districts Equitably Consideration of costs, incentives, and transparency suggests a number of factors that might be used to allocat e revenue across school districts. For example, the factors might be average daily attendance, percentage of students living in poverty , and a regional wage index . Once the factors have been decided upon , every district with the same values for those factors should receive the same revenue, a concept sometimes referred to a s horizontal equity. Horizontal equity is closely related to transparency. If the law is clear about the factors to be considered in allocating revenue, then districts with the same values for th ose factors sho uld receive the same revenue. Balance State and Local Authority The four principles discussed above concern the allocation of revenue to school districts. Th is last principle concerns the conditions place d on the use of those funds. The bulk of revenue provided to schools is determined by the state legislature, which must weigh the needs of school districts against those of other state agencies and local governments. In this situation, it is only natural that the legislature may consi der some uses of school district revenue to have a higher priority than other s. In arguing for funds, school districts themselves will tend to empha size some uses, such as reducing class sizes, over other uses, such as http://www.ppic.org/main/home.asp Pathways for School Finance in California 11 hiring district administrators. Thus, i t should come as no surprise when the legislature places restrictions on how school districts are to use their funds. Many district administrators also favor some external restriction on the use of funds . In interviews with randomly selected s uperintendents, two advantages of such restrictions were commonly cited (Rose, Sonstelie, and Reinhard 2006). First, restrictions may protect funds from the collective bargaining process. For example, f unds to reduce class size require districts t o hire more teacher s rather than increas e the salaries of current teachers. Second, restrictions may thwart local political pressures. F or example, fu nds for disadva ntaged students require districts to provide additional resources to schools with large enrollments of these students. Some superintendents believed that w ithout such restrictions, locally elected school boards might tend to allocate resources equally to all schools, regardless of differences in need. On the other hand, California is home to a large and complicated K –12 system —more than 8,000 public schools with widely di fferent students and staff. Loc al administrators have considerable information about the strengths and weaknesses of their personnel and the abilities and backgrounds of their students, more than any central authority could have. The decentralization of knowledge argues for a decentralization of decisions about how revenue should be employed and for few restrictions on the use of funds (Brewer and Smith 2006) . Some have suggested that t he tension between state finance and local go vernance can be eased because of the recent emphasis on standards and accountability (Little Hoover Commission 2008) . The state has defined what it expects schools to achieve. It can therefore give sch ools more authority in how they achieve those objectives. It has defined outputs, so it can loosen its grip on inputs. Although this suggestion seems right in theory , it depends on the clear and fair measurement of outputs and on the power to hold district administrators accountable for meeting those measured objectives. Although California has made progress in measur ing student achievement , the state has only limited authority when it comes to holding local administrators accounta ble, a limit that is a natural consequence of local governance. It seems to us that the tension between state finance and local governance is unlikely to be resolved and that the restrictions that spring from that tension should be judged on a case -by -cas e basis. Some restrictions are clearly motivated by a difference in the objectives of the state legislature and those of locally elected school boards. Others, however, can only be rationalized by a difference of opinion about how a common objective is bes t pursued. In those case s, we believe the state should defer to local authorities. http://www.ppic.org/main/home.asp Pathways for School Finance in California 12 Assessing California’s System The five principles discussed above provide us with a lens for examining California’s current school finance system. Broadly speaking, the system has four funding components. The first is revenue limit funding, which combines local property tax revenue with state aid to generate a source of funds that school districts can use on any educational purpose. Revenue limit funding constitutes approximately 60 percent of the fundi ng received by school districts and forms the foundation of the state’s school finance system. The second component is a collection of programs that channel sta te aid to districts and place restrictions on how that aid is used. These programs , generally referred to as state categorical programs, constitute more than 20 percent of the revenue received by districts. The third component is a collection of federal categorical programs, const ituting approximately 10 percent of funding. The last component is discretionary local revenue such as parcel taxes and interest income. These local funds also constitute approximately 10 percent of district funding. In this discussion, we focus on the rev enues controlled by the state legislature—i.e., revenue limit funding and state categorical progr ams. And i n our analysis of state categorical programs, we are concerned about two programs in particular. The first is special education, which funds services for students with learning disabilit ies. The second is Economic Im pact Aid, which targets English learners and economically disadvantaged students. These two programs are the primary vehicles for addressing differences in studen t need. We analyze t he remaining categorical programs together as a single group . Our analysis does not include a consideration of the state’s approach to providing funds for students in sparsely populated areas. Although the provisions for necessary small schools and the pr ogram for pupil transportation recognize the higher cost of educating students in sparsely populated areas, the programs do not provide incentives for school districts to find efficient methods to educating students in these areas. When a school achieves necessary small school status, it has little incentive to merge with other schools, even if that merger would reduce costs without diminishing the education of students. Similarly, transportation funds are allocated according to historic costs, removing inc entives for districts to find cheaper solutions for educating students in sparsely populated areas. A program that allocated funds according to population density might address the needs of these areas without inadvertantly rewarding inefficiencies, a subj ect which deserves further study. In the follow ing discussion , we analyze revenue sources throught the lens of the principles of school finance described above. Different principles come into play in different areas, but one issue cuts across all areas. U nlike other large states such as Florida, New York , and Texas, California does not adjust revenue in any of its programs for regional salary differences . The state ignores the very large variations in the costs of the most important resource school distric ts employ, a clear violation of our first principle. Ou r analysis presents funding rates (dollars per pupil) for school districts. For our purpose s, districts are separated into nine groups based on their type (elementary, high school, and unified) and siz e (small, medium , and large). We chose the size partitions to yield a roughly equal number of districts in the three size classifications for each district t ype. Our analysis excludes necessary small schools and charter schools. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools. The schools and districts we do include represent 95 percent http://www.ppic.org/main/home.asp Pathways for School Finance in California 13 of California’s public school enrollment. As Table 1 shows, large un ified districts include nearly 60 percent of these students. 4 TABLE 1 Type and size of district School d istricts by t ype and size, 2009– 2010 Number of d istricts Average daily attendance (ADA) Percent of total ADA Elementary Small (0 –250) 132 17,474 0.3 Medium (251–1,500) 171 110,693 2.0 Large (1,501+) 175 969,368 17.5 High s chool Small (0 –1,500) 23 19,607 0.4 Medium (1,501–6,000) 26 82,848 1.5 Large (6,001+) 31 444,893 8.0 Unified Small (0 –3,000) 120 151,138 2.7 Medium (3,001–10,000) 96 556,209 10.2 Large (10,001+) 110 3,181,060 57.4 All districts 884 5,543,291 100.0 SOURCE: 2009 Principal Apportionment, California Department of Education. NOTE: Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools. Revenue Limit Funding A simple formula determines a school district’s revenue limit funding . Every district has a base rate, a dollar amount per pupil. That base rate is multiplied by the district’s average daily attendance (ADA) to determine its total funding entitlement. This entitlement is met through local property taxes and state aid. Students in necessary small schools are funded through a different formula. In addition, the calculation of a district’s revenue limit entitlement involves several other adjustments that generally stem from policy decisions made over the years. As Weston (2010a) shows, these adjustments do not contribute much to the variation in revenue limit funding per pupil across school districts in California. The biggest source of variation stems from differences in the base rate among districts. These variations are represented in Figure 1. The boxes show the distance between the base rate in the 75 th and 25 th percentile for a group. Percentiles are weighted by the number of students in a district. Within each group, students are assigned the base rate of their district and ranked according to this rate. The 75 th percentile is the base rate of the student in the 75 th percentile of this ranking. The upper light part of each box is the distance between the median base rate and the base rate in the 75 th percentile. The v ertical lines show the distances between the 10 th and 90th 4 The percentile. Each group also has three horizontal hash marks above the box and three below it. These marks show the highest three and the lowest three base rates in each group. When the two or more of the individual base rates are nearly identical, the hash marks for those rates are indistinguishable and appear as just one mark. For example, large unified districts appear to have two hash marks above the box because the second and third highest base rates are nearly the same. For almost every group, the bottom three hash marks are very close or indistinguishable. technical appendix provides more details on students and districts excluded in our analysis. http://www.ppic.org/main/home.asp Pathways for School Finance in California 14 FIGURE 1 Revenue l imit base r ates, 2009– 2010 SOURCE: 2009 Principal Apportionment, California Department of Education. NOTE: Base rat es are statutory rates for 2009 –10, with 18.355 deficit factor . Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools. Base rates vary significantly across groups. Some of the vari ation is due to differences between high school districts and other districts. The median rate for each of the three high school groups is approximately $6,000 per pupil. For the elementary groups, all three median s are close to $5,000 per pupil. The medians for the medium and large unified districts are about $5,200. For the small unified districts, the median is $5,517. As the figure demonstrates, t here are also variations withi n groups, primarily among the small elementary and small unified groups . This variation clear ly violat es the principle of horizontal equity. These variations reflect a historical process. When the revenue limit system was first introduced in 1973, a distr ict’s base rate was its expenditures per pupil in 1972–73. Over time, the state has gradually raised the lowest base rates. To determine the relative position of base rates, districts were classified by type and size. Equalization reduced differences withi n groups but did not necessarily reduce differences across groups. The higher average rates of high school districts are sometimes justified by the notion that high schools are more expensive to operate than other schools. Although this may be the case, th e differences in base rates among district types are not an explicit state policy , and the research that might justify these differences is not conclusive (Sonstelie 2007) . Furthermore, if state policy did mandate higher funding rates for high school students, the base rates of unified districts should reflect the percentage of their students attending high school. 4,500 4,750 5,000 5,250 5,500 5,750 6,000 6,250 6,500 6,750 7,000 7,250 7,500 7,750 8,000 Small ElementaryMedium ElementaryLarge ElementarySmall High SchoolMedium High SchoolLarge High SchoolSmall UnifiedMedium UnifiedLarge Unified Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 15 On occasion, the state has increased variations in base rates as it introduced new policies or phased out old ones . In 1997, it changed the methods for calculating the ADA of a district. To offset the negative effects this change had on some districts, the state increased their base rates . 5 Special Education The rate differences it created by this increase are temporary , however, because of the formula for an nually adjusting base rates. Each year, the state calculates an amount per pupil necessary to a djust the average base rate for inflation. Different amounts are calculated for each type of district: high school, elementary , and unified. The state then increa ses the base rate of every district by th e amount of this inflation increase . For districts with base rates above the average, this adjustment is not high enough to keep up with inflation. The base rate of these districts falls in real terms. For districts with base rates below the average, real rates rise. When inflation is low , as it has been for some time, these equalizing changes are small. With this policy, it will take many years to equalize base rates. The state allocates funding f or special education through special education local planning areas (SELPA). SELPAs are groups of districts, county offices of education, and charter schools that agree to share special education funding and services. Over 90 percent of this funding is allocated through a simple formula. Each SELPA has a base rate, expressed in dollars per ADA. Its entitlement is this rate multiplied by its ADA. This entitlement is met through property taxes and federal and state aid. This formula was created by Assembl y Bill 602 in 1997. Before that time, special education funding was allocated according to the costs and needs of special education students within each SELPA, creating a fiscal incentive for districts to classify students as disabled. The new bill ended t his incentive by allocating special education funds according to the attendance of all students, not just special education students. Although the current formula is consistent with the principle of structuring incentives properly, it does raise the question of whether the variations in special education costs among districts are adequately addressed, the first of our principles. These variations are partly addressed by three separate funding sources for relatively rare, but severe, disabilities. One provi des additional revenue for districts that must purchase special materials and equipment, and the other two fund the placement of students in special facilities. The funding formula also includes a Special Disability Adjustment that provides additional fund s for SELPAs that had unusually high incidences of learning disabilities when the new formula was created in 1997. These adjustments and additional funding sources certainly address some of the cost differences across districts. However, a fundamental question remains: Is the incidence of learning disabilities randomly distributed across districts? Recent research indicates that this is not the case. Using data from a large survey of families, Lipscomb (2009) found that the incidence of severe disabilities among children is negatively correlated with family income. This finding suggests that the formula for allocating special education revenue among SELPAs should include poverty rates as well as ADA. Because poverty rates are outside a district’s control, t his change would address cost differences among districts without creating incentives for districts to identify students as learning disabled. Those possible changes notwithstanding, the current allocation of special education revenue clearly violates the p rinciple of horizontal equity. As Figure 2 shows, base rates for special education vary across districts. 5 For details, see Weston 2010, p. 11. http://www.ppic.org/main/home.asp Pathways for School Finance in California 16 For example, for large unified districts , the rate in the 75th percentile is 10 percent higher than the rate in the 25 th A second issue is whether the level of special education funding is adequate overall to meet the cost of that education. In 2006– 07, California school districts spent more than twice as much on special education services as they received in special education revenue (Lip scomb 2009). In the jargon of school finance, special education services “encroached” on general education services. If special education revenue were increased substantially, however, some districts would surely have more revenue for special education ser vices than they otherwise might spend on those services, creating incentives for districts to identify learning disabilities and to spend too generously on special education services. To us, encroachment is a much less serious issue than is the allocation o f existing revenue according to need. percentile. FIGURE 2 Special e ducation base rates , 2009–2010 SOURCE: 2009 Principal Apportionment, California Department of Education. NOTE: The rate for each district is the rate of the SELPA to which it belongs. Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools.. Economic Impact Aid Economic Impact Aid (EIA) funds supplemental services for English learners and economically disadvantaged students. Each district’s entitlement is determined by multiplying its EIA rate by a weighted count of eligible students. The count starts with the number of English learne rs in the district plus the 500 550 600 650 700 750 800 850 900 950 1,000 Small ElementaryMedium ElementaryLarge ElementarySmall High SchoolMedium High SchoolLarge High SchoolSmall UnifiedMedium UnifiedLarge Unified Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 17 district’s count for the federal Title I program, a Census-based estimate of the number of students living in poverty. Every student in this count in excess of 50 percent of all students receives a weight of 1.5. Every other student receives a weight of unity. The EIA count is the sum of these weights. The EIA program is the primary mechanism through which California addresses differences in student need correlated with family income and native language. The EIA formula places a heavy weight on English learners because 85 percent of English learners are also economically disadvantaged (Legislative Analyst’s Office 2007). Accordingly, most English learners generate twice as much EIA revenue for a district as an economically disadv antaged student who is fluent in English. 6 Including English learners in the base for Economic Impact Aid also creates incenti ves for districts that are not consistent with state goals. In a district with a particularly effective program for English learners, students move relatively quickly to fluency, and the district receives less revenue than it would if students were slower to make this transition. After a review of several studies of the resource needs of English learners, Gandara and Rumberger (2006) question the implicit assumption underlying this revenue premium. They argue that although English learners may need different services than economically disadvantaged students who are fluent in English, the cost of additional services may be similar for both groups of students. Like revenue limits and special education funding , EIA funding rates vary across distri cts, violating the principle of horizontal equity. Figure 3 shows these variations. The median funding rate for every group is close to $300 per pupil. However, f or every group except large high sc hool districts, the funding rate at the 7 5 th percentile is at least 10 percent higher than the rate at the 25th The level of funding is also an issue. EIA provides funding for additi onal student needs correlated with poverty and lack of English fluency. As argued above, we do not believe that cur rent research provides precise estimates of the relationship between school resources and student achievement . Consequently , that research cannot tell us how much additional revenue schools will need because they have high percentages of English learners or economically disadvantaged students. In this circumstance , we believe the best guidance comes from educational practitioners. On the basis of budget exercises with over 500 randomly selected California teachers, principals , and superintendents, Sonstelie (2007) concluded that to meet the state’s academic standards, a school in which every student was economically disadvantaged would need abou t $1,200 per pupil more in 2003– 04 than a school in which no student was disadvantaged. Adjusting for inflation, this figure would be about $1,500 per pupil in 2009 dollars. Part of this gap is closed by federal Title I funds , which target economically dis advantaged students. These funds are allocated to the state, which then allocates them to school districts. The process is complicated, b ut, on average, districts received $450 per di sadvantaged student in 2009 –10. Accordingly, based on the expertise of educational practitioners, EIA funding rates should be about $1,050 per disadvantaged student, instead of the current rate of about $300 per student. percentile 6 About 60 percent of disadvantaged students are fluent in English (Legislative Analyst’s Office 2007). http://www.ppic.org/main/home.asp Pathways for School Finance in California 18 FIGURE 3 Economic Impact Aid funding rates, 2009–2010 SOURCE: 2009 Economic Impact Aid Funding Results, California Department of Education. NOTE: Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in wh ich all schools are charter schools. Other Categorical Programs The most recent description of California’s categorical programs is presented in Weston, Sonstelie, and Rose (2009). Th is report describes each of the state’s more than 60 programs and provides data on the allocation of revenu es in each program for the 2005 –06 academic year. The largest program that year was special education. Economic Impact Aid was the fifth largest. In order of expenditure, t he other programs among the five largest were K –3 Class Size Reduction, the Targeted Instruction Improvement Block Grant, and Adult Education. Each state categorical program has its own funding procedures and restrictions. Although each program has its rationale, the sheer number of programs raises two key issues. First, because the funding procedures for programs can be complicated and because these procedures vary from program to program, it is very difficult to determine why one district receives more or less categorical revenue than another. Taken as a whole, the allocation of categorical revenue is not transparent, violating one of our principles. Second, the cumulative effect of program restrictions may have tipped the balance too far in the direction of state control over the use of funds. Th e latter issue received new attention in the Budget Act of 2009. In an effort to give local administrators more flexibility to absorb revenue cuts, the legislature granted spending flexibility for approximately 40 0 500 1,000 1,500 2,000 2,500 3,000 3,500 Small ElementaryMedium ElementaryLarge ElementarySmall High SchoolMedium High SchoolLarge High SchoolSmall UnifiedMedium UnifiedLarge Unified Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 19 categorical programs.7 A recent survey shows how districts have responded to this flexibility (Legislative Analyst’s Office 2010). Most districts reported shifting funds away from uses financed by programs in the flex item. In general, these funds have been shifted tow ard core classroom instruction. For programs in this so -called flex item, the legislature suspended restrictions on the use of funding, making these programs unrestricted general support. Special education and Economic Imp act Aid are not included in the fl ex item. This suspension of funding rest rictions continues through 2012 –13. At that point, the legislature could re - impose the restrictions it suspended. Alternatively, it could make the suspension permanent, in effect turning the pr ograms in the flex item into a source of unrestricted revenue similar to revenue limit funds. By doing so, the legislature would be taking a large step toward decentralizing decisions about how funds are spent. Before the legislature took this step, i t would surely review the composition of programs in the flex item. Additional programs might reasonably be added, and some programs now in the flex item might just as reasonably be excluded. Adult Education is a good example. In some areas of the state, commu nity colleges have assumed the primary role in providing adult education. In other areas, school districts have assumed this role. If the Adult Education program were turned into unrestricted revenue, adult education might suffer in the lat ter areas, but s chool districts in those areas would have a new source of unrestricted revenue not enjoyed by other districts. Other examples are the Regional Occupational Program and the Teacher Credentialing Block Grant, both of which support regional structures serving many school districts. A more general discussion of categorical flexibility might also consider changes to some programs presently excluded from the flex item. For example, the Targeted Instruction Improvement Block Grant, which is currently included in the flex item, might instead be consolidated with Economic Impact Aid, as was recently done with the English Language Acquisition Program in the 2010 Budget Act. If the legislature were to turn some version of the flex item into a permanent source of unrestricted revenue, it would also be forced to confront another issue. The revenue that districts receive from the flex item would no longer have a clear rationale, a reason why one district receives more or less revenue than another. As Figure 4 shows, funding rates for the flex item differ considerably across districts. 8 The median rates range from $711 per ADA for large elementary districts to $893 for medium high school districts. There are also large variations in funding rates within groups of districts. The difference between the rate in the 75 th percentile and the rate in the 25 th 7 For more details, see Weston, forthcoming. A list of programs is provided in the percentile is $566 per ADA for large unified districts. The difference is more than 20 percent of the median rate in every group. If the flex item were to become unrestricted support, it would clearly violate the principle of horizontal equity. technical appendix, Table A2. 8 Although K –3 Class Size Reduction (K –3 CSR) was not in the flex item, we have included 70 percent of revenue in that program in the flex item. The Budget Act of 2009 allowed districts to retain at least 80 percent of their class size reduction funds as lo ng as class sizes do not exceed 25 students. Even if class sizes exceed 25 students, districts retain 70 percent of their previous funds. Thus, at the very leas t, 70 percent of the previous year funding for K –3 CSR ought to be considered part of a district ’s flex item, a practice we follow in the data presented in Figur e 4. http://www.ppic.org/main/home.asp Pathways for School Finance in California 20 FIGURE 4 Flex item funding rates , 2009–2010 SOURCE: Funding Results (various programs) and 2009 Principal Apportionment , California Department of Education, Deferred Maintenance Program funding, Office of Public School Construction. NOTE: Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools. 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000 8,500 9,000 9,500 10,000 10,500 11,000 Small ElementaryMedium ElementaryLarge ElementarySmall High SchoolMedium High SchoolLarge High SchoolSmall UnifiedMedium UnifiedLarge Unified Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 21 Simulating Pathways Our assessment of California’s school finance system has identified sever al potential improvements. Each would require additi onal funds if districts w ere held harmless as improvements were implemented. For example, horizontal equity requires funding rates to be equal ized. Equalization could be achieve d without additiona l revenue by reducing the rates above the average while raising rates below the average. However, l eveling down may not be politically feasible . A mor e acceptable approach would be to increase low rates more rapidly tha n high rates over time as the state invests in its public schools . In this section, we simulate that process using a number of assumptions about fut ure growth in revenue and enrollments. We have chos en four improvements to simulate , illustrating a range of possibilities. The first is to equalize funding rates for revenue limits. The second scenario simulates the process of turning the flex item into a source of unrestricted aid and equalizing funding rates over time . The third scenario increase s the EIA funding rate to $1,050 per student. Over time, the finance s ystem in this scenario would resemble the recommendation of the Governor’s Committee on Education Excellence (2007). The fourth scenario adjust s program rates for regional cost differences. The endpoint in this scenario is a school finance system similar to that recommended by Bersin, Kirst, and Liu (2008) . Parameters The simulations begin with a baseline in 2009– 10. School districts receive the revenue in the simulation that they actually received in that year from revenue limits, special education, Economic Impact Aid, and th e flex item. Th e funds in these programs total ed $39 billion in 2009–10. Each subsequent year in the simulation, the total increases as economic and demograph ic conditions permit, and the additional revenue allows for increases in each district’s funding rates for each of the four programs. T he additional revenue comes from normal economic growth and from demographic trends likely to unfold over the next 2 0 years. Growth in school revenue is tied to economic growth and demographic trends through the following equation: The equation isolates two important factors in the state’s ability to provide revenue for its public schools. The first is a demographic fac tor, the number of taxpayers per student. A given tax burden for the average taxpayer will produce more revenue per pupil for schools if there are more taxpayers per student. The second factor, average income per taxpayer, represents the ability of taxpayers to bear a given tax burden. Below, we investigate trends since 1970 in each of t he three terms in the equation above. Our objective is t o identify reasonable assumptions for our simulations and to put those assumptions in perspective. We conduct our simulations in real terms, adjusting for inflati on. Because our focus is the purchasing power of school districts , we use the Implicit Price Deflator for State and Local Government to deflate nominal figures for bot h school expenditures and personal income. http://www.ppic.org/main/home.asp Pathways for School Finance in California 22 The first term in the equation is public school expenditures in California as a percentage of state personal income. We do not have data on revenue in the four funding programs extending back in time. The best data we have is the current e xpenditures of public schools, which is approximately equal to total revenue schools received for operating expenses. The revenue in the four programs we focus on is currently about three- fourths of that total. In the early 1970s, public school expenditures were over 4 percent of state personal income (Figure 5). After Proposition 13 in 1978, that percentage fell steadily, reaching a low of 3 percent in 1984. It has rebounded steadily since that time with a high near 4 percent in 2003. In 2006–07, current expenditures for California public schools were 3.8 percent of state person al income. From 1980–81 to 2006– 07, that ratio has averaged 3.6 percent. FIGURE 5 Public school expenditures as a percent of state personal income, 1970 –2007 SOURCE: Public school expenditures in California are from the National Center for Education Statistics, State Education Facts , 1969– 1987, and Digest of Education Statistics, 1998, 2001, and 2009. State personal income is from the Bureau of Economic Analysis, Department of Commerce. NOTE: Public school expenditures are current expenditures for elementary and secondary education. The expenditure data in Figure 5 comes from the National Center of Education Statistics . Although the Center provides a consistent and reliable source of data extending back for many years, the most recent data are for 2006– 07. However, w e can update th ese data from other sources. From the data in the Governor’s Budget, revenue limit funds pl us state categorical revenue fell about 11 percent from 2006 –07 to 2009 –10. Some of the reduction in funds was replaced by a temporary increase in federal revenue. Over the same period, personal income in California rose by about 4.6 percent. Ignoring the temporary increase in federal revenue, the ratio of current expenditures to personal income fell to approximately 3.2 in 2009 –10. The first assumption for our simulation is that this ratio will return gradually to its average level of 3.6 percent and that revenue in the four programs will rise by the same percenta ge. In particular, as total revenue rises from 3.2 percent of personal income to 3.6 percent of personal income, revenue in the four programs is assumed to rise from its current value of 2.5 percent of personal income to 2.8 percent of personal income . In the simulations, this increase is phased in uniformly ov er time with the ratio increasing by 0.014 percentage points each year until it reaches 2.8 percent in 2030. 0% 1% 2% 3% 4% 5% Expenditures as Percentage of Personal Income http://www.ppic.org/main/home.asp Pathways for School Finance in California 23 We believe this assumption is relatively conservative for four reasons. First , in computing the average ratio of exp enditures to personal income, we excluded the years before Proposition 13, in which the rate was significantly higher than in subsequent years. Second, as noted by the California Budget Project (2010), public school expend itures as a percentage of personal income have been lower in California than in the rest of the nation for many years. Our simulations assume that in the case of K –12 education, California continues to be a relati vely low-spending state. Third , for every y ear in the simulation except the last, the ratio of revenue to personal income is below its average from 1980 to 2006. Fourth, we assume that revenue in the four programs is the same fraction of total revenue as it is currently. Revenue for programs we have not included, such as necessary small schools, is implicitly assumed to grow at the same rate as revenue for the programs we have included. With this assumption, the growth in personal income is an important factor in determining the growth in revenue. To separate economic and demographic factors, we have decomposed personal income per student into two parts: personal income per taxpayer and taxpayers per student. We have used age to partition California residents into taxpayers and others, including stu dents. Taxpayers are residents 18 years and older. In general, younger residents are either in school or younger than school age. As Figure 6 shows, personal income per taxpayer has grown steadily since 1970, punctuated by economic recessions in 1981, 1990, 2001, and 2008. In the figure, personal income is expressed in real terms, using the Implicit Deflator for State and Local Government to adjust for inflation. With that adjustment, real personal income per taxpayer has grown at an average rate of 0.5 percent since 1970. The dashed line in the figure represents this average growth rate. FIGURE 6 Real personal income per taxpayer , 1970–2008 SOURCE: State Personal Income and the Implicit Price Deflator for State and Local Government is from the Bureau of Economic Analysis, Department of Commerce. Data on California residents 18 years of age and older are from the State of California, Department of Finance, Race/Ethnic Population with Age and Sex Detail, 1970 –1989, 1990– 1999, and 2000 – NOTE: Taxpayers are California residents 18 years of age and older. State Personal Income is deflated by the Implicit Price Deflator for State and Local Governments. 2050. In our simulation s, we assume that real personal income per taxpayer continues to grow at 0.5 percent per year. To be clear, this assumption is not a forecast of economic growth in California. Our purpose is not to 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 Personal Income per Taxpayer (2009 $) http://www.ppic.org/main/home.asp Pathways for School Finance in California 24 forecast , but to work through the long -term consequences of certai n policies using realistic assumptions. Fo r that purpose, we believe our assumed growth rate is appropriate. By adjusting for inflation using the Implicit Deflator for State and Local Government instead of the Consumer Price Index, we are also accounting for the fact that the salaries of state and local government workers has tended to grow faster than the general price level. In that sense, our assumed growth rate represents the growth in personal income in excess of that necessary to maintain a constant level of public education services. Demographic trends unfold more smoothly than economic trends over time. As shown in Figure 7, the number of taxpayers per student rose steadily in California from 1970 through 1986. In 1970, there were slightly less than three taxpayers per student. By 1986, that ratio had risen to 4.65, a significan t decline in the tax burden public schools placed on taxpayers. From 1986 to 1999, the ratio fell to a low of 4.08. It has risen slightly since then. FIGURE 7 Taxpayers per student, 1970– 2007 SOURCE: Data on California residents 18 years of age and older are from the State of California, Department of Finance, Race/Ethnic Population with Age and Sex Detail, 1970–1989, 1990 –1999, and 2000 –2050. Data on public NOTE: Taxpayers are California residents 18 years of age and older. Students are students enrolled in California public schools . school enrollments in California are from the Nation al Center for Education Statistics, State Education Facts , 1969 –1987 , and Digest of Education Statistics, 1998, 2001, and 2009. Because demographic trends are tied to the aging of the existing population, they are more predictable than economic trends. According to our projections based on California Department of Finance data, school enrollment should rise by about 20 percent ov er the next twenty years, while the population of California adults over age 18 rises almost 30 percent. 9 9 Although much of this growth will be among older adults ages 65 to 74, seniors in California tend to have good economic outco mes. For example, poverty rates for 65 - to 74 -year -olds in California are lower than for any other age group (7.9% in 2008 compared to 13.6% for all other age groups, based on American Communtiy Survey data). Of course, use of public health programs is higher at older ages, but f ederal programs provide most of tha t support. Thus , taxpayers per student should rise somewhat. 0 1 2 3 4 5 Taxpayers per Student http://www.ppic.org/main/home.asp Pathways for School Finance in California 25 Our simulation uses enrollment projections from the Department of Finance to project attendance and EIA counts for each school district. For years beyond the base year, we project ADA by applying growth rates to the base year ADA. These growth rates are based on county -level enrollment derived from population projections by the Department of Finance. For the EIA student count, which is not AD A, we use the EIA count in 2008 –09 as a percentage of ADA in the district. Table 2 summarizes our economic and demographic assumptions. The implication of thes e assumptions is a 31 percent increase in real expenditures per student. They rise from $7,022 in 2009 to $9,206 in 2030. TABLE 2 Year Projections of economic and demographic trends Taxpayers Personal income per taxpayer (2009$) Average daily attendance Taxpayers per student Revenue per student s (2009$) 2009 28,695,960 54,516 5,543, 291 5.18 7,022 2010 29,146,279 54,789 5,546, 543 5.25 7,199 2015 31,312,124 57,583 5,630, 917 5.56 8,007 2020 33,244,039 60,521 5,896, 815 5.64 8,532 2025 35,157,589 63,608 6,329, 462 5.55 8,835 2030 37,076,944 66,852 6,732, 870 5.51 9,206 This growth is consistent with the basic tenets of Proposition 98, which provides a minimum guarantee for revenue in California public schools and community colleges. The guarantee involves several complicated conditions, but the central condition is that, in normal economic times, revenue per student should grow at least as fast as per capita personal income. In our simulations, the growth rate for revenue per student exceeds this Prop 98 growth rate initially as taxpayers per student ri se s. It then falls below the Prop 98 growth rate for the remaining years. In 2030, revenue per student in our simulation is approximately $500 less per student than if revenue per student had grown at the rate of personal income from 2009 to 2030. Scenarios Each scenario involves a different allocation of additional revenue. In the base year of 2009 –10 , the revenue each district received in each of the four programs we focus on can be expressed as a funding rate (dollars per student) multiplied by a particular count of students in the district. The simulations change those per - pupil funding rates each year. We describe four scenarios below which reflect a range of state priorities. To provide a baseline for these scenarios, we also demonstrate how funding rates change over the next 20 years if the state relies solely on its current mechanisms of adjusting revenue rates for inflation. Status Quo In this status quo baseline, each district’s statutory revenue limi t rate increases annually for inflation. As described earlier, this inflation increase is the same dollar amount for all districts of the same type, but the increases vary in proportional terms , depending on whether the district’s rev enue limit is above or below the average for its type. In simulating these increases, we have assumed an inflation rate of 4.83 percent, the 30 -year average rate for the Implicit Price Deflator for State and Local Government. This baseline trajectory also adjusts http://www.ppic.org/main/home.asp Pathways for School Finance in California 26 the flex item, special education, and EIA rates for inflation, but the adjustment works differently. In these three programs, the funding rates all increase by the inflation rate. In real terms, these rates do not change. This baseline incorporate s an additional feature of California’s school finance system. Recent declines in the state budget have caused the state to appropriate revenue limit funds based on rates that were 18 percent lower than each district was entitled to by statute. We calculat e the revenue limit inflation increases using the statutory rates and add th e increases to th ose rates. However, the actual revenue limit rates assigned to districts are based on the available funds in a given year. Those available funds are driven by the economic and demographic factors highlighted in Table 2. Each year, the additional revenue is divided into two parts. The first is the amount necessary to adjust the prior year’s rates for the flex item, EIA, and special education for inflation , and to fun d the new year’s level of ADA at those rates. This amount can differ in real terms from the total revenue provided to districts in the previous year in those programs because ADA changes from year to year. After adjusting the three programs for inflation , the remaining new funds are used to bring revenue limit rates up to their new statutory levels. If the total revenue available for this purpose is only 90 percent of the amount required to bring all districts to their statutory levels, each district’s rate is then 90 percent of its statutory rate. Once state revenue in this model has grown sufficiently to fund all districts at their statutory revenue limits, we continue to make inflation adjustments in all programs . At this point, we also t rack the addition al funds available for reform efforts other than inflation adjustments. Figure 8 show s the results from simulating this baseline. For each year, the figures plot the 90 th, 75th, 50th, 25th, and 10 th percentiles of per -pupil revenue in the specified program. These percentiles are based on students not districts. To compute these values, all students are ranked based on the funding level associated with their district, and the percentiles are extracted from this list. For example, the 10 th percentile refers to the funding level received by the district of the student who ranks above the bottom 10 percent of the students in the list. The difference between the 90 th and 10th percentiles gives the range of funding levels experienced by the middle 80 percent of students. The figure also shows the funds available for education once the deficit factor has been restored to the revenue limits. These figures show revenue in real 2009 dollars. http://www.ppic.org/main/home.asp Pathways for School Finance in California 27 FIGURE 8 Status quo baseline In this status quo approach, revenue limit funds increase over time, but the amount of equalization across all districts is modest. The rates rise dramatically from 2009 through 2017, at which point they level off in real terms . The year 2017 marks the point at which the deficit factor has been eliminated and districts receive their inflation -adjusted statutory revenue limit rate. At the same time, funds begin to accumulate for other potential purposes. In 2017, those funds are small, averaging $63 per pupil. By 2030, however, those funds grow to $1,039 per pupil, providing scope for many reform efforts. The flex item, EIA, and special education programs maintain their original real rates throughout the scenario’s horizon. The dip in the 90 th percentile of the flex item occurs because enrollment s in districts with the highest flexible funding in the base year are projected to shrink relatively faster than other districts. Thus, in the future, the 90 th In Figure 8, the percentile lines for revenue limits essentially represent different district types. The 90 percentile student is in a district with l ower flexible funds. th percentile district is a high school district, the median line represents a unified district, and the 25 th 02004006008001,0001,2001,4001,6001,8002,000Flex Item90th75thMedian25th10th$ 2009 0 200 400 600 800 1,000 1,200Ec onom ic Im pa c t Aid $ 2009 $ 2009 0 200 400 600 800 1,000 1,200Special Education $ 2009 0 200 400 600 800 1,000 1,200Additiona l R e v e nue $ 2009 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000Revenue Limit $ 2009 percentile is an elementary district. Because the inflation increases are different for each district type, th ese percentiles do not converge. http://www.ppic.org/main/home.asp Pathways for School Finance in California 28 The lack of overall convergence among revenue limit rates is not surprising, given that the goal of the inflation increase mecha nism in revenue limits is to equalize rates within district types and not across all types. As Figure 1 showed, rates vary dramatically across district types but within the same type of district, there is less variation. In 2009, elementary districts avera ged a rate of $5,007, with only $62 separating the 90 th from the 10th percentile. High school districts averaged a rate of $6,016 with $87 separating the 90th from the 10 th percentile. Unified districts averaged $5,239 with a $79 gap between the 90th and 1 0th percentile. The inflation equalization mechanism does work within district types; these gaps do shrink over time. More dramatic are the narrowing gaps between the maximum revenue limit rates and the 90 th As discussed earlier, justificati ons for the current differences in revenue limit rates by district type are not entirely transparent, nor are the differences an explicit state policy. Thus, our subsequent scenarios provide a way of equalizing these rates across district types as well. percentile. By 2030, they fall in half from thei r 2009 levels of $2,944, $693, and $1,542 for elementary, high, and unified districts, respectively. This narrowing occurs primarily because the inflation increase reduces the real funding rate for high revenue districts. Re venue Limit Focus The first of these scenario s presents a more aggressive approach to equalizing revenue limits. The economic and demographic trends in Table 2 still determine the additional flow of revenue into the system, but our simulations divide that revenue in a different way. For each year, the revenue is divided into two amounts. The first is the amount necessary to fund the four programs in each district using the real per- pupil funding rates of the previous year. It is the amount necessary to hold districts harmless. Like our status quo example, this amount can differ from the total revenue provided to districts in the previous year because of changes in ADA. The difference between revenue for the four programs and the hold- harmless amount is new revenue available to increase funding rates. This first scenario allocates all of those additional funds to the revenue limit. We divide these additional funds among districts in a way that achieves revenue equalization across all districts over time. We f ocus on equalizing per -pupil revenue at the 90 th percentile of base year funding. Districts that begin a year with a funding rate below th at equalization target are entitled to a portion of the a dditional revenue. We allocate the additional revenue in the following way. We determine the difference between the total revenue each district would receive at the funding rates of the previous year and what it would receive at the equalization target r ate. We add up those deficits across all districts and compare that sum to the additional revenue available for revenue limits. If the additional revenue represents 30 percent of the total deficit, then each district below the target receives additional funds in that program equal to 30 percent of its deficit . That allocation defines the rate for the next year, and we repeat the process. This equalization mechanism directs the largest amounts of funding in absolute terms to districts furthest from the 90 th percentile. Districts only receive enough funds to brin g them to the equalization target. Districts with current rates above the target receive no additional funds until all districts have reached the target. Once all districts have reached th at original 90 th percentile target, we consider revenue limits funds to have equalized, and additional funds are dispersed on an equal per -pupil basis. At this point, the absolute gap between the 90 th percentile and higher percentiles will be constant. However, this process achieves some equalization even after all distric ts have achieved the original 90 th percentile rate, because districts above the 90 th percentile receive less in percentage terms than those at the 90 th percentile. Although the equalization targets are constant throughout the planning horizon, the rates and deficits change yearly, so additional funds are http://www.ppic.org/main/home.asp Pathways for School Finance in California 29 allocated differently each year. Figure 9 shows the pathway for revenue limits under this more aggressive equalization scheme. Since additional funds are allocated to the flex item, special education , and EIA in amounts that keep their real rates constant, the pathways in those programs look like their corresponding trajectories in Figure 8. FIGU RE 9 Revenue limit focus scenario Directing the additional funds to the revenue limit program brings all districts up to the 90th percentile of nearly $6,000 by 2013. From that year on, all additional revenue continues to flow into the revenue limit. By 2030, the 90 th Although this scenario directs all available additional resources to the revenue limit program and ultimately restores the current 18 percent deficit factor for all districts, the complete restoration occurs at different times for different districts. Elementary and unified districts eliminate their deficit factor first, because the revenue limit rates are below those of high school districts. Another option would be to restore the revenue limit rates to all districts before equalizing across district types. The pathway for such an approach would look like a combination of t he revenue limit trajectories in Figures 8 and 9. percentile rate climbs to $7,486, an increase of about 25 percent from the equalized value. This scenario demonstrates that , with focused attention on this goal, it is possible to achieve horizontal equity for revenue limits in a short amount of time. This revenue limit scenario demonstrates how the state could equalize the largest revenue program quickly, but it ignores the unequal distributions in other programs. Although Figures 2 and 3 demonstrate some dispersion in the funding rates for EIA and special education in 2009, the gaps are small relative to the level of funding in the revenue limit program . Furthermore, the gaps could be closed over time by redirecting a very small share of additional r evenue from the revenue limit rate to these programs that serve disadvantaged students. Only $96 separates the 90 th and 10th percentile students in special education , and the range is only $66 for EIA. Directing 1 percent of new revenue to EIA and 3 percen t to special education would bring all districts up to the current 90 th 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000 Revenue Limit 90th 75th Median 25th 10th$ 2009 percentile by 2030. This example assumes we use the same equalization mechanism for these programs as for the revenue limit. http://www.ppic.org/main/home.asp Pathways for School Finance in California 30 This modified scenario leaves 96 percent of additional fun ds available for the revenue limit program. This change causes a one -year delay for revenue limit equalization. The rates equalize in 2014, and that equalized rate grows to $7,396 by 2030. This level represents only a slight dip from the original revenue limit scenario. Unlike EIA and special education, the flex item rates exhibit substantial variation in 2009. Equalizing this program is the focus of our next scenario. Flex Item Focus As shown in Figure 8, the flex item exhibits a great deal of variation i n the base year of 2009, with $1,335 separating the 90 th and 10th percentiles. This gap is nearly twice the median flex item rate, and it represents about 21 percent of the average revenue limit and flex item rates combined. Nearly two-thirds of this gap o ccurs between the 90 th and 75th Compared with equalization mechanism in the prior scenario, this scenario slightly modifies the equalization target for the flex item. Because of the natural attrition of students in distric ts that currently receive high flex item rates, the 90 percentiles. Equalizing this program requires substantial additional funds. Our second scenario equalizes the flex item program by 2030, while continuing to achieve equalization in the special education and EIA programs in t hat year as well. To achieve this goal, 30 percent of new revenue each year is allocated to the flex item. As in the modified revenue limit scenario, 1 percent of additional funds are allocated to the EIA program and 3 percent to special education. The revenue limit receives the remaining 66 percent of additional funds. th percentile decreases naturally over time. Rather than use the equalization target of $1,915 ( the 90 th percentile in the base year ), this scenario uses a target of $1,539 ( the 90th percentile in year 2030 in the status quo scenario ). Every year in the simulation, districts with rates less than this target are entitled to a portion of the additional funds directed to the flex item. Like the previous scenario, the gap between each district ’s rate and the target is closed by an identical proportion, where the proportion is the share of new flex item funds to the total gap. Figure 10 shows the results from this simulation. FIGU RE 10 Flex item focus scenario 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000Revenue Limit $ 2009 02004006008001,0001,2001,4001,6001,8002,000Flex Item90th 75th Median 25th 10th$ 2009 0 200 400 600 800 1,000 1,200Ec onom ic Im pa c t Aid $ 2009 0 200 400 600 800 1,000 1,200Special Education $ 2009 http://www.ppic.org/main/home.asp Pathways for School Finance in California 31 Diverting such a large share of funds away fro m the revenue limit and to the flex item delays revenue limit equalization until 2019. Although the median revenue limit increases 30 percent from 2009 to 2030, the equalized revenue limit is about $675 lower than in the last scenario; it reaches only $6,721 instead of $7,396. Although this second scenario satisfies the principle of horizontal equity, it ignores the issue that the costs of educating disadvantaged students may be higher than for other students. Our next scenario addresses this issue. Econo mic Impact Aid Focus Our third scenario focuses on increasing funding to EIA , following recommendations by the Governor’s Committee on Education Excellence (2007) and Bersin, Kirst, and Liu (2008). Both recommendation s aim to close the achievement gap between disadvantaged students and other students. This scenario allocates 17 percent of new funds to the EIA program . The goal is a n EIA funding rate of $1,050 . To accommodate this emphasis on disadvantaged students, the share of new funds going to the revenue limit program drops from 66 percent, as in the prior scenario, to 50 percent. The share of new funds going to the flex item and special education remain at their levels from the previous scenario. Figure 11 shows the pathways for the revenue limit and EIA under this new allocation. Pathways for the special e ducation and the flexible categorical programs do not change from the previous figure. FIGURE 11 Economic Impact Aid focus scenario With this shift in priorities, all students are in districts receiving an EIA rate of at least $1,081 by 2030. This growth in EIA rates come at the expense of revenue limit rates. Compared with the previous scenario, revenue limit rates equalize in the same year, but the equalized rate reaches only $6,368, a $353 shortfall from the prior scenario. Comparing Scenarios The scenarios we present a re meant to demonstrate a range of possibilities for future funding. They all involve tradeoffs. We have presented them in a way that shows their cost relative to funding in the revenue limit program. If a relatively high fraction of additional funding is allocated to the flex item, funding rates in the program converge by 2030 and the median funding rate grows from $852 per ADA in 2010 to $1 ,541 per ADA by 2030 (Table 3). However, the projected 2030 revenue limit rate is about $675 per ADA lower, and equalization of revenue limit rates occurs five years later. The goal of increasing funding for disadvantaged students, as in our EIA focus scena rio, does not delay the revenue limit equalization by eve n a full year, but 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000Revenue Limit $ 2009 0 200 400 600 800 1,000 1,200Ec onom ic Im pa c t Aid90th 75th Median 25th 10th$ 2009 http://www.ppic.org/main/home.asp Pathways for School Finance in California 32 it does reduce the 2030 projection of that rate by about $353 per ADA. H owever, t he EIA funding rate is $1, 081 per ADA in 2030, as compared to $317 under the flex item or modified revenue limit focus. Other scenarios are certainly worth investigating , and hopefully our simulations have provided a framework for considering such extensions. TABLE 3 Programs Summary of scenarios Revenue limit f ocus Revenue limit + m odifications Flex item focus Economic Impact Aid focus Revenue limit Allocation share (%) 100 96 66 50 Year of equalization 2013 2014 2019 2019 Median rate 2010 5403 5395 5337 5306 Median rate 2020 6806 6743 6275 6030 Median rate 2030 7486 7396 6721 6368 Flex item Allocation share (%) 0 0 30 30 Year of equalization Never Never 2030 2030 Median rate 2010 790 790 852 852 Median rate 2020 780 780 1312 1313 Median rate 2030 776 776 1540 1541 Economic Impact Aid Allocation share (%) 0 1 1 17 Year of equalization Never 2030 2030 2010 Median rate 2010 313 317 317 377 Median rate 2020 314 346 346 846 Median rate 2030 314 361 361 1081 Special education Allocation share (%) 0 3 3 3 Year of equalization Never 2030 2030 2030 Median rate 2010 638 644 644 644 Median rate 2020 637 693 693 693 Median rate 2030 637 717 717 717 Regional Cost Differences None of the previous scenarios takes into account differences in regional wages , yet teacher compensation (salaries plus benefits) varies substantially across the state. In 2003– 04, teachers with the same level of education and experience averaged compensation just under $55,000 per year in the North Coast and Yolo Counties but over $70,000 annually in Santa Clara and Orange Counties (Rose et al. 2008). 10 10 The North Co ast Counties include Del Norte, Humboldt, Lake, and Mendocino C ounties. These differences in teacher compensation a re highly correlated with regional differences in the wages of college -educated workers who are not teachers. Non-teacher wages provide one way of measuring the purchasing power of revenue in a scho ol district. For example, if labor costs are 20 percent higher than average in District A and http://www.ppic.org/main/home.asp Pathways for School Finance in California 33 this district receives the same level of per -pupil funding for hiring teachers as the average district , then Di strict A could only purchase 83 percent (1.0/1.2) o f what the average district could purchase. To show how differences in purchasing power affect the distribution of reso urces, we adjust each district’s original rate in each of the four programs by an index of non -teacher wages. This regional wage index ( RWI) is described in Rose and Sengupta (2007). 11 This index groups counties into 30 regions, based on the U.S. Census Bureau’s definition of metropolitan statistical areas, and computes the average wage of college- educated non -teachers in each region. Each region is assigned an index value based on the ratio of its regional average wage to the state’s average wage, where the state average is weighted by the number of students in each region. The index ranges from a high of 1.2 in Santa Clara C ounty to a low of 0.8 in the North Coast Counties. Because salaries comprise about 80 percent of district budgets, we divide that portion of a district’s program rates by the RWI to obtain an RWI adjusted rate. Figure 12 shows the variation in funding before and after th ese adjustments for 2009. FIGURE 12 Program rates adjust ed by regional wage index, 2009– 2010 The most striking feature of the adjustment is the significant widening in revenue limit rates. Although actual revenue limits are quite similar, their purchasing power varies widely. Without the adjustment, the gap between the 75 th and 25th percentiles is imperceptible, measuring only $22. With the adjustment, $353 separates those percentiles. With the regional wage adjustment, the distribution of specia l education funds widens slightly. The gap between the 75 th and 25th 11 We update the index to include 2005 data from the Occupational Employment Survey. percentiles increases by $13, suggesting that some districts with the higher special education rates exhibited a slight tendency to be located in areas with lower regional wages. The dist ribution of flex item and EIA program rates, however, changed very little with the regional wage adjustment, suggesting there is no systematic relationship between those rates and regional wages. 5,000 5,250 5,500 5,750 6,000 ActualRWI-Adjusted Revenue Limit Dollars per ADA 500 750 1,000 1,250 1,500 1,750 2,000 ActualRWI-Adjusted Flex Item Dollars per ADA 600 650 700 750 ActualRWI-Adjusted Special Education Dollars per ADA 250 300 350 400 ActualRWI-Adjusted Ec onom ic Im pa c t Aid Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 34 To demonstrate the effect of compensating districts for reg ional differences in labor costs, we conduct a final simulation that repeats our original revenue limit scenario but equalizes the RWI -adjusted rates rather than the actual rates. The equalization mechanism works like that in the revenue limit scenario wit h the following exception. The equalization target is based on the 90 th percentile of the RWI -adjusted revenue limit rate in 2009, and districts are only entitled to additional revenue in a given year if their program ’s RWI -adjusted rate is below that targ et rate. Figure 13 shows the resulting pathway for the RWI -adjusted revenue limit rates. FIGURE 13 Simulated pathway for revenue limit rates adjusted by regional wage index In this scenario, the districts with RWI-adjusted rates below the target reach the target in 2013, the same year as the original revenue limit scenario. The pathways for the original and adjusted revenue limit rates look remarkably similar. However, the set of districts receiving revenue in these two scenarios differs. The correlation coefficient between gains in these two scenarios is 0.56. Table 4 shows the average gains for districts with and without the regional wage adjustment. Districts are categorized by type and whether they are in a high -, medium -, or low -wage regio n. About one -third of students are in low -wage districts and one - quarter are in high -wage districts . We chose these thresholds because they best accommodate the distribution of regional wages. 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000 Revenue Limit 90th 75th Median 25th 10th $ 2009 http://www.ppic.org/main/home.asp Pathways for School Finance in California 35 TABLE 4 District classification Average gains in revenue limit rates between 2009 an d 2013 ($) Original revenue limit scenario RWI-a djusted revenue limit scenario Elementary districts Low wage 878 393 Medium wage 886 824 High wage 880 1,215 High school districts Low wage 21 40 Medium wage 21 40 High wage 21 367 Unified districts Low wage 583 246 Medium wage 666 604 High wage 689 883 Without adjusting for regional wages, elementary districts gained the most while high school districts gained very little. With the adjustments, however, the gains for elementary districts in low -wage regions fall by more than 50 percent and those in high wage regions rise by about 40 percent . Furthermore, high school districts in high wage regions receive some additional funds when regional wages are taken int o account. This scenario demonstrates that, over time, the state could level the playing field by increasing the revenue of districts in high -wage regions. http://www.ppic.org/main/home.asp Pathways for School Finance in California 36 Conclusions By some accounts, California’s school finance system is so fundamentally flawed that only a complete overhaul could fix it. We do not share this view. The current system is flawed, but the basic elements of a sound system are already in place. These elements need to be strengthened, which can be accomplished steadily over time as economic and demographic conditions permit. Our numerical simulations illustrate the possibilities. Of course, t he results of our simulations depend on assumptions we have made about several economic and demographic factors and trends. For the most part, we have b een conservative in our assumption s, accepting a relatively slow growth rate in revenue per pupil. That is not to say that the future is certain to be rosier than our assumptions. One area of current concern is the looming fiscal pressure from a host of entitlement programs that promise to expand as the population ages. And in ten years, that concern will surely be replaced by another that is not apparent to anyone now. Despite these uncertainties, our simulations illustrate a simple and clear message. Even if the growth in revenue for public schools is relatively slow, steady improvement can achieve a great deal in time. Nonetheless, s teady improvement does require two difficult steps. The first is to formulate a clear vision of what the system should look like. The second is to create a mechanism to ensure that steady progress is made toward that goal. Our assessment of California’s system has identified a long menu of potential improvements. The simulations have focused on four: equalizing funding rates f or the core program, turning some current categorical programs into unrestricted support, increasing funding for districts with high percentages of economically disadvantaged students, and adjusting funding rates for regional differences in labor costs. We present these simulations , not as recommendations for a plan the state should adopt, but as illustrations of the type of analysis that can help the state develop a long -range plan. We welcome the opportunity to simulate other options if the state undertak es such an effort . Any long -range policy should also consider a number of other issues and questions:  Small schools in rural areas. The current approach to funding small schools in rural areas does not provide incentives for districts to find efficient way s of educating students in these areas. Would an approach based on an external measure such as population density be preferable?  Funding base for special education. The current system for funding special education has removed fiscal incentives to identify learning disabilities, but it may not adequately recognize cost differences among districts. Should the percentage of economically disadvantaged students in a district be part of the funding base for special education?  Funding base for Economic Impact Aid. English learners are now part of the funding base for Economic Impact Aid, which inadvertlently reduces funding for districts that are particularly effective in educating these students. Because most English learners are also economically disadvantaged, s hould the funding base for Economic Impact Aid be economically disadvantaged students instead of economically disadvantaged students and English learners?  Categorical flexibility. The legislature temporarily suspended restrictions on forty categorical programs, and we used this list of programs to demonstrate the process of turning categorical funding into permanent unrestricted support. If the suspensions were made perm anent, which programs should be excluded, which should remain, and what other progra ms should be included? http://www.ppic.org/main/home.asp Pathways for School Finance in California 37  Grade span. Revenue limit base rates tend to be higher for high school districts than for elementary districts. Should these differences be made explicit by enacting different base rates for students in different grade spans , as is currently the case in charter schools ?  Adjustments to revenue limit entitlements. Although most of a district’s revenue limit entitlement is determined by its base rate and attendance, a number of other adjustments are made to reach the district’s entitleme nt. Should those adjustments be phased out over time?  Excess taxes. If a district’s property tax revenue exceeds its revenue limit, it retains these excess taxes. Should excess taxes be refunded to taxpayers instead? Decisions on these issues would help form a clea r vision for the state to follow over time. Of course, c onsistency over time is a challenge, particularly for a term -limit legislature. However, California does have a precedent. Each year, every district’s base rate for revenue limit funding ch anges according to a formula set in statute. This formula updates base rates for inflation, but in a way that gradually equalizes base rates over time. This annual updating produces an appropriation of state aid, which does not require legislative action. Statutory appropriation is certainly an efficient mechanism for implementing a steady change over time. The statutory appropriation for revenue limits also has a provision for accommodating variations in state revenue. If economic conditions cause a downtu rn in state revenue, the legislature may decide that it cannot afford to fully fund the revenue limit appropriation in a particular year, creating a deficit factor for that year. In other years, it has used an abundance of revenue to raise funding rates fo r low-revenue districts. These ad hoc decisions could be made a routine part of an effort to create a rainy day fund for the state. In years in which the growth in state revenue exceeded the growth in the statutory appropriation, the difference could be se t aside in a separate fund. In years in which the revenue growth rate fell short of the growth rate in the statutory appropriation, the fund could be tapped to fund the statutory appropriation. In that way, changes would be phased in steadily over time, al lowing school districts to implement long -term strategies based on realistic assumptions of future revenue. None of the reforms discussed in this report will magically transform California’s public schools. A state’s school finance system is only a foundation. If well designed, it provides districts the revenue to employ the resources they need. Given this opportunity, the question then b ecomes whether districts will use their revenue effectively. But if they are not given th e opportunity, it is hard to see how they can otherwise be successful. http://www.ppic.org/main/home.asp Pathways for School Finance in California 38 References Angrist, Joshua, and Victor Lavy . 1999. “Using Maimonides’ Rule to Estimate the Effect of Class Size on Children’s Academic Achievement .” Quarterly Journal of Economics 114(2) : 533– 75. Angrist, Joshua D., and Jorn -Steffen Pischke. 2010. “The Credibility Revolution in Empirical Economics: How Better Research Design is Taking the Con O ut of Econometrics.” Journal of Economic Perspectives 24(2) : 3 –30. Bersin, Alan, Michael W. Kirst, and Goodwin Liu . 2008. Getting Beyond the Facts: Reforming California School Finance . Chief Justice Earl Warren Institute on Race, Ethnicity, and Diversity Issue B rief. University of California, Berkeley, California. Betts, Julian R., Andrew C. Zau, and Cory Koedel . 2010. Lessons in Reading Reform: Finding What Works Brewer, Dominic J., and Joanna Smith . 2006. “Evaluating the ‘Crazy Quilt’: Educational Governance in California .” Center on Educational Governance, Rossier School of Education , University of Southern California. . San Francisco: Public Policy Institute of California. California Budget Project . 2010. “Race to the Bottom? California’s Support for Schools Lags the Nation .” School Finance Facts. Chambers, Jay, Jesse Levin, and Danielle De Lancey . 2006. “Efficiency and Adequacy in California School Finance: A Professional Judgement Approach .” American Institutes for Research . G andara, Patricia, and Russell W . Rummberger. 2006, Resource Needs for California English Learners. University of California Linguistic Minority Research Institute. Governor’s Committee on Education Excellence. 2007. Students First: Renewing Hope for California’s Future. Hanushek, Eric A. 1997. “Assessing the Effects of School Resources on Student Performance: An Update .” Educational Evaluation and Policy Analysis 19(2) : 141– 64. Hoxby, Caroline M. 2000. “The Effects of Class Size on Student Achievement: New Evidence from Population Varia tion.” Quarterly Journal of Economics 115(4): 1239–85. Imazeki, Jennifer . 2006. “Assessing the Cost of K –12 Education in California Public Schools .” San Diego State University . Krueger, Alan B. 1999. “Experimental Estimates of Education Production Functions. ” Quarterly Journal of Economics 114(2) : 497 –532. Krueger, Alan B. 2002. “Understanding the Magnitude and Effect of Class Size on Student Achievement .” In The Class Size Debate, ed. Lawrence Mishel and Richard Rothstein (Washington, DC: Economic Policy Institute ). Legislative Analyst’s Office . 2007. “English Learners .” Analysis of the 2007 –08 Budget Bill: Education (February 21) . Legislative Analyst’s Office . 2010. “Update on School District Finan ce and Flexibility” (May 4). Little Hoover Commission . 2008. “Educational Governance and Accountability: Taking the Next Step .” Lipscomb, Stephen . 2009. “ Special Education Financing in California: A Decade After Reform ” Public Policy Institute of Calif ornia . Rivkin, Steven G., Eric A. Hanushek, and John F. Kain . 2005. “Teachers, Schools, and Academic Achievement .” Econometrica 73(2) : 417 –58. Rose, Heather, Jon Sonstelie, and Ray Reinhard . 2006. School Resources and Academic Standards: Lessons from the Schoolhouse . San Francisco: P ublic Policy Institute of California Rose, Heather, and Ria Sengupta . 2007. “ Teacher Compensation and Local Labor Market Conditions in California: Implications for School Funding .” Public Policy Institute of California . Rose , Heather, Ria Sengupta, Jon Sonstelie, and Ray Reinhard . 2008. “ Funding Formulas for California Schools: Simulations and Supporting Data .” Public Policy Institute of California . Sonstelie, Jon . 2007. “ Aligning School Finance with Academic Standards: A Weighted -Student Formula Based on a Survey of Practitioners .” Public Policy Institute of California . Weston, Margaret, Jon Sonstelie, and Heather Rose. 2009. “ California School Finance Revenue Manual .” Public Policy Institute of California . Weston, Marga ret. 20 10. “ Funding California Schools: The Revenue Limit System .” Public Policy Institute of California . Weston, Margaret . Forthcoming. “ California’s New School Funding Flexibility: One Year After Reform .” Public Policy Ins titute of California . http://www.ppic.org/main/home.asp Pathways for School Finance in California 39 About the Author s Heather Rose is an adjunct fellow at PPIC and an assistan t p rofessor in the School of Education at the University of California, Davis. She specializes in the economics of education. She has published work on school fin ance, college affirmative action policies, and the relationship between high school curriculum, test scores, and subsequent earnings. Her current research projects focus on school finance reform in California as well as school board politics and teacher sa laries. Previously, she was a research fellow at PPIC. She holds a B.A. in economics from the University of California, Berkeley, and an M.A. and Ph.D. in economics from the University of California, San Diego. Jon Sonstelie is a n adjunct fellow at PPIC a nd professor of economics at the University of California, Santa Barbara. His research interests include several areas in public finance and urban economics, including the effect of public school quality on private school enrollment, the incidence of the p roperty tax, the demand for public school spending, the economics of rationing by waiting, and the effect of transportation innovations on residential locations. He was previously a research fellow at Resources for the Future. He holds a B.A. from Washingt on State University and a Ph.D. from Northwestern University. Margaret Weston is a research associate at the Public Policy Institute of California’s Sacramento Center, where her work f ocuses on K–12 school finance. Prior to joining PPIC, she taught high school English and drama in Baltimore City Public Scho ols through Teach For America. She holds a master’s degree in teaching from Johns Hopkins University and a master of public policy degree from the University of Michigan. Acknowledgments We thank Carol Bingham and Heather Carlson from the California Department of Education for providing the data used in this report. We thank Gary Bjork, Ellen Hanak, Michael Kirst, Eric McGhee, John Mockler, Kim Rueben, Nicolas Schweizer, and Lynette Ubois for useful comm ents on previous drafts. PUBLIC POLICY INSTITUTE OF CALIFORNIA Board of Directors Walter B. Hewlett, Chair Director Center for Computer Assisted Research in the Humanities Mark Baldassare President and CEO Public Policy Institute of California Ruben Barrales President and CEO San Diego Chamber of Commerce Maria Blanco Executive Director Chief Justice Earl Warren Institute on Race, Ethnicity and Diversity University of California, Berkeley School of Law John E. Bryson Reti red Chairman and CEO Edison International Gary K. Hart Former State Senator and Secretary of Education State of California Robert M. Hertzberg Partner Mayer Brown LLP Donna Lucas Chief Executive Officer Lucas Public Affairs David Mas Masumoto Author and farmer Steven A. Merksamer Senior Partner Nielsen, Merksamer, Parrinello, Mueller & Naylor, LLP Constance L. Rice Co -Director The Advancement Project Thomas C. Sutton Retired Chairman and CEO Pacific Life Insurance Company The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research on major economic, social, and political issues. The institute’s goal is to raise public awareness and to give elected representatives and other decisionmakers a more informed basis for developing policies and programs. The institute’s research focuses on the underlying forces shaping California’s future, cutting across a wide range of public policy concerns, including economic development, education, environment and resources, governance, population, public finance, and social and health policy. PPIC is a private operating foundation. It does not take or support positions on any ball ot measures or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. PPIC was established in 1994 with an endowment from William R. Hewlett. Mark Baldassare is President and Chief Executive Officer of PPIC. Walter B. Hewlett is Chair of the Board of Directors. Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the abo ve copyright notice is included. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Copyright © 2010 Public Policy Institute of California All rights reserved. San Francisco, CA PUBLIC POLICY INSTITUTE OF CALIFORNIA 500 Washington Street, Suite 600 San Francisco, California 94111 phone: 415.291.4400 fax: 415.291.4401 www.ppic.org PPIC SACRAMENTO CENT ER Senator Office Building 1121 L Street, Suite 801 Sacramento, California 95814 phone: 916.440.1120 fax: 916.440.1121" } ["___content":protected]=> string(104) "

R 1110MWR

" ["_permalink":protected]=> string(85) "https://www.ppic.org/publication/pathways-for-school-finance-in-california/r_1110mwr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8747) ["ID"]=> int(8747) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:40:27" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(4064) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(9) "R 1110MWR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(9) "r_1110mwr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(13) "R_1110MWR.pdf" ["wpmf_size"]=> string(6) "561187" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(108380) "Pathways for School Finance in California November 2010 Heather Rose, Jon Sonstelie, and Margaret Weston with contributions from Hans Johnson Supported with funding from The William and Flora Hewlett Foundation http://www.ppic.org/main/home.asp Pathways for School Finance in California 2 Summary This report demonstrates how California can improve its school finance system steadily over time as economic and demographic conditions permit. The improvements we suggest here are derived from our analysis of California’s current system using the following five principles :  Meet resource needs : Schools should have the resources necessary for their studen ts to meet state academic standards, and the cost of those resources may vary from school to school for a variety of reasons.  Structure incentives properly: The formulas allocating revenue to schools should not give schools incentives to deviate from actio ns in the best interest of students and taxpayers .  Allocate f unds transparently : The formulas for allocating revenue to schools should be clear and relatively simple.  Treat s imilar districts equitably : When the state has chosen the factors that determine t he revenue a school district receives, school districts with the same values for those factors sho uld receive the same revenue.  Balance s tate and local authority : Restrictions on the use of funds must properly balance the state objectives with the realities that schools differ widely across the state and that school administrators have unique knowledge about local conditions. California’s school finance system violates th ese principles in many ways. Under the current system, different district s are funded at different rates, a clear violation of horizontal equity. Unlike school finance systems in other large states, California does not adjust revenue to school districts based on regional differences in the cost of hiring employees, failing to recogn ize a large and obvious cost difference among districts. Because of the many state categorical programs directing revenue to public schools, the allocation of revenue to districts is not transparent, and the many restrictions on the use of funds in those p rograms unduly constrain local school administrators. Moreover, although California does provide additional funds for school districts with many economically disadvantaged students, the additional funds are not large enough to compensate for the differences in student need correlated with poverty. Our analysis also reveals several other areas in which California’s system could be improved. Making improvements without making some districts worse off would require additional revenue, which is now in short sup ply. However, as the economy improves, state tax revenue will rise, and the state can afford to invest again in its schools. At the same time, school enrollments are projected to rise relatively slowly, allowing an increase in revenue per pupil. This increase will not be dramatic, but it promises to be relatively steady, permitting the state to make slow and steady progress over time. To illustrat e the possibilities, this report simulates this process for a variety of potential improvements. One scenario eq ualizes funding rates for the main programs in the current system. In another scenario, funding is increased in districts with many economically disadvantaged students. A final scenario demonstrates the consequences of adjusting funding rates for regional differences in the cost of hiring personnel. Contents Summary 2 Figures 4 Tables 5 Introduction 6 School Finance Principles 7 Meet Resource Needs 7 Structure Incentives Properly 9 Allocate Funds Transparently 10 Treat Similar Districts Equitably 10 Balance State and Local Authority 10 Assessing California’s System 12 Revenue Limit Funding 13 Special Education 15 Economic Impact Aid 16 Other Categorical Programs 18 Simulating Pathways 21 Parameters 21 Scenarios 25 Conclusions 36 References 38 About the Authors 39 Acknowledgments 39 All technical appendices to this paper are available on the PPIC website: http://www.ppic.org/content/pubs/other/1110MWR_appendix.pdf http://www.ppic.org/main/home.asp Pathways for School Finance in California 4 Figures Figure 1 . Revenue limit base rates, 2009 –2010 .................................................................... 14 Figure 2 . Special education base rates, 2009 –2010 ............................................................ 16 Figure 3 . Economic Impact Aid f unding rates, 2009 –2010 ................................................... 18 Figure 4. Flex item funding rates, 2009 –2010 ..................................................................... 20 Figure 5 . Public school expenditures as a percent of state personal income, 1970 –2007 ......... 22 Figure 6 . Real personal income per t axpayer, 1970 –2008 .................................................. 23 Figure 7 . Taxpayers per s tudent, 1970–2007 ...................................................................... 24 Figure 8 . Status quo baseline ............................................................................................. 27 Figure 9 . Revenue limit focus scenario ............................................................................... 29 Figure 10 . Flex item focus scenario .................................................................................... 30 Figure 11. Economic Impact Aid focus scenario ................................................................... 31 Figure 12 . Program rates adjusted by regional wage index, 2009 –2010 ............................. 33 Figure 13 . Simulated pathway for revenue limit rates adjusted by regional wage index ............. 34 http://www.ppic.org/main/home.asp Pathways for School Finance in California 5 Tables Table 1 . School districts by type and size, 2009 –2010 ......................................................... 13 Table 2 . Projections of economic and demographic trends ............................................... 25 Table 3 . Summary of scenarios .......................................................................................... 32 Table 4. Average gains in revenue limit rates between 2009 and 2013 ($) ......................... 35 http://www.ppic.org/main/home.asp Pathways for School Finance in California 6 Introduction California’s bu dget crisis has diminish ed educational resources for the state’s current cohort of public school students . Because school districts have less revenue, class sizes are larg er and struggling students receiv e less assistance. Under these circumstances , it seems beside the point to suggest that California should begin plan ning for t he next cohort of students . Yet, history demonstrates that a failure to plan now will leave the state unprepared for what will surely follow . Although the current recession is deep, economic recovery will come , offering increas ing tax revenue and an opportunity for the state legislature to be more generous . How wi ll the state take advantage of this opportunity ? It may respond as it has in the past by allocating new revenue to schools for specific purposes . By 2005 –06, the state had more than 6 0 pro grams targeting a variety of purposes such as reducing class sizes, hiring counselors, purchasing textboo ks, and involving parents (Weston, Sonstelie, and Rose 2009 ). Altern atively, the state might use additional revenue to address underlying weaknesses in its school finance system . Our paper explores this alternative. We begin by discussing five broad principles for assessing California’s school finance system . Th ese principles do not lead to a single superior system, but they do suggest sev eral ways in which California could improve its current system . T hrough numerical s imulations, we illustrate the effects of pursuing some of th ese improvements steadily over time. The simulatio ns demonstrate that corrective long-term polic ies could significant ly strengthen California’s school finance system. http://www.ppic.org/main/home.asp Pathways for School Finance in California 7 School Finance Principles California’s school finance system is fundamentally different from the systems of most other states. In most state s, school districts have the power to set tax rates on real property. They have a robust source of discretionary local revenue. In California, school districts have limited taxing authority . They receive property tax revenue, but the state determines the amount they receive. 1 Because the instit utions governing schools are not the institutions financing them, conflict between the two is inevitable. Aligning these institutions should be a high priority . Our goal is this paper is more modest , however. We take as a given California’s current mixture of state finance and local governance and ask how California’s school finance system might be improved , given that mixture. We believe that five principles can be useful in guid ing this improvement . Yet, school districts are not agencies of the state. Each district has an elected school board that determines how its revenue is spent. Meet Resource N eeds We expect many things from our schools. Chief among th ese expectations is that students graduate from high school with a sound education. Over the p ast several years, California has spent considerable effort defining that education. The result is the state’s Academic Content Standards. The state has also implemented a battery of tests to determine whether students meet those standards. Although the tests are imperfect measures of knowledge and the s tandards do not include everything we expect students to l earn, a fundamental goal for any school finance system is to ensure that schools have the resources (teachers, textbooks, aides, counselors, and so on) necessary for their students to meet the state’s standards. B ecause the relationship between resources and academic achievement has not been firmly established, it is difficult to determine th ese resources with certainty . 2 Although these two highly regarded experts disagree on the conclus ions to be drawn from existing research, both agree that better research on the rel ationship between resources and achievement is needed. In particular, researchers need longitudinal data that tracks the academic improvement of individual students over tim e. Using such data , Rivkin, Hanushek, and Kain (2005 ) found that students in small classes did improve more rapidly than students in large classes. In that study, however, class sizes for individual students were determ ined through a process that the researchers did not explicitly account for in their analysis , raising concerns that class size might be related to unmeasured characteristics of students that also affect their achievement . The best response to th ese concerns would be an experiment in which students are randomly assigned to classes of different sizes. Using data from the onl y large experiment with class sizes, For example, several studies have focused on the rela tionship between class size and student achievement. After reviewing 59 of these studies, Hanushek (1997) judged the research inc onclusive: Most studies failed to find a statistically significant effect of class size on achievement, and the positive findings were offset by an equivalent number o f negative findings. Reviewing th e same set of studies, however, Krueger (2002) concluded that the research supports the belief that s maller class size lead s to higher student achie vement. 1 School districts may enact a parcel tax if it is approved by tw o-thirds of the voters. In 2005 –06, 98 districts had a parcel tax, representing 0.4 percent of school dist rict revenue. 2 Recent studies exploring the resource needs of California schools include Chambers, Levin, and DeLancey(2006); Imazeki (2006) ; and Sonstelie (2007). http://www.ppic.org/main/home.asp Pathways for School Finance in California 8 Krueger (1999) found that students in smaller classes did progress faster. Taking advantage of a natural experiment in whi ch class sizes were determined through a well-understood process th at was unlikely to be affected by unmeasured student characteristics , Angrist and Lavy (1999) reached the same conclusion. Examining data from a similar natural experiment, however, Hoxby ( 2000) found no significant effect of class size on achievement. In reviewing recent research, Angrist and Pischke conclude that reductions in class size do increase student achievement and that the estimated effects are consistent across studies. Reasonabl y well-identified studies from a number of advanced countries, at different grade levels and subjects, and for class sizes ranging anywhere from a few students to about 40, have produc ed estimates within a remarkably narrow band. 3 While w e agree that the best recent research tends to find a statistically significant relationship betwee n class size and student achievement , we do not believe this research is sufficient to give precise guidelines about the class sizes sufficient to achieve the state’ s academic standards . R educing class size is also very costly. A more efficient use of resources might be to focus on struggling students through interventions such as after -school tutoring or summer school ( Betts, Zau, and Koedel 2010 Recent research has also confirmed the importance of effective teaching (Rivkin, Hanushek, and Kain 2005) . This research suggests t hat effective teachers may have a more important influence on student achievement than reduction s in class size . Accordingly, i dentifying, recruiting, developing , and retaining s uch teachers should be a high priority for schools. From th is perspective , the most efficient use of a school’s revenue may be in providing the compensation and support that will attract and retain excellent teachers. Considering all the possible uses of school revenue, w e conclude that although the best research is consistent wi th a positive relationship between resources an d achievement, the parameters of th is relationship are not yet well understood. ). Howeve r, we are not aware of research on these inter ventions with the scale and statistical sophistication of the best recent research on class size. On the other hand, it is well understood that achievement varies dramatically among students provided with the same level of educ ational resources. Learning disabilities hinder the progress of some students. O thers may lack English language skills. P reparation, motivation, and aptitude may also be issues. To achieve gra de-level proficiency, some students may need additional attentio n from their teachers or after- school tutor ing. Because these needs are not uniformly distributed across sc hools, some schools will require more resources than others to meet the state’s standards. The cost of resources also varies across school districts in the state. More than half of a district’s budget consists of the salaries and benefits of its teachers. For the services of these and other employees, school districts must compete with other employers in local labor market s. As Rose and Sengupta (2007) show , the compensation offered by these employers differs significantly across regions of California, and thus the compensation of public school teachers also varies by region. In regions where other employers are offering relatively high salaries and benefits , school districts must do so also . To offer similar levels of educational services to their students, districts in high compensation regions must have higher revenue than similar districts in other regions. 3 Angrist and Pischke 2010, p. 24. http://www.ppic.org/main/home.asp Pathways for School Finance in California 9 Other costs may also vary across di stricts. As shown in Rose et al. ( 2008), the cost of transporting pupils to school is higher in rural areas . In the 100 districts with the lowest population density, transportation costs in 2003 –04 averaged more than $700 per pupil. In contrast, in the 300 districts with the highest population density, th e cost was about $100 per pupil. Utility costs also vary among districts, although not as widely as transportation cost s. Structure Incentives Properly In addressing cost differences, a school finance sys tem must not inadvertently reward districts for actions not in the best interests of students and taxpayers. For example, to account for differences in transportation costs, the state might reimburse school districts for the cost of transporting students t o school. Cost reimbursements would certainly neutralize cost differences across districts, but reimbursements would also remove any incentive school districts might have to control the costs of pupil transportation. This dilemma could be resolved, however, by using a measure of cost outside the control of school districts—for example, population density (see Rose et al. 2008), which is negatively correlated with transportation costs but independent of any action a district might take. The state could address the special needs of rural districts by allocating additional funds to districts with low population density.Districts would have flexibility in the use of these funds and thus have an incentive to use them wisely. This same concept applies to student achievement. Obviously, students who fall behind need additional instruction. However, if funds were allocated to school districts based on the share of students who fail to achieve proficiency on statewide tests, the districts with a lower share of profic ient students would receive more revenue than other districts, reducing funding for districts that were particularly successful in raising student achievement. As in the case of transportation, the resolution is to find a measure that is unaffected by dist rict actions but that is related to the likelihood that a student will fall behind. As many studies have shown, one such measure is the income of a student’s parents. Each year, the Census Bureau estimates the percentage of a district’s students living bel ow the federal poverty level. This measure is negatively correlated with student achievement, but it cannot be affected by any action taken by the district. Furthermore, this negative correlation exists within schools as well as across them, implying that the observed variation across schools cannot be solely due to a negative correlation between poverty and school effectiveness. Unfortunately, no external measure can precisely capture all of the differences i n cost across districts. For example, in the case of transportation, we might have two districts with the same population density; but in one, almost all students are concentrated in one town, while in the other, students are spread evenly throughout the d istrict. Transportation costs in the second district will be much higher than in the first, even though the population density is the same. Likewise, the percentage of students living in poverty is an imperfect indicator of average family income. Districts in which all of the families are just above the poverty line would be quite different from districts in which all of the families are well above the poverty line. These examples demonstrate that the principle of accounting for cost differences can conflic t with the principle of structuring incentives properly. A school finance system must find a balance between the two. And if a source of cost variation cannot be closely related to an external measure, it is hard to see how the school finance system can ta ke account of that cost without rewarding inefficiency. http://www.ppic.org/main/home.asp Pathways for School Finance in California 10 Allocate Funds Transparently Transparency is important in al l areas of government. The lack of transparency breeds distrust and undermines support for public institutions. Transparency is particularly important in the allocation of funds to school districts. Compared to other public services, the resources employed in public schools are clearly evident. Parents generally know the class sizes of their schools and the opportunities available to their children. At this level of the bureaucracy , public schools are relatively transparent , and parents rightly believe that they should be able to understand why resources differ across schools and districts , an understanding that ultimately requires them to k now why revenues var y across districts. Parents are more likely to understand why revenue s var y if the rules for allocating school funding are simple. Of course, t he simplest and thus most transparent rule is to a llocate revenue to districts in proportion to their enrollment , ignoring cost differences among districts. On the other hand , a set of rules for allocati ng revenue that account ed for every cost difference would be extremely complicated and thus not very transparent. These two extremes illustrate the tension between transparency and the recognition that costs are likely to differ across districts. This tension requires a pragmatic approach. If cost differences are small , they should be ignored. For example, a fter investigating the relationship between climate and utility costs, Rose et al. (2008) argue that variations in cost due to climate are not large enough to make climate a significant factor in all ocating revenue to schools. In contrast, regional salary difference s are large (Rose and Sengupta 2007) and should, in principle, be incorporated in a finance formula , although this would involve a number of complicated practical issues. Regardless of how regional boundaries are drawn, some adjacent districts would end up in different regions and thus receive different cost adjustments. With many regions, these differences would be small ; but nonetheless, a system involving many regions would be quite complicated, violating the principle of transparency . Other large stat es (Florida, New York, Texas) have been able to overcome this obstacle , however. Treat Similar Districts Equitably Consideration of costs, incentives, and transparency suggests a number of factors that might be used to allocat e revenue across school districts. For example, the factors might be average daily attendance, percentage of students living in poverty , and a regional wage index . Once the factors have been decided upon , every district with the same values for those factors should receive the same revenue, a concept sometimes referred to a s horizontal equity. Horizontal equity is closely related to transparency. If the law is clear about the factors to be considered in allocating revenue, then districts with the same values for th ose factors sho uld receive the same revenue. Balance State and Local Authority The four principles discussed above concern the allocation of revenue to school districts. Th is last principle concerns the conditions place d on the use of those funds. The bulk of revenue provided to schools is determined by the state legislature, which must weigh the needs of school districts against those of other state agencies and local governments. In this situation, it is only natural that the legislature may consi der some uses of school district revenue to have a higher priority than other s. In arguing for funds, school districts themselves will tend to empha size some uses, such as reducing class sizes, over other uses, such as http://www.ppic.org/main/home.asp Pathways for School Finance in California 11 hiring district administrators. Thus, i t should come as no surprise when the legislature places restrictions on how school districts are to use their funds. Many district administrators also favor some external restriction on the use of funds . In interviews with randomly selected s uperintendents, two advantages of such restrictions were commonly cited (Rose, Sonstelie, and Reinhard 2006). First, restrictions may protect funds from the collective bargaining process. For example, f unds to reduce class size require districts t o hire more teacher s rather than increas e the salaries of current teachers. Second, restrictions may thwart local political pressures. F or example, fu nds for disadva ntaged students require districts to provide additional resources to schools with large enrollments of these students. Some superintendents believed that w ithout such restrictions, locally elected school boards might tend to allocate resources equally to all schools, regardless of differences in need. On the other hand, California is home to a large and complicated K –12 system —more than 8,000 public schools with widely di fferent students and staff. Loc al administrators have considerable information about the strengths and weaknesses of their personnel and the abilities and backgrounds of their students, more than any central authority could have. The decentralization of knowledge argues for a decentralization of decisions about how revenue should be employed and for few restrictions on the use of funds (Brewer and Smith 2006) . Some have suggested that t he tension between state finance and local go vernance can be eased because of the recent emphasis on standards and accountability (Little Hoover Commission 2008) . The state has defined what it expects schools to achieve. It can therefore give sch ools more authority in how they achieve those objectives. It has defined outputs, so it can loosen its grip on inputs. Although this suggestion seems right in theory , it depends on the clear and fair measurement of outputs and on the power to hold district administrators accountable for meeting those measured objectives. Although California has made progress in measur ing student achievement , the state has only limited authority when it comes to holding local administrators accounta ble, a limit that is a natural consequence of local governance. It seems to us that the tension between state finance and local governance is unlikely to be resolved and that the restrictions that spring from that tension should be judged on a case -by -cas e basis. Some restrictions are clearly motivated by a difference in the objectives of the state legislature and those of locally elected school boards. Others, however, can only be rationalized by a difference of opinion about how a common objective is bes t pursued. In those case s, we believe the state should defer to local authorities. http://www.ppic.org/main/home.asp Pathways for School Finance in California 12 Assessing California’s System The five principles discussed above provide us with a lens for examining California’s current school finance system. Broadly speaking, the system has four funding components. The first is revenue limit funding, which combines local property tax revenue with state aid to generate a source of funds that school districts can use on any educational purpose. Revenue limit funding constitutes approximately 60 percent of the fundi ng received by school districts and forms the foundation of the state’s school finance system. The second component is a collection of programs that channel sta te aid to districts and place restrictions on how that aid is used. These programs , generally referred to as state categorical programs, constitute more than 20 percent of the revenue received by districts. The third component is a collection of federal categorical programs, const ituting approximately 10 percent of funding. The last component is discretionary local revenue such as parcel taxes and interest income. These local funds also constitute approximately 10 percent of district funding. In this discussion, we focus on the rev enues controlled by the state legislature—i.e., revenue limit funding and state categorical progr ams. And i n our analysis of state categorical programs, we are concerned about two programs in particular. The first is special education, which funds services for students with learning disabilit ies. The second is Economic Im pact Aid, which targets English learners and economically disadvantaged students. These two programs are the primary vehicles for addressing differences in studen t need. We analyze t he remaining categorical programs together as a single group . Our analysis does not include a consideration of the state’s approach to providing funds for students in sparsely populated areas. Although the provisions for necessary small schools and the pr ogram for pupil transportation recognize the higher cost of educating students in sparsely populated areas, the programs do not provide incentives for school districts to find efficient methods to educating students in these areas. When a school achieves necessary small school status, it has little incentive to merge with other schools, even if that merger would reduce costs without diminishing the education of students. Similarly, transportation funds are allocated according to historic costs, removing inc entives for districts to find cheaper solutions for educating students in sparsely populated areas. A program that allocated funds according to population density might address the needs of these areas without inadvertantly rewarding inefficiencies, a subj ect which deserves further study. In the follow ing discussion , we analyze revenue sources throught the lens of the principles of school finance described above. Different principles come into play in different areas, but one issue cuts across all areas. U nlike other large states such as Florida, New York , and Texas, California does not adjust revenue in any of its programs for regional salary differences . The state ignores the very large variations in the costs of the most important resource school distric ts employ, a clear violation of our first principle. Ou r analysis presents funding rates (dollars per pupil) for school districts. For our purpose s, districts are separated into nine groups based on their type (elementary, high school, and unified) and siz e (small, medium , and large). We chose the size partitions to yield a roughly equal number of districts in the three size classifications for each district t ype. Our analysis excludes necessary small schools and charter schools. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools. The schools and districts we do include represent 95 percent http://www.ppic.org/main/home.asp Pathways for School Finance in California 13 of California’s public school enrollment. As Table 1 shows, large un ified districts include nearly 60 percent of these students. 4 TABLE 1 Type and size of district School d istricts by t ype and size, 2009– 2010 Number of d istricts Average daily attendance (ADA) Percent of total ADA Elementary Small (0 –250) 132 17,474 0.3 Medium (251–1,500) 171 110,693 2.0 Large (1,501+) 175 969,368 17.5 High s chool Small (0 –1,500) 23 19,607 0.4 Medium (1,501–6,000) 26 82,848 1.5 Large (6,001+) 31 444,893 8.0 Unified Small (0 –3,000) 120 151,138 2.7 Medium (3,001–10,000) 96 556,209 10.2 Large (10,001+) 110 3,181,060 57.4 All districts 884 5,543,291 100.0 SOURCE: 2009 Principal Apportionment, California Department of Education. NOTE: Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools. Revenue Limit Funding A simple formula determines a school district’s revenue limit funding . Every district has a base rate, a dollar amount per pupil. That base rate is multiplied by the district’s average daily attendance (ADA) to determine its total funding entitlement. This entitlement is met through local property taxes and state aid. Students in necessary small schools are funded through a different formula. In addition, the calculation of a district’s revenue limit entitlement involves several other adjustments that generally stem from policy decisions made over the years. As Weston (2010a) shows, these adjustments do not contribute much to the variation in revenue limit funding per pupil across school districts in California. The biggest source of variation stems from differences in the base rate among districts. These variations are represented in Figure 1. The boxes show the distance between the base rate in the 75 th and 25 th percentile for a group. Percentiles are weighted by the number of students in a district. Within each group, students are assigned the base rate of their district and ranked according to this rate. The 75 th percentile is the base rate of the student in the 75 th percentile of this ranking. The upper light part of each box is the distance between the median base rate and the base rate in the 75 th percentile. The v ertical lines show the distances between the 10 th and 90th 4 The percentile. Each group also has three horizontal hash marks above the box and three below it. These marks show the highest three and the lowest three base rates in each group. When the two or more of the individual base rates are nearly identical, the hash marks for those rates are indistinguishable and appear as just one mark. For example, large unified districts appear to have two hash marks above the box because the second and third highest base rates are nearly the same. For almost every group, the bottom three hash marks are very close or indistinguishable. technical appendix provides more details on students and districts excluded in our analysis. http://www.ppic.org/main/home.asp Pathways for School Finance in California 14 FIGURE 1 Revenue l imit base r ates, 2009– 2010 SOURCE: 2009 Principal Apportionment, California Department of Education. NOTE: Base rat es are statutory rates for 2009 –10, with 18.355 deficit factor . Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools. Base rates vary significantly across groups. Some of the vari ation is due to differences between high school districts and other districts. The median rate for each of the three high school groups is approximately $6,000 per pupil. For the elementary groups, all three median s are close to $5,000 per pupil. The medians for the medium and large unified districts are about $5,200. For the small unified districts, the median is $5,517. As the figure demonstrates, t here are also variations withi n groups, primarily among the small elementary and small unified groups . This variation clear ly violat es the principle of horizontal equity. These variations reflect a historical process. When the revenue limit system was first introduced in 1973, a distr ict’s base rate was its expenditures per pupil in 1972–73. Over time, the state has gradually raised the lowest base rates. To determine the relative position of base rates, districts were classified by type and size. Equalization reduced differences withi n groups but did not necessarily reduce differences across groups. The higher average rates of high school districts are sometimes justified by the notion that high schools are more expensive to operate than other schools. Although this may be the case, th e differences in base rates among district types are not an explicit state policy , and the research that might justify these differences is not conclusive (Sonstelie 2007) . Furthermore, if state policy did mandate higher funding rates for high school students, the base rates of unified districts should reflect the percentage of their students attending high school. 4,500 4,750 5,000 5,250 5,500 5,750 6,000 6,250 6,500 6,750 7,000 7,250 7,500 7,750 8,000 Small ElementaryMedium ElementaryLarge ElementarySmall High SchoolMedium High SchoolLarge High SchoolSmall UnifiedMedium UnifiedLarge Unified Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 15 On occasion, the state has increased variations in base rates as it introduced new policies or phased out old ones . In 1997, it changed the methods for calculating the ADA of a district. To offset the negative effects this change had on some districts, the state increased their base rates . 5 Special Education The rate differences it created by this increase are temporary , however, because of the formula for an nually adjusting base rates. Each year, the state calculates an amount per pupil necessary to a djust the average base rate for inflation. Different amounts are calculated for each type of district: high school, elementary , and unified. The state then increa ses the base rate of every district by th e amount of this inflation increase . For districts with base rates above the average, this adjustment is not high enough to keep up with inflation. The base rate of these districts falls in real terms. For districts with base rates below the average, real rates rise. When inflation is low , as it has been for some time, these equalizing changes are small. With this policy, it will take many years to equalize base rates. The state allocates funding f or special education through special education local planning areas (SELPA). SELPAs are groups of districts, county offices of education, and charter schools that agree to share special education funding and services. Over 90 percent of this funding is allocated through a simple formula. Each SELPA has a base rate, expressed in dollars per ADA. Its entitlement is this rate multiplied by its ADA. This entitlement is met through property taxes and federal and state aid. This formula was created by Assembl y Bill 602 in 1997. Before that time, special education funding was allocated according to the costs and needs of special education students within each SELPA, creating a fiscal incentive for districts to classify students as disabled. The new bill ended t his incentive by allocating special education funds according to the attendance of all students, not just special education students. Although the current formula is consistent with the principle of structuring incentives properly, it does raise the question of whether the variations in special education costs among districts are adequately addressed, the first of our principles. These variations are partly addressed by three separate funding sources for relatively rare, but severe, disabilities. One provi des additional revenue for districts that must purchase special materials and equipment, and the other two fund the placement of students in special facilities. The funding formula also includes a Special Disability Adjustment that provides additional fund s for SELPAs that had unusually high incidences of learning disabilities when the new formula was created in 1997. These adjustments and additional funding sources certainly address some of the cost differences across districts. However, a fundamental question remains: Is the incidence of learning disabilities randomly distributed across districts? Recent research indicates that this is not the case. Using data from a large survey of families, Lipscomb (2009) found that the incidence of severe disabilities among children is negatively correlated with family income. This finding suggests that the formula for allocating special education revenue among SELPAs should include poverty rates as well as ADA. Because poverty rates are outside a district’s control, t his change would address cost differences among districts without creating incentives for districts to identify students as learning disabled. Those possible changes notwithstanding, the current allocation of special education revenue clearly violates the p rinciple of horizontal equity. As Figure 2 shows, base rates for special education vary across districts. 5 For details, see Weston 2010, p. 11. http://www.ppic.org/main/home.asp Pathways for School Finance in California 16 For example, for large unified districts , the rate in the 75th percentile is 10 percent higher than the rate in the 25 th A second issue is whether the level of special education funding is adequate overall to meet the cost of that education. In 2006– 07, California school districts spent more than twice as much on special education services as they received in special education revenue (Lip scomb 2009). In the jargon of school finance, special education services “encroached” on general education services. If special education revenue were increased substantially, however, some districts would surely have more revenue for special education ser vices than they otherwise might spend on those services, creating incentives for districts to identify learning disabilities and to spend too generously on special education services. To us, encroachment is a much less serious issue than is the allocation o f existing revenue according to need. percentile. FIGURE 2 Special e ducation base rates , 2009–2010 SOURCE: 2009 Principal Apportionment, California Department of Education. NOTE: The rate for each district is the rate of the SELPA to which it belongs. Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools.. Economic Impact Aid Economic Impact Aid (EIA) funds supplemental services for English learners and economically disadvantaged students. Each district’s entitlement is determined by multiplying its EIA rate by a weighted count of eligible students. The count starts with the number of English learne rs in the district plus the 500 550 600 650 700 750 800 850 900 950 1,000 Small ElementaryMedium ElementaryLarge ElementarySmall High SchoolMedium High SchoolLarge High SchoolSmall UnifiedMedium UnifiedLarge Unified Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 17 district’s count for the federal Title I program, a Census-based estimate of the number of students living in poverty. Every student in this count in excess of 50 percent of all students receives a weight of 1.5. Every other student receives a weight of unity. The EIA count is the sum of these weights. The EIA program is the primary mechanism through which California addresses differences in student need correlated with family income and native language. The EIA formula places a heavy weight on English learners because 85 percent of English learners are also economically disadvantaged (Legislative Analyst’s Office 2007). Accordingly, most English learners generate twice as much EIA revenue for a district as an economically disadv antaged student who is fluent in English. 6 Including English learners in the base for Economic Impact Aid also creates incenti ves for districts that are not consistent with state goals. In a district with a particularly effective program for English learners, students move relatively quickly to fluency, and the district receives less revenue than it would if students were slower to make this transition. After a review of several studies of the resource needs of English learners, Gandara and Rumberger (2006) question the implicit assumption underlying this revenue premium. They argue that although English learners may need different services than economically disadvantaged students who are fluent in English, the cost of additional services may be similar for both groups of students. Like revenue limits and special education funding , EIA funding rates vary across distri cts, violating the principle of horizontal equity. Figure 3 shows these variations. The median funding rate for every group is close to $300 per pupil. However, f or every group except large high sc hool districts, the funding rate at the 7 5 th percentile is at least 10 percent higher than the rate at the 25th The level of funding is also an issue. EIA provides funding for additi onal student needs correlated with poverty and lack of English fluency. As argued above, we do not believe that cur rent research provides precise estimates of the relationship between school resources and student achievement . Consequently , that research cannot tell us how much additional revenue schools will need because they have high percentages of English learners or economically disadvantaged students. In this circumstance , we believe the best guidance comes from educational practitioners. On the basis of budget exercises with over 500 randomly selected California teachers, principals , and superintendents, Sonstelie (2007) concluded that to meet the state’s academic standards, a school in which every student was economically disadvantaged would need abou t $1,200 per pupil more in 2003– 04 than a school in which no student was disadvantaged. Adjusting for inflation, this figure would be about $1,500 per pupil in 2009 dollars. Part of this gap is closed by federal Title I funds , which target economically dis advantaged students. These funds are allocated to the state, which then allocates them to school districts. The process is complicated, b ut, on average, districts received $450 per di sadvantaged student in 2009 –10. Accordingly, based on the expertise of educational practitioners, EIA funding rates should be about $1,050 per disadvantaged student, instead of the current rate of about $300 per student. percentile 6 About 60 percent of disadvantaged students are fluent in English (Legislative Analyst’s Office 2007). http://www.ppic.org/main/home.asp Pathways for School Finance in California 18 FIGURE 3 Economic Impact Aid funding rates, 2009–2010 SOURCE: 2009 Economic Impact Aid Funding Results, California Department of Education. NOTE: Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in wh ich all schools are charter schools. Other Categorical Programs The most recent description of California’s categorical programs is presented in Weston, Sonstelie, and Rose (2009). Th is report describes each of the state’s more than 60 programs and provides data on the allocation of revenu es in each program for the 2005 –06 academic year. The largest program that year was special education. Economic Impact Aid was the fifth largest. In order of expenditure, t he other programs among the five largest were K –3 Class Size Reduction, the Targeted Instruction Improvement Block Grant, and Adult Education. Each state categorical program has its own funding procedures and restrictions. Although each program has its rationale, the sheer number of programs raises two key issues. First, because the funding procedures for programs can be complicated and because these procedures vary from program to program, it is very difficult to determine why one district receives more or less categorical revenue than another. Taken as a whole, the allocation of categorical revenue is not transparent, violating one of our principles. Second, the cumulative effect of program restrictions may have tipped the balance too far in the direction of state control over the use of funds. Th e latter issue received new attention in the Budget Act of 2009. In an effort to give local administrators more flexibility to absorb revenue cuts, the legislature granted spending flexibility for approximately 40 0 500 1,000 1,500 2,000 2,500 3,000 3,500 Small ElementaryMedium ElementaryLarge ElementarySmall High SchoolMedium High SchoolLarge High SchoolSmall UnifiedMedium UnifiedLarge Unified Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 19 categorical programs.7 A recent survey shows how districts have responded to this flexibility (Legislative Analyst’s Office 2010). Most districts reported shifting funds away from uses financed by programs in the flex item. In general, these funds have been shifted tow ard core classroom instruction. For programs in this so -called flex item, the legislature suspended restrictions on the use of funding, making these programs unrestricted general support. Special education and Economic Imp act Aid are not included in the fl ex item. This suspension of funding rest rictions continues through 2012 –13. At that point, the legislature could re - impose the restrictions it suspended. Alternatively, it could make the suspension permanent, in effect turning the pr ograms in the flex item into a source of unrestricted revenue similar to revenue limit funds. By doing so, the legislature would be taking a large step toward decentralizing decisions about how funds are spent. Before the legislature took this step, i t would surely review the composition of programs in the flex item. Additional programs might reasonably be added, and some programs now in the flex item might just as reasonably be excluded. Adult Education is a good example. In some areas of the state, commu nity colleges have assumed the primary role in providing adult education. In other areas, school districts have assumed this role. If the Adult Education program were turned into unrestricted revenue, adult education might suffer in the lat ter areas, but s chool districts in those areas would have a new source of unrestricted revenue not enjoyed by other districts. Other examples are the Regional Occupational Program and the Teacher Credentialing Block Grant, both of which support regional structures serving many school districts. A more general discussion of categorical flexibility might also consider changes to some programs presently excluded from the flex item. For example, the Targeted Instruction Improvement Block Grant, which is currently included in the flex item, might instead be consolidated with Economic Impact Aid, as was recently done with the English Language Acquisition Program in the 2010 Budget Act. If the legislature were to turn some version of the flex item into a permanent source of unrestricted revenue, it would also be forced to confront another issue. The revenue that districts receive from the flex item would no longer have a clear rationale, a reason why one district receives more or less revenue than another. As Figure 4 shows, funding rates for the flex item differ considerably across districts. 8 The median rates range from $711 per ADA for large elementary districts to $893 for medium high school districts. There are also large variations in funding rates within groups of districts. The difference between the rate in the 75 th percentile and the rate in the 25 th 7 For more details, see Weston, forthcoming. A list of programs is provided in the percentile is $566 per ADA for large unified districts. The difference is more than 20 percent of the median rate in every group. If the flex item were to become unrestricted support, it would clearly violate the principle of horizontal equity. technical appendix, Table A2. 8 Although K –3 Class Size Reduction (K –3 CSR) was not in the flex item, we have included 70 percent of revenue in that program in the flex item. The Budget Act of 2009 allowed districts to retain at least 80 percent of their class size reduction funds as lo ng as class sizes do not exceed 25 students. Even if class sizes exceed 25 students, districts retain 70 percent of their previous funds. Thus, at the very leas t, 70 percent of the previous year funding for K –3 CSR ought to be considered part of a district ’s flex item, a practice we follow in the data presented in Figur e 4. http://www.ppic.org/main/home.asp Pathways for School Finance in California 20 FIGURE 4 Flex item funding rates , 2009–2010 SOURCE: Funding Results (various programs) and 2009 Principal Apportionment , California Department of Education, Deferred Maintenance Program funding, Office of Public School Construction. NOTE: Necessary small schools and charter schools are excluded. We also exclude 79 districts with more than 75 percent of students in necessary small schools and two districts in which all schools are charter schools. 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000 8,500 9,000 9,500 10,000 10,500 11,000 Small ElementaryMedium ElementaryLarge ElementarySmall High SchoolMedium High SchoolLarge High SchoolSmall UnifiedMedium UnifiedLarge Unified Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 21 Simulating Pathways Our assessment of California’s school finance system has identified sever al potential improvements. Each would require additi onal funds if districts w ere held harmless as improvements were implemented. For example, horizontal equity requires funding rates to be equal ized. Equalization could be achieve d without additiona l revenue by reducing the rates above the average while raising rates below the average. However, l eveling down may not be politically feasible . A mor e acceptable approach would be to increase low rates more rapidly tha n high rates over time as the state invests in its public schools . In this section, we simulate that process using a number of assumptions about fut ure growth in revenue and enrollments. We have chos en four improvements to simulate , illustrating a range of possibilities. The first is to equalize funding rates for revenue limits. The second scenario simulates the process of turning the flex item into a source of unrestricted aid and equalizing funding rates over time . The third scenario increase s the EIA funding rate to $1,050 per student. Over time, the finance s ystem in this scenario would resemble the recommendation of the Governor’s Committee on Education Excellence (2007). The fourth scenario adjust s program rates for regional cost differences. The endpoint in this scenario is a school finance system similar to that recommended by Bersin, Kirst, and Liu (2008) . Parameters The simulations begin with a baseline in 2009– 10. School districts receive the revenue in the simulation that they actually received in that year from revenue limits, special education, Economic Impact Aid, and th e flex item. Th e funds in these programs total ed $39 billion in 2009–10. Each subsequent year in the simulation, the total increases as economic and demograph ic conditions permit, and the additional revenue allows for increases in each district’s funding rates for each of the four programs. T he additional revenue comes from normal economic growth and from demographic trends likely to unfold over the next 2 0 years. Growth in school revenue is tied to economic growth and demographic trends through the following equation: The equation isolates two important factors in the state’s ability to provide revenue for its public schools. The first is a demographic fac tor, the number of taxpayers per student. A given tax burden for the average taxpayer will produce more revenue per pupil for schools if there are more taxpayers per student. The second factor, average income per taxpayer, represents the ability of taxpayers to bear a given tax burden. Below, we investigate trends since 1970 in each of t he three terms in the equation above. Our objective is t o identify reasonable assumptions for our simulations and to put those assumptions in perspective. We conduct our simulations in real terms, adjusting for inflati on. Because our focus is the purchasing power of school districts , we use the Implicit Price Deflator for State and Local Government to deflate nominal figures for bot h school expenditures and personal income. http://www.ppic.org/main/home.asp Pathways for School Finance in California 22 The first term in the equation is public school expenditures in California as a percentage of state personal income. We do not have data on revenue in the four funding programs extending back in time. The best data we have is the current e xpenditures of public schools, which is approximately equal to total revenue schools received for operating expenses. The revenue in the four programs we focus on is currently about three- fourths of that total. In the early 1970s, public school expenditures were over 4 percent of state personal income (Figure 5). After Proposition 13 in 1978, that percentage fell steadily, reaching a low of 3 percent in 1984. It has rebounded steadily since that time with a high near 4 percent in 2003. In 2006–07, current expenditures for California public schools were 3.8 percent of state person al income. From 1980–81 to 2006– 07, that ratio has averaged 3.6 percent. FIGURE 5 Public school expenditures as a percent of state personal income, 1970 –2007 SOURCE: Public school expenditures in California are from the National Center for Education Statistics, State Education Facts , 1969– 1987, and Digest of Education Statistics, 1998, 2001, and 2009. State personal income is from the Bureau of Economic Analysis, Department of Commerce. NOTE: Public school expenditures are current expenditures for elementary and secondary education. The expenditure data in Figure 5 comes from the National Center of Education Statistics . Although the Center provides a consistent and reliable source of data extending back for many years, the most recent data are for 2006– 07. However, w e can update th ese data from other sources. From the data in the Governor’s Budget, revenue limit funds pl us state categorical revenue fell about 11 percent from 2006 –07 to 2009 –10. Some of the reduction in funds was replaced by a temporary increase in federal revenue. Over the same period, personal income in California rose by about 4.6 percent. Ignoring the temporary increase in federal revenue, the ratio of current expenditures to personal income fell to approximately 3.2 in 2009 –10. The first assumption for our simulation is that this ratio will return gradually to its average level of 3.6 percent and that revenue in the four programs will rise by the same percenta ge. In particular, as total revenue rises from 3.2 percent of personal income to 3.6 percent of personal income, revenue in the four programs is assumed to rise from its current value of 2.5 percent of personal income to 2.8 percent of personal income . In the simulations, this increase is phased in uniformly ov er time with the ratio increasing by 0.014 percentage points each year until it reaches 2.8 percent in 2030. 0% 1% 2% 3% 4% 5% Expenditures as Percentage of Personal Income http://www.ppic.org/main/home.asp Pathways for School Finance in California 23 We believe this assumption is relatively conservative for four reasons. First , in computing the average ratio of exp enditures to personal income, we excluded the years before Proposition 13, in which the rate was significantly higher than in subsequent years. Second, as noted by the California Budget Project (2010), public school expend itures as a percentage of personal income have been lower in California than in the rest of the nation for many years. Our simulations assume that in the case of K –12 education, California continues to be a relati vely low-spending state. Third , for every y ear in the simulation except the last, the ratio of revenue to personal income is below its average from 1980 to 2006. Fourth, we assume that revenue in the four programs is the same fraction of total revenue as it is currently. Revenue for programs we have not included, such as necessary small schools, is implicitly assumed to grow at the same rate as revenue for the programs we have included. With this assumption, the growth in personal income is an important factor in determining the growth in revenue. To separate economic and demographic factors, we have decomposed personal income per student into two parts: personal income per taxpayer and taxpayers per student. We have used age to partition California residents into taxpayers and others, including stu dents. Taxpayers are residents 18 years and older. In general, younger residents are either in school or younger than school age. As Figure 6 shows, personal income per taxpayer has grown steadily since 1970, punctuated by economic recessions in 1981, 1990, 2001, and 2008. In the figure, personal income is expressed in real terms, using the Implicit Deflator for State and Local Government to adjust for inflation. With that adjustment, real personal income per taxpayer has grown at an average rate of 0.5 percent since 1970. The dashed line in the figure represents this average growth rate. FIGURE 6 Real personal income per taxpayer , 1970–2008 SOURCE: State Personal Income and the Implicit Price Deflator for State and Local Government is from the Bureau of Economic Analysis, Department of Commerce. Data on California residents 18 years of age and older are from the State of California, Department of Finance, Race/Ethnic Population with Age and Sex Detail, 1970 –1989, 1990– 1999, and 2000 – NOTE: Taxpayers are California residents 18 years of age and older. State Personal Income is deflated by the Implicit Price Deflator for State and Local Governments. 2050. In our simulation s, we assume that real personal income per taxpayer continues to grow at 0.5 percent per year. To be clear, this assumption is not a forecast of economic growth in California. Our purpose is not to 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 Personal Income per Taxpayer (2009 $) http://www.ppic.org/main/home.asp Pathways for School Finance in California 24 forecast , but to work through the long -term consequences of certai n policies using realistic assumptions. Fo r that purpose, we believe our assumed growth rate is appropriate. By adjusting for inflation using the Implicit Deflator for State and Local Government instead of the Consumer Price Index, we are also accounting for the fact that the salaries of state and local government workers has tended to grow faster than the general price level. In that sense, our assumed growth rate represents the growth in personal income in excess of that necessary to maintain a constant level of public education services. Demographic trends unfold more smoothly than economic trends over time. As shown in Figure 7, the number of taxpayers per student rose steadily in California from 1970 through 1986. In 1970, there were slightly less than three taxpayers per student. By 1986, that ratio had risen to 4.65, a significan t decline in the tax burden public schools placed on taxpayers. From 1986 to 1999, the ratio fell to a low of 4.08. It has risen slightly since then. FIGURE 7 Taxpayers per student, 1970– 2007 SOURCE: Data on California residents 18 years of age and older are from the State of California, Department of Finance, Race/Ethnic Population with Age and Sex Detail, 1970–1989, 1990 –1999, and 2000 –2050. Data on public NOTE: Taxpayers are California residents 18 years of age and older. Students are students enrolled in California public schools . school enrollments in California are from the Nation al Center for Education Statistics, State Education Facts , 1969 –1987 , and Digest of Education Statistics, 1998, 2001, and 2009. Because demographic trends are tied to the aging of the existing population, they are more predictable than economic trends. According to our projections based on California Department of Finance data, school enrollment should rise by about 20 percent ov er the next twenty years, while the population of California adults over age 18 rises almost 30 percent. 9 9 Although much of this growth will be among older adults ages 65 to 74, seniors in California tend to have good economic outco mes. For example, poverty rates for 65 - to 74 -year -olds in California are lower than for any other age group (7.9% in 2008 compared to 13.6% for all other age groups, based on American Communtiy Survey data). Of course, use of public health programs is higher at older ages, but f ederal programs provide most of tha t support. Thus , taxpayers per student should rise somewhat. 0 1 2 3 4 5 Taxpayers per Student http://www.ppic.org/main/home.asp Pathways for School Finance in California 25 Our simulation uses enrollment projections from the Department of Finance to project attendance and EIA counts for each school district. For years beyond the base year, we project ADA by applying growth rates to the base year ADA. These growth rates are based on county -level enrollment derived from population projections by the Department of Finance. For the EIA student count, which is not AD A, we use the EIA count in 2008 –09 as a percentage of ADA in the district. Table 2 summarizes our economic and demographic assumptions. The implication of thes e assumptions is a 31 percent increase in real expenditures per student. They rise from $7,022 in 2009 to $9,206 in 2030. TABLE 2 Year Projections of economic and demographic trends Taxpayers Personal income per taxpayer (2009$) Average daily attendance Taxpayers per student Revenue per student s (2009$) 2009 28,695,960 54,516 5,543, 291 5.18 7,022 2010 29,146,279 54,789 5,546, 543 5.25 7,199 2015 31,312,124 57,583 5,630, 917 5.56 8,007 2020 33,244,039 60,521 5,896, 815 5.64 8,532 2025 35,157,589 63,608 6,329, 462 5.55 8,835 2030 37,076,944 66,852 6,732, 870 5.51 9,206 This growth is consistent with the basic tenets of Proposition 98, which provides a minimum guarantee for revenue in California public schools and community colleges. The guarantee involves several complicated conditions, but the central condition is that, in normal economic times, revenue per student should grow at least as fast as per capita personal income. In our simulations, the growth rate for revenue per student exceeds this Prop 98 growth rate initially as taxpayers per student ri se s. It then falls below the Prop 98 growth rate for the remaining years. In 2030, revenue per student in our simulation is approximately $500 less per student than if revenue per student had grown at the rate of personal income from 2009 to 2030. Scenarios Each scenario involves a different allocation of additional revenue. In the base year of 2009 –10 , the revenue each district received in each of the four programs we focus on can be expressed as a funding rate (dollars per student) multiplied by a particular count of students in the district. The simulations change those per - pupil funding rates each year. We describe four scenarios below which reflect a range of state priorities. To provide a baseline for these scenarios, we also demonstrate how funding rates change over the next 20 years if the state relies solely on its current mechanisms of adjusting revenue rates for inflation. Status Quo In this status quo baseline, each district’s statutory revenue limi t rate increases annually for inflation. As described earlier, this inflation increase is the same dollar amount for all districts of the same type, but the increases vary in proportional terms , depending on whether the district’s rev enue limit is above or below the average for its type. In simulating these increases, we have assumed an inflation rate of 4.83 percent, the 30 -year average rate for the Implicit Price Deflator for State and Local Government. This baseline trajectory also adjusts http://www.ppic.org/main/home.asp Pathways for School Finance in California 26 the flex item, special education, and EIA rates for inflation, but the adjustment works differently. In these three programs, the funding rates all increase by the inflation rate. In real terms, these rates do not change. This baseline incorporate s an additional feature of California’s school finance system. Recent declines in the state budget have caused the state to appropriate revenue limit funds based on rates that were 18 percent lower than each district was entitled to by statute. We calculat e the revenue limit inflation increases using the statutory rates and add th e increases to th ose rates. However, the actual revenue limit rates assigned to districts are based on the available funds in a given year. Those available funds are driven by the economic and demographic factors highlighted in Table 2. Each year, the additional revenue is divided into two parts. The first is the amount necessary to adjust the prior year’s rates for the flex item, EIA, and special education for inflation , and to fun d the new year’s level of ADA at those rates. This amount can differ in real terms from the total revenue provided to districts in the previous year in those programs because ADA changes from year to year. After adjusting the three programs for inflation , the remaining new funds are used to bring revenue limit rates up to their new statutory levels. If the total revenue available for this purpose is only 90 percent of the amount required to bring all districts to their statutory levels, each district’s rate is then 90 percent of its statutory rate. Once state revenue in this model has grown sufficiently to fund all districts at their statutory revenue limits, we continue to make inflation adjustments in all programs . At this point, we also t rack the addition al funds available for reform efforts other than inflation adjustments. Figure 8 show s the results from simulating this baseline. For each year, the figures plot the 90 th, 75th, 50th, 25th, and 10 th percentiles of per -pupil revenue in the specified program. These percentiles are based on students not districts. To compute these values, all students are ranked based on the funding level associated with their district, and the percentiles are extracted from this list. For example, the 10 th percentile refers to the funding level received by the district of the student who ranks above the bottom 10 percent of the students in the list. The difference between the 90 th and 10th percentiles gives the range of funding levels experienced by the middle 80 percent of students. The figure also shows the funds available for education once the deficit factor has been restored to the revenue limits. These figures show revenue in real 2009 dollars. http://www.ppic.org/main/home.asp Pathways for School Finance in California 27 FIGURE 8 Status quo baseline In this status quo approach, revenue limit funds increase over time, but the amount of equalization across all districts is modest. The rates rise dramatically from 2009 through 2017, at which point they level off in real terms . The year 2017 marks the point at which the deficit factor has been eliminated and districts receive their inflation -adjusted statutory revenue limit rate. At the same time, funds begin to accumulate for other potential purposes. In 2017, those funds are small, averaging $63 per pupil. By 2030, however, those funds grow to $1,039 per pupil, providing scope for many reform efforts. The flex item, EIA, and special education programs maintain their original real rates throughout the scenario’s horizon. The dip in the 90 th percentile of the flex item occurs because enrollment s in districts with the highest flexible funding in the base year are projected to shrink relatively faster than other districts. Thus, in the future, the 90 th In Figure 8, the percentile lines for revenue limits essentially represent different district types. The 90 percentile student is in a district with l ower flexible funds. th percentile district is a high school district, the median line represents a unified district, and the 25 th 02004006008001,0001,2001,4001,6001,8002,000Flex Item90th75thMedian25th10th$ 2009 0 200 400 600 800 1,000 1,200Ec onom ic Im pa c t Aid $ 2009 $ 2009 0 200 400 600 800 1,000 1,200Special Education $ 2009 0 200 400 600 800 1,000 1,200Additiona l R e v e nue $ 2009 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000Revenue Limit $ 2009 percentile is an elementary district. Because the inflation increases are different for each district type, th ese percentiles do not converge. http://www.ppic.org/main/home.asp Pathways for School Finance in California 28 The lack of overall convergence among revenue limit rates is not surprising, given that the goal of the inflation increase mecha nism in revenue limits is to equalize rates within district types and not across all types. As Figure 1 showed, rates vary dramatically across district types but within the same type of district, there is less variation. In 2009, elementary districts avera ged a rate of $5,007, with only $62 separating the 90 th from the 10th percentile. High school districts averaged a rate of $6,016 with $87 separating the 90th from the 10 th percentile. Unified districts averaged $5,239 with a $79 gap between the 90th and 1 0th percentile. The inflation equalization mechanism does work within district types; these gaps do shrink over time. More dramatic are the narrowing gaps between the maximum revenue limit rates and the 90 th As discussed earlier, justificati ons for the current differences in revenue limit rates by district type are not entirely transparent, nor are the differences an explicit state policy. Thus, our subsequent scenarios provide a way of equalizing these rates across district types as well. percentile. By 2030, they fall in half from thei r 2009 levels of $2,944, $693, and $1,542 for elementary, high, and unified districts, respectively. This narrowing occurs primarily because the inflation increase reduces the real funding rate for high revenue districts. Re venue Limit Focus The first of these scenario s presents a more aggressive approach to equalizing revenue limits. The economic and demographic trends in Table 2 still determine the additional flow of revenue into the system, but our simulations divide that revenue in a different way. For each year, the revenue is divided into two amounts. The first is the amount necessary to fund the four programs in each district using the real per- pupil funding rates of the previous year. It is the amount necessary to hold districts harmless. Like our status quo example, this amount can differ from the total revenue provided to districts in the previous year because of changes in ADA. The difference between revenue for the four programs and the hold- harmless amount is new revenue available to increase funding rates. This first scenario allocates all of those additional funds to the revenue limit. We divide these additional funds among districts in a way that achieves revenue equalization across all districts over time. We f ocus on equalizing per -pupil revenue at the 90 th percentile of base year funding. Districts that begin a year with a funding rate below th at equalization target are entitled to a portion of the a dditional revenue. We allocate the additional revenue in the following way. We determine the difference between the total revenue each district would receive at the funding rates of the previous year and what it would receive at the equalization target r ate. We add up those deficits across all districts and compare that sum to the additional revenue available for revenue limits. If the additional revenue represents 30 percent of the total deficit, then each district below the target receives additional funds in that program equal to 30 percent of its deficit . That allocation defines the rate for the next year, and we repeat the process. This equalization mechanism directs the largest amounts of funding in absolute terms to districts furthest from the 90 th percentile. Districts only receive enough funds to brin g them to the equalization target. Districts with current rates above the target receive no additional funds until all districts have reached the target. Once all districts have reached th at original 90 th percentile target, we consider revenue limits funds to have equalized, and additional funds are dispersed on an equal per -pupil basis. At this point, the absolute gap between the 90 th percentile and higher percentiles will be constant. However, this process achieves some equalization even after all distric ts have achieved the original 90 th percentile rate, because districts above the 90 th percentile receive less in percentage terms than those at the 90 th percentile. Although the equalization targets are constant throughout the planning horizon, the rates and deficits change yearly, so additional funds are http://www.ppic.org/main/home.asp Pathways for School Finance in California 29 allocated differently each year. Figure 9 shows the pathway for revenue limits under this more aggressive equalization scheme. Since additional funds are allocated to the flex item, special education , and EIA in amounts that keep their real rates constant, the pathways in those programs look like their corresponding trajectories in Figure 8. FIGU RE 9 Revenue limit focus scenario Directing the additional funds to the revenue limit program brings all districts up to the 90th percentile of nearly $6,000 by 2013. From that year on, all additional revenue continues to flow into the revenue limit. By 2030, the 90 th Although this scenario directs all available additional resources to the revenue limit program and ultimately restores the current 18 percent deficit factor for all districts, the complete restoration occurs at different times for different districts. Elementary and unified districts eliminate their deficit factor first, because the revenue limit rates are below those of high school districts. Another option would be to restore the revenue limit rates to all districts before equalizing across district types. The pathway for such an approach would look like a combination of t he revenue limit trajectories in Figures 8 and 9. percentile rate climbs to $7,486, an increase of about 25 percent from the equalized value. This scenario demonstrates that , with focused attention on this goal, it is possible to achieve horizontal equity for revenue limits in a short amount of time. This revenue limit scenario demonstrates how the state could equalize the largest revenue program quickly, but it ignores the unequal distributions in other programs. Although Figures 2 and 3 demonstrate some dispersion in the funding rates for EIA and special education in 2009, the gaps are small relative to the level of funding in the revenue limit program . Furthermore, the gaps could be closed over time by redirecting a very small share of additional r evenue from the revenue limit rate to these programs that serve disadvantaged students. Only $96 separates the 90 th and 10th percentile students in special education , and the range is only $66 for EIA. Directing 1 percent of new revenue to EIA and 3 percen t to special education would bring all districts up to the current 90 th 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000 Revenue Limit 90th 75th Median 25th 10th$ 2009 percentile by 2030. This example assumes we use the same equalization mechanism for these programs as for the revenue limit. http://www.ppic.org/main/home.asp Pathways for School Finance in California 30 This modified scenario leaves 96 percent of additional fun ds available for the revenue limit program. This change causes a one -year delay for revenue limit equalization. The rates equalize in 2014, and that equalized rate grows to $7,396 by 2030. This level represents only a slight dip from the original revenue limit scenario. Unlike EIA and special education, the flex item rates exhibit substantial variation in 2009. Equalizing this program is the focus of our next scenario. Flex Item Focus As shown in Figure 8, the flex item exhibits a great deal of variation i n the base year of 2009, with $1,335 separating the 90 th and 10th percentiles. This gap is nearly twice the median flex item rate, and it represents about 21 percent of the average revenue limit and flex item rates combined. Nearly two-thirds of this gap o ccurs between the 90 th and 75th Compared with equalization mechanism in the prior scenario, this scenario slightly modifies the equalization target for the flex item. Because of the natural attrition of students in distric ts that currently receive high flex item rates, the 90 percentiles. Equalizing this program requires substantial additional funds. Our second scenario equalizes the flex item program by 2030, while continuing to achieve equalization in the special education and EIA programs in t hat year as well. To achieve this goal, 30 percent of new revenue each year is allocated to the flex item. As in the modified revenue limit scenario, 1 percent of additional funds are allocated to the EIA program and 3 percent to special education. The revenue limit receives the remaining 66 percent of additional funds. th percentile decreases naturally over time. Rather than use the equalization target of $1,915 ( the 90 th percentile in the base year ), this scenario uses a target of $1,539 ( the 90th percentile in year 2030 in the status quo scenario ). Every year in the simulation, districts with rates less than this target are entitled to a portion of the additional funds directed to the flex item. Like the previous scenario, the gap between each district ’s rate and the target is closed by an identical proportion, where the proportion is the share of new flex item funds to the total gap. Figure 10 shows the results from this simulation. FIGU RE 10 Flex item focus scenario 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000Revenue Limit $ 2009 02004006008001,0001,2001,4001,6001,8002,000Flex Item90th 75th Median 25th 10th$ 2009 0 200 400 600 800 1,000 1,200Ec onom ic Im pa c t Aid $ 2009 0 200 400 600 800 1,000 1,200Special Education $ 2009 http://www.ppic.org/main/home.asp Pathways for School Finance in California 31 Diverting such a large share of funds away fro m the revenue limit and to the flex item delays revenue limit equalization until 2019. Although the median revenue limit increases 30 percent from 2009 to 2030, the equalized revenue limit is about $675 lower than in the last scenario; it reaches only $6,721 instead of $7,396. Although this second scenario satisfies the principle of horizontal equity, it ignores the issue that the costs of educating disadvantaged students may be higher than for other students. Our next scenario addresses this issue. Econo mic Impact Aid Focus Our third scenario focuses on increasing funding to EIA , following recommendations by the Governor’s Committee on Education Excellence (2007) and Bersin, Kirst, and Liu (2008). Both recommendation s aim to close the achievement gap between disadvantaged students and other students. This scenario allocates 17 percent of new funds to the EIA program . The goal is a n EIA funding rate of $1,050 . To accommodate this emphasis on disadvantaged students, the share of new funds going to the revenue limit program drops from 66 percent, as in the prior scenario, to 50 percent. The share of new funds going to the flex item and special education remain at their levels from the previous scenario. Figure 11 shows the pathways for the revenue limit and EIA under this new allocation. Pathways for the special e ducation and the flexible categorical programs do not change from the previous figure. FIGURE 11 Economic Impact Aid focus scenario With this shift in priorities, all students are in districts receiving an EIA rate of at least $1,081 by 2030. This growth in EIA rates come at the expense of revenue limit rates. Compared with the previous scenario, revenue limit rates equalize in the same year, but the equalized rate reaches only $6,368, a $353 shortfall from the prior scenario. Comparing Scenarios The scenarios we present a re meant to demonstrate a range of possibilities for future funding. They all involve tradeoffs. We have presented them in a way that shows their cost relative to funding in the revenue limit program. If a relatively high fraction of additional funding is allocated to the flex item, funding rates in the program converge by 2030 and the median funding rate grows from $852 per ADA in 2010 to $1 ,541 per ADA by 2030 (Table 3). However, the projected 2030 revenue limit rate is about $675 per ADA lower, and equalization of revenue limit rates occurs five years later. The goal of increasing funding for disadvantaged students, as in our EIA focus scena rio, does not delay the revenue limit equalization by eve n a full year, but 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000Revenue Limit $ 2009 0 200 400 600 800 1,000 1,200Ec onom ic Im pa c t Aid90th 75th Median 25th 10th$ 2009 http://www.ppic.org/main/home.asp Pathways for School Finance in California 32 it does reduce the 2030 projection of that rate by about $353 per ADA. H owever, t he EIA funding rate is $1, 081 per ADA in 2030, as compared to $317 under the flex item or modified revenue limit focus. Other scenarios are certainly worth investigating , and hopefully our simulations have provided a framework for considering such extensions. TABLE 3 Programs Summary of scenarios Revenue limit f ocus Revenue limit + m odifications Flex item focus Economic Impact Aid focus Revenue limit Allocation share (%) 100 96 66 50 Year of equalization 2013 2014 2019 2019 Median rate 2010 5403 5395 5337 5306 Median rate 2020 6806 6743 6275 6030 Median rate 2030 7486 7396 6721 6368 Flex item Allocation share (%) 0 0 30 30 Year of equalization Never Never 2030 2030 Median rate 2010 790 790 852 852 Median rate 2020 780 780 1312 1313 Median rate 2030 776 776 1540 1541 Economic Impact Aid Allocation share (%) 0 1 1 17 Year of equalization Never 2030 2030 2010 Median rate 2010 313 317 317 377 Median rate 2020 314 346 346 846 Median rate 2030 314 361 361 1081 Special education Allocation share (%) 0 3 3 3 Year of equalization Never 2030 2030 2030 Median rate 2010 638 644 644 644 Median rate 2020 637 693 693 693 Median rate 2030 637 717 717 717 Regional Cost Differences None of the previous scenarios takes into account differences in regional wages , yet teacher compensation (salaries plus benefits) varies substantially across the state. In 2003– 04, teachers with the same level of education and experience averaged compensation just under $55,000 per year in the North Coast and Yolo Counties but over $70,000 annually in Santa Clara and Orange Counties (Rose et al. 2008). 10 10 The North Co ast Counties include Del Norte, Humboldt, Lake, and Mendocino C ounties. These differences in teacher compensation a re highly correlated with regional differences in the wages of college -educated workers who are not teachers. Non-teacher wages provide one way of measuring the purchasing power of revenue in a scho ol district. For example, if labor costs are 20 percent higher than average in District A and http://www.ppic.org/main/home.asp Pathways for School Finance in California 33 this district receives the same level of per -pupil funding for hiring teachers as the average district , then Di strict A could only purchase 83 percent (1.0/1.2) o f what the average district could purchase. To show how differences in purchasing power affect the distribution of reso urces, we adjust each district’s original rate in each of the four programs by an index of non -teacher wages. This regional wage index ( RWI) is described in Rose and Sengupta (2007). 11 This index groups counties into 30 regions, based on the U.S. Census Bureau’s definition of metropolitan statistical areas, and computes the average wage of college- educated non -teachers in each region. Each region is assigned an index value based on the ratio of its regional average wage to the state’s average wage, where the state average is weighted by the number of students in each region. The index ranges from a high of 1.2 in Santa Clara C ounty to a low of 0.8 in the North Coast Counties. Because salaries comprise about 80 percent of district budgets, we divide that portion of a district’s program rates by the RWI to obtain an RWI adjusted rate. Figure 12 shows the variation in funding before and after th ese adjustments for 2009. FIGURE 12 Program rates adjust ed by regional wage index, 2009– 2010 The most striking feature of the adjustment is the significant widening in revenue limit rates. Although actual revenue limits are quite similar, their purchasing power varies widely. Without the adjustment, the gap between the 75 th and 25th percentiles is imperceptible, measuring only $22. With the adjustment, $353 separates those percentiles. With the regional wage adjustment, the distribution of specia l education funds widens slightly. The gap between the 75 th and 25th 11 We update the index to include 2005 data from the Occupational Employment Survey. percentiles increases by $13, suggesting that some districts with the higher special education rates exhibited a slight tendency to be located in areas with lower regional wages. The dist ribution of flex item and EIA program rates, however, changed very little with the regional wage adjustment, suggesting there is no systematic relationship between those rates and regional wages. 5,000 5,250 5,500 5,750 6,000 ActualRWI-Adjusted Revenue Limit Dollars per ADA 500 750 1,000 1,250 1,500 1,750 2,000 ActualRWI-Adjusted Flex Item Dollars per ADA 600 650 700 750 ActualRWI-Adjusted Special Education Dollars per ADA 250 300 350 400 ActualRWI-Adjusted Ec onom ic Im pa c t Aid Dollars per ADA http://www.ppic.org/main/home.asp Pathways for School Finance in California 34 To demonstrate the effect of compensating districts for reg ional differences in labor costs, we conduct a final simulation that repeats our original revenue limit scenario but equalizes the RWI -adjusted rates rather than the actual rates. The equalization mechanism works like that in the revenue limit scenario wit h the following exception. The equalization target is based on the 90 th percentile of the RWI -adjusted revenue limit rate in 2009, and districts are only entitled to additional revenue in a given year if their program ’s RWI -adjusted rate is below that targ et rate. Figure 13 shows the resulting pathway for the RWI -adjusted revenue limit rates. FIGURE 13 Simulated pathway for revenue limit rates adjusted by regional wage index In this scenario, the districts with RWI-adjusted rates below the target reach the target in 2013, the same year as the original revenue limit scenario. The pathways for the original and adjusted revenue limit rates look remarkably similar. However, the set of districts receiving revenue in these two scenarios differs. The correlation coefficient between gains in these two scenarios is 0.56. Table 4 shows the average gains for districts with and without the regional wage adjustment. Districts are categorized by type and whether they are in a high -, medium -, or low -wage regio n. About one -third of students are in low -wage districts and one - quarter are in high -wage districts . We chose these thresholds because they best accommodate the distribution of regional wages. 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000 Revenue Limit 90th 75th Median 25th 10th $ 2009 http://www.ppic.org/main/home.asp Pathways for School Finance in California 35 TABLE 4 District classification Average gains in revenue limit rates between 2009 an d 2013 ($) Original revenue limit scenario RWI-a djusted revenue limit scenario Elementary districts Low wage 878 393 Medium wage 886 824 High wage 880 1,215 High school districts Low wage 21 40 Medium wage 21 40 High wage 21 367 Unified districts Low wage 583 246 Medium wage 666 604 High wage 689 883 Without adjusting for regional wages, elementary districts gained the most while high school districts gained very little. With the adjustments, however, the gains for elementary districts in low -wage regions fall by more than 50 percent and those in high wage regions rise by about 40 percent . Furthermore, high school districts in high wage regions receive some additional funds when regional wages are taken int o account. This scenario demonstrates that, over time, the state could level the playing field by increasing the revenue of districts in high -wage regions. http://www.ppic.org/main/home.asp Pathways for School Finance in California 36 Conclusions By some accounts, California’s school finance system is so fundamentally flawed that only a complete overhaul could fix it. We do not share this view. The current system is flawed, but the basic elements of a sound system are already in place. These elements need to be strengthened, which can be accomplished steadily over time as economic and demographic conditions permit. Our numerical simulations illustrate the possibilities. Of course, t he results of our simulations depend on assumptions we have made about several economic and demographic factors and trends. For the most part, we have b een conservative in our assumption s, accepting a relatively slow growth rate in revenue per pupil. That is not to say that the future is certain to be rosier than our assumptions. One area of current concern is the looming fiscal pressure from a host of entitlement programs that promise to expand as the population ages. And in ten years, that concern will surely be replaced by another that is not apparent to anyone now. Despite these uncertainties, our simulations illustrate a simple and clear message. Even if the growth in revenue for public schools is relatively slow, steady improvement can achieve a great deal in time. Nonetheless, s teady improvement does require two difficult steps. The first is to formulate a clear vision of what the system should look like. The second is to create a mechanism to ensure that steady progress is made toward that goal. Our assessment of California’s system has identified a long menu of potential improvements. The simulations have focused on four: equalizing funding rates f or the core program, turning some current categorical programs into unrestricted support, increasing funding for districts with high percentages of economically disadvantaged students, and adjusting funding rates for regional differences in labor costs. We present these simulations , not as recommendations for a plan the state should adopt, but as illustrations of the type of analysis that can help the state develop a long -range plan. We welcome the opportunity to simulate other options if the state undertak es such an effort . Any long -range policy should also consider a number of other issues and questions:  Small schools in rural areas. The current approach to funding small schools in rural areas does not provide incentives for districts to find efficient way s of educating students in these areas. Would an approach based on an external measure such as population density be preferable?  Funding base for special education. The current system for funding special education has removed fiscal incentives to identify learning disabilities, but it may not adequately recognize cost differences among districts. Should the percentage of economically disadvantaged students in a district be part of the funding base for special education?  Funding base for Economic Impact Aid. English learners are now part of the funding base for Economic Impact Aid, which inadvertlently reduces funding for districts that are particularly effective in educating these students. Because most English learners are also economically disadvantaged, s hould the funding base for Economic Impact Aid be economically disadvantaged students instead of economically disadvantaged students and English learners?  Categorical flexibility. The legislature temporarily suspended restrictions on forty categorical programs, and we used this list of programs to demonstrate the process of turning categorical funding into permanent unrestricted support. If the suspensions were made perm anent, which programs should be excluded, which should remain, and what other progra ms should be included? http://www.ppic.org/main/home.asp Pathways for School Finance in California 37  Grade span. Revenue limit base rates tend to be higher for high school districts than for elementary districts. Should these differences be made explicit by enacting different base rates for students in different grade spans , as is currently the case in charter schools ?  Adjustments to revenue limit entitlements. Although most of a district’s revenue limit entitlement is determined by its base rate and attendance, a number of other adjustments are made to reach the district’s entitleme nt. Should those adjustments be phased out over time?  Excess taxes. If a district’s property tax revenue exceeds its revenue limit, it retains these excess taxes. Should excess taxes be refunded to taxpayers instead? Decisions on these issues would help form a clea r vision for the state to follow over time. Of course, c onsistency over time is a challenge, particularly for a term -limit legislature. However, California does have a precedent. Each year, every district’s base rate for revenue limit funding ch anges according to a formula set in statute. This formula updates base rates for inflation, but in a way that gradually equalizes base rates over time. This annual updating produces an appropriation of state aid, which does not require legislative action. Statutory appropriation is certainly an efficient mechanism for implementing a steady change over time. The statutory appropriation for revenue limits also has a provision for accommodating variations in state revenue. If economic conditions cause a downtu rn in state revenue, the legislature may decide that it cannot afford to fully fund the revenue limit appropriation in a particular year, creating a deficit factor for that year. In other years, it has used an abundance of revenue to raise funding rates fo r low-revenue districts. These ad hoc decisions could be made a routine part of an effort to create a rainy day fund for the state. In years in which the growth in state revenue exceeded the growth in the statutory appropriation, the difference could be se t aside in a separate fund. In years in which the revenue growth rate fell short of the growth rate in the statutory appropriation, the fund could be tapped to fund the statutory appropriation. In that way, changes would be phased in steadily over time, al lowing school districts to implement long -term strategies based on realistic assumptions of future revenue. None of the reforms discussed in this report will magically transform California’s public schools. A state’s school finance system is only a foundation. If well designed, it provides districts the revenue to employ the resources they need. Given this opportunity, the question then b ecomes whether districts will use their revenue effectively. But if they are not given th e opportunity, it is hard to see how they can otherwise be successful. http://www.ppic.org/main/home.asp Pathways for School Finance in California 38 References Angrist, Joshua, and Victor Lavy . 1999. “Using Maimonides’ Rule to Estimate the Effect of Class Size on Children’s Academic Achievement .” Quarterly Journal of Economics 114(2) : 533– 75. Angrist, Joshua D., and Jorn -Steffen Pischke. 2010. “The Credibility Revolution in Empirical Economics: How Better Research Design is Taking the Con O ut of Econometrics.” Journal of Economic Perspectives 24(2) : 3 –30. Bersin, Alan, Michael W. Kirst, and Goodwin Liu . 2008. Getting Beyond the Facts: Reforming California School Finance . Chief Justice Earl Warren Institute on Race, Ethnicity, and Diversity Issue B rief. University of California, Berkeley, California. Betts, Julian R., Andrew C. Zau, and Cory Koedel . 2010. Lessons in Reading Reform: Finding What Works Brewer, Dominic J., and Joanna Smith . 2006. “Evaluating the ‘Crazy Quilt’: Educational Governance in California .” Center on Educational Governance, Rossier School of Education , University of Southern California. . San Francisco: Public Policy Institute of California. California Budget Project . 2010. “Race to the Bottom? California’s Support for Schools Lags the Nation .” School Finance Facts. Chambers, Jay, Jesse Levin, and Danielle De Lancey . 2006. “Efficiency and Adequacy in California School Finance: A Professional Judgement Approach .” American Institutes for Research . G andara, Patricia, and Russell W . Rummberger. 2006, Resource Needs for California English Learners. University of California Linguistic Minority Research Institute. Governor’s Committee on Education Excellence. 2007. Students First: Renewing Hope for California’s Future. Hanushek, Eric A. 1997. “Assessing the Effects of School Resources on Student Performance: An Update .” Educational Evaluation and Policy Analysis 19(2) : 141– 64. Hoxby, Caroline M. 2000. “The Effects of Class Size on Student Achievement: New Evidence from Population Varia tion.” Quarterly Journal of Economics 115(4): 1239–85. Imazeki, Jennifer . 2006. “Assessing the Cost of K –12 Education in California Public Schools .” San Diego State University . Krueger, Alan B. 1999. “Experimental Estimates of Education Production Functions. ” Quarterly Journal of Economics 114(2) : 497 –532. Krueger, Alan B. 2002. “Understanding the Magnitude and Effect of Class Size on Student Achievement .” In The Class Size Debate, ed. Lawrence Mishel and Richard Rothstein (Washington, DC: Economic Policy Institute ). Legislative Analyst’s Office . 2007. “English Learners .” Analysis of the 2007 –08 Budget Bill: Education (February 21) . Legislative Analyst’s Office . 2010. “Update on School District Finan ce and Flexibility” (May 4). Little Hoover Commission . 2008. “Educational Governance and Accountability: Taking the Next Step .” Lipscomb, Stephen . 2009. “ Special Education Financing in California: A Decade After Reform ” Public Policy Institute of Calif ornia . Rivkin, Steven G., Eric A. Hanushek, and John F. Kain . 2005. “Teachers, Schools, and Academic Achievement .” Econometrica 73(2) : 417 –58. Rose, Heather, Jon Sonstelie, and Ray Reinhard . 2006. School Resources and Academic Standards: Lessons from the Schoolhouse . San Francisco: P ublic Policy Institute of California Rose, Heather, and Ria Sengupta . 2007. “ Teacher Compensation and Local Labor Market Conditions in California: Implications for School Funding .” Public Policy Institute of California . Rose , Heather, Ria Sengupta, Jon Sonstelie, and Ray Reinhard . 2008. “ Funding Formulas for California Schools: Simulations and Supporting Data .” Public Policy Institute of California . Sonstelie, Jon . 2007. “ Aligning School Finance with Academic Standards: A Weighted -Student Formula Based on a Survey of Practitioners .” Public Policy Institute of California . Weston, Margaret, Jon Sonstelie, and Heather Rose. 2009. “ California School Finance Revenue Manual .” Public Policy Institute of California . Weston, Marga ret. 20 10. “ Funding California Schools: The Revenue Limit System .” Public Policy Institute of California . Weston, Margaret . Forthcoming. “ California’s New School Funding Flexibility: One Year After Reform .” Public Policy Ins titute of California . http://www.ppic.org/main/home.asp Pathways for School Finance in California 39 About the Author s Heather Rose is an adjunct fellow at PPIC and an assistan t p rofessor in the School of Education at the University of California, Davis. She specializes in the economics of education. She has published work on school fin ance, college affirmative action policies, and the relationship between high school curriculum, test scores, and subsequent earnings. Her current research projects focus on school finance reform in California as well as school board politics and teacher sa laries. Previously, she was a research fellow at PPIC. She holds a B.A. in economics from the University of California, Berkeley, and an M.A. and Ph.D. in economics from the University of California, San Diego. Jon Sonstelie is a n adjunct fellow at PPIC a nd professor of economics at the University of California, Santa Barbara. His research interests include several areas in public finance and urban economics, including the effect of public school quality on private school enrollment, the incidence of the p roperty tax, the demand for public school spending, the economics of rationing by waiting, and the effect of transportation innovations on residential locations. He was previously a research fellow at Resources for the Future. He holds a B.A. from Washingt on State University and a Ph.D. from Northwestern University. Margaret Weston is a research associate at the Public Policy Institute of California’s Sacramento Center, where her work f ocuses on K–12 school finance. Prior to joining PPIC, she taught high school English and drama in Baltimore City Public Scho ols through Teach For America. She holds a master’s degree in teaching from Johns Hopkins University and a master of public policy degree from the University of Michigan. Acknowledgments We thank Carol Bingham and Heather Carlson from the California Department of Education for providing the data used in this report. We thank Gary Bjork, Ellen Hanak, Michael Kirst, Eric McGhee, John Mockler, Kim Rueben, Nicolas Schweizer, and Lynette Ubois for useful comm ents on previous drafts. PUBLIC POLICY INSTITUTE OF CALIFORNIA Board of Directors Walter B. Hewlett, Chair Director Center for Computer Assisted Research in the Humanities Mark Baldassare President and CEO Public Policy Institute of California Ruben Barrales President and CEO San Diego Chamber of Commerce Maria Blanco Executive Director Chief Justice Earl Warren Institute on Race, Ethnicity and Diversity University of California, Berkeley School of Law John E. Bryson Reti red Chairman and CEO Edison International Gary K. Hart Former State Senator and Secretary of Education State of California Robert M. Hertzberg Partner Mayer Brown LLP Donna Lucas Chief Executive Officer Lucas Public Affairs David Mas Masumoto Author and farmer Steven A. Merksamer Senior Partner Nielsen, Merksamer, Parrinello, Mueller & Naylor, LLP Constance L. Rice Co -Director The Advancement Project Thomas C. Sutton Retired Chairman and CEO Pacific Life Insurance Company The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research on major economic, social, and political issues. The institute’s goal is to raise public awareness and to give elected representatives and other decisionmakers a more informed basis for developing policies and programs. The institute’s research focuses on the underlying forces shaping California’s future, cutting across a wide range of public policy concerns, including economic development, education, environment and resources, governance, population, public finance, and social and health policy. PPIC is a private operating foundation. It does not take or support positions on any ball ot measures or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. PPIC was established in 1994 with an endowment from William R. Hewlett. Mark Baldassare is President and Chief Executive Officer of PPIC. Walter B. Hewlett is Chair of the Board of Directors. Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the abo ve copyright notice is included. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Copyright © 2010 Public Policy Institute of California All rights reserved. San Francisco, CA PUBLIC POLICY INSTITUTE OF CALIFORNIA 500 Washington Street, Suite 600 San Francisco, California 94111 phone: 415.291.4400 fax: 415.291.4401 www.ppic.org PPIC SACRAMENTO CENT ER Senator Office Building 1121 L Street, Suite 801 Sacramento, California 95814 phone: 916.440.1120 fax: 916.440.1121" ["post_date_gmt"]=> string(19) "2017-05-20 09:40:27" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(9) "r_1110mwr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:40:27" ["post_modified_gmt"]=> string(19) "2017-05-20 09:40:27" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(51) "http://148.62.4.17/wp-content/uploads/R_1110MWR.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }