xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam a22 u 4500
controlfield tag 008 c20019999azu 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E11-00227
Educational policy analysis archives.
n Vol. 9, no. 30 (August 17, 2001).
Tempe, Ariz. :
b Arizona State University ;
Tampa, Fla. :
University of South Florida.
c August 17, 2001
Committing to class-size reduction and finding the resources to implement it : a case study of resource reallocation / Allan Odden [and] Sarah Archibald.
Arizona State University.
University of South Florida.
t Education Policy Analysis Archives (EPAA)
xml version 1.0 encoding UTF-8 standalone no
mods:mods xmlns:mods http:www.loc.govmodsv3 xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govmodsv3mods-3-1.xsd
mods:relatedItem type host
mods:identifier issn 1068-2341mods:part
mods:detail volume mods:number 9issue 30series Year mods:caption 20012001Month August8Day 1717mods:originInfo mods:dateIssued iso8601 2001-08-17
1 of 14 Education Policy Analysis Archives Volume 9 Number 30August 17, 2001ISSN 1068-2341 A peer-reviewed scholarly journal Editor: Gene V Glass, College of Education Arizona State University Copyright 2001, the EDUCATION POLICY ANALYSIS ARCHIVES. Permission is hereby granted to copy any article if EPAA is credited and copies are not sold. Articles appearing in EPAA are abstracted in the Current Index to Journals in Education by the ERIC Clearinghouse on Assessment and Evaluation and are permanently archived in Resources in Education .Committing to Class-Size Reduction and Finding the Resources to Implement It: A Case Study of Resource Reallocation Allan Odden University of WisconsinÂ—Madison Sarah Archibald University of WisconsinÂ—MadisonAbstractThis article discusses how a medium-sized school di strict in Wisconsin was able to reallocate resources to reduce class si zes in grades K-5 without spending more money or increasing its tax r ate. Previous research on resource reallocation found that the bu lk of reallocated resources were those supporting categorical program services. This district was able to use a different strategy. As a growing district, its marginal costs of adding an extra class of students were much less than its average expenditures per pupil, which were reim bursed by the state school finance formula. As the district grew, there fore, it acquired excess revenues. Also, by implementing full-day kindergart en, the district
2 of 14acquired more excess revenues because this added (0 .5 pupil) X (Number of kindergartners) to its current enrollmen t, and the cost of educating these students was less than the amount t hey received from the state funding formula. It then used these revenues to reduce class sizes to between 15 and 20 in all Kindergarten through grade 3 classrooms and to between 15 and 22 for grades 4-5. Introduction Like most districts, the Kenosha Public School Dist rict in Southeastern Wisconsin faced a number of different educational and fiscal challenges as the new millennium approached. A medium-sized urban district in the Mi dwest, Kenosha served a population of approximately 20,000 students in the 2000-01 school year. Demographically, its student population was 77 perc ent white with 11 percent Hispanic, 9 percent African American, and 2 percent other. Hi spanics represented the largest and fastest growing minority group. About 30 percent of Kenosha's students lived in families with incomes below the poverty level, as indicated by eligibility for the federal free and reduced-price lunch program. Over the past decade, the percentage of students from low-income backgrounds continued to grow. But, rather than being evenly distributed across a ll schools, many of the district's low-income, minority and limited-English-speaking s tudents were concentrated in a small number of its schools. Several years ago, it became obvious that the growing concentration of educational challenges that usuall y accompany these demographic characteristics was making it more difficult for st udents in these schools to achieve the high standards set for them by the state and the di strict. Although a growing student population, combined with a generous state educatio n aid program, helped ensure budget stability, the district nevertheless struggled more and more to find the fiscal resources required to deploy programs successful in boosting the performance of its students, particularly its low income and minority students. In an effort to meet the achievement goals for all students over the 1990s, the district implemented a number of education reform i nitiatives. A new out-of-state superintendent who arrived in 1996 stimulated these reforms, plus, as will be shown, a number of related management and fiscal changes. At the district level, the school system shifted from a junior high to a middle schoo l approach to ease the transition into high school; opened three new charter schools to he lp foster innovation and new educational approaches; and most recently, implemen ted a full-day kindergarten program, in part to help low-income children learn to read in the early elementary grades. At the site level, many schools began to ad opt new, structured literacy programs as well as comprehensive school reform designs, suc h as Direct Instruction, Success For All, the Literacy Collaborative, Marva Collins and Accelerated Schools. Although these reforms helped the district and its schools make progress toward their student achievement goals, teachers as well a s district and school leaders wanted to see even greater progress. One additional reform th at teachers and administrators alike believed would provide a greater boost in student a chievement was class-size reduction. Indeed, class size reduction, particularly in the early grades, was an education reform sweeping the country during this time. Odden and Picus (2000) found that 19 states and scores of districts enacted various vers ions of this education policy in the
3 of 141990s, believing that despite the high cost, it was a policy that research showed could dramatically raise student performance (Grissmer, 1 999). Further, Wisconsin had adopted a targeted class si ze reduction program in 1995 as part of a statewide approach to improve student performance, particularly the performance of its low-income and minority students Called the Student Achievement Guarantee in Education or SAGE program, it was init ially limited to schools that had 50 percent or more of its students eligible for free a nd reduced price lunch, i.e., from a family with a poverty level income. In the 1996-97 school year, this included 30 elementary schools in 21 districts, including one school in Kenosha. For the 2000-01 school year, the program expanded its eligibility a nd had a dramatic increase in funding; in 2001, it helped hundreds of schools afford small er classes around Wisconsin. Although there was strong support for a district-w ide class-sizereduction program, there was still the issue of how to financ e it. The expansion of the SAGE program provided one source. But two additional fac tors helped the district frame a strategy for how they could come up with the money to fund such a high-cost education reform. First, a 1994 legislative change in the fed eral Title I program allowed schools with at least 50 percent of its students in poverty to use their Title I funds for school-wide programs, rather than programs targeted to just its low-income students. The school-wide programs selected had to be researc h-based, and class-size reduction was a strategy that qualified. Second, prior to the opening of the 1998-99 school year, district administrators heard a presentation on res ource reallocation in education; one of the major ideas discussed was school use of Title I funds for schoolwide programs, including both class-size reduction and comprehensi ve school design models (see, for example, Odden & Archibald, 2000). The notion of resource reallocation for Â“whole sch ool reformÂ” triggered thoughts about additional programmatic and fiscal changes am ong many district and school leaders. The general notion was that the district c ould redesign schools from Â“the ground upÂ” and use all of a site's resources to finance a new, researchbased educational strategy, which could include small classes in at l east Kindergarten through grade three. Further, a few elementary schools had access to oth er funds that could help reduce class sizes. In addition to Title I and the possibility o f SAGE money, some elementary schools qualified for a small state program, called P5, tha t provided extra resources to schools with high poverty concentrations (schools could eit her receive P5 or SAGE money, but not both). In addition, the district knew that at t he federal level, Title VI of the reauthorized Elementary and Secondary Education Act (ESEA) would provide an additional pot of money to help fund class-size red uction. A combination of the above ideas convinced the sch ool board in July 1998 that it could begin immediately to reduce class size in a s elected number of its high-poverty elementary schools just by reallocating Title I (ma inly to school-wide programs) and P-5, and using SAGE and Title VI funds where possib le. They did so, and in the first year they saw a positive impact on student academic performance as measured by state and district tests. In light of the positive results at these elementary schools, as well as even firmer belief in research supporting smaller classe s (Finn & Achilles, 1999; Molnar, Smith, Zahorik, Palmer, Halbach & Ehrle, 1999, see also http://www.uwm.edu/Dept/CERAI), the district decide d to find the money to reduce class sizes for Kindergarten through grade three in all elementary schools the following school year. Whether this was the reform that would best help all students is beyond the scope of this study. Instead, this case study tells the story of how the district was able to fund this ambitious class-size reduction initiative
4 of 14Finding the Money Previous research on resource reallocation (Achill es, 1999; Odden & Archibald, 2000) showed that categorical program dollarsÂ—Title I, special education, programs for limitedEnglish proficient students, etc.Â—comprise d the bulk of funds that were reallocated, whether the goal was classsize reduc tion or implementing a specific whole school design, such as Roots and Wings/Success for All. Kenosha initially used this strategy as well, as it was primarily federal Title I, Title VI, and state SAGE and P5 money that allowed the highest poverty schools to b egin to reduce class sizes. Studies have shown that reduced class sizes are especially beneficial for low-income and minority students (Finn and Achilles, 1999). Believing these benefits would extend to all stude nts, the district decided to reduce class sizes in all elementary schools, inclu ding those that were not eligible for categorical program funds, but to do so it needed a new, larger source of funding. This challenge sent the district back to analyze the wor kings of its overall budget as well as the state of Wisconsin's school finance equalizatio n formula. These analyses also prompted the district to consider implementing a fu ll-day kindergarten program. Although they had been considering this step for a long time, this decision, combined with the reality of rising student enrollments, hel ped the district generate the resources that would enable it to fund a larger class-size re duction strategy. Funding for the larger class-size reduction initia tive came from two sources: rising enrollment and the implementation of full-da y kindergarten. In the next few paragraphs, these funding sources will be addressed in turn. First, Kenosha was able to generate resources from rising enrollments in the f ollowing way: In Wisconsin, districts receive aid on the basis of a three tiered, guarant eed tax base, school finance formula. The state aid formula functions so that in any one year, for each additional student Kenosha gains (or loses), the district receives (or loses) approximately $7000. Since the district has experienced rising enrollments in the past few years, its budget has constantly grown. And because each new student cost s less than $7000 to educate, the marginal cost of educating each new student is less than the average per pupil expenditure of the district. To illustrate this poi nt, assume the district enrolls 25 new students. The district's budget would rise by $175, 000 (25 x $7000). But the district would have to hire only one additional teacher for a new class of 25 kids. This would require one teacher at an average cost of $50,000, plus the cost of 20 percent of another teacher for the first teacher's planning and prepar ation time, and some additional costs for materials, supplies and operations and maintena nce. Assume these costs totaled $75,000. That would leave the district with excess revenues of $100,000, which would be sufficient for the district to hire close to two additional teachers, at an average cost of $50,000 each. The marginal costs for the additional 25 students were only $3000 per student ($75,000 divided by 25 students), while the average additional dollars were $7000 a child, for a possible difference between av erage and marginal costs of about $4000 a child. To be sure, the above is a simplified analysis. Fo r the numbers to work exactly, all students would need to be in the same grade lev el and in the same school. But since the district enrollment grew by approximately 500 s tudents between 1999-2000 and 2000-01, a rise of 25 per grade was possible, thoug h maybe not in each school. For these reasons, it may be true that the difference between average and marginal costs for each additional student is not precisely $4000. Even so, because of the high level of the school finance equalization formula and the phenome non of rising enrollments, the
5 of 14district was in the enviable position of generating substantial excess revenues for each new student in the district. This situation was fur ther strengthened by the fact that the district had some excess classroom space, so it did not need to build new classrooms because of enrollment growth or because of class si ze reductions. How the district found the space for the extra classes will be discussed l ater in this case study. Another complication for the above example is that Wisconsin's education aid formula does not use the current student count in d etermining each district's aid, but uses the average of the number of students from the past three years. Initially, this was implemented to limit state aid losses for declining enrollment districts, but for growing enrollment districts it limits the increase in reve nues as well. Thus, for Kenosha to enjoy the full measure of the excess of average over marg inal cost increases, it needed a full three years. If this were the only source of new fu nding, the district would have had to phase in its class size reduction strategy over sev eral years. However, as previously mentioned, the district fou nd another way to raise revenue for the small class-size policy that furthe r aided all its students, but particularly its students from low-income backgrounds, which was the implementation of full-day kindergarten. In the mid-1990s, the district provid ed only a half-day kindergarten program. But it knew that full day kindergarten was desired by many families and also was a research-based strategy to help students from low-income backgrounds learn reading and writing in the first three years of ele mentary school (Slavin and Madden, 1996). Therefore, the district decided to change fr om a half-day to a full-day kindergarten program. This move generated more fund s for class-size reduction for the same reason identified with new enrollments: having kindergartners for an extra half day gave the district additional revenues because of ex cess average over marginal costs. Because the state allowed the district to count the child as a 1.0 student for a full-day program as compared to 0.5 for a half-day program, the shift to a full day program essentially allowed the district to Â“increaseÂ” enro llments on its own. For every two students who shifted from a halfto a full-day kin dergarten program, the district was able to add 1.0 (2 x 0.5) students to its pupil cou nt. It received the full $7000 for this Â“extraÂ” student, but again its marginal costs were much lower than this $7000 figure. Another example helps illustrate this point. When a school has 20 students for morning kindergarten and 20 students for afternoon kindergarten, the school receives 40 x 0.5 x $7000, or $140,000. Assuming that the same teacher is working in the morning and the afternoon, the school will spend approximat ely $50,000 of that amount on the teacher's salary and benefits, plus planning and pr eparation time, materials, etc. which might total $65,000 70,000, leaving an upper-boun d estimate of $70,000 ($140,000 70,000) in excess revenues. When those 40 students change to spending a full d ay at school, the school receives $7000 for each student, or a total of $280 ,000. The school will now need to hire two teachers, who will each have a class of 20 stud ents, which will cost about $140,000 ($100,000 for salary and benefits plus additional m oney for planning time and materials). This leaves $140,000 ($280,000-140,000) for other purposes, compared to the $70,000 that was left when the 40 students only attended for half a day. Because Kenosha had most of the necessary classroom space, the district primarily needed only to build in the cost of operating those rooms all d ay to cover the expenses of essentially doubling the number of kindergarten classes by movi ng to full-day kindergarten. As is true with the growing number of students enrolled, the marginal cost of educating kindergartners for a full day was less than the ave rage amount that they received via the state equalization program, resulting in a net fisc al gain for the district. Again, because of the three-year average pupil count, these excess revenue numbers took three years to
6 of 14be fully realized. So the district was able to take advantage of two rising enrollment phenomena to produce the revenues to help fund a district-wide c lass-size reduction programÂ—first, excess revenues from naturally occurring enrollment increases, and second, excess revenues from shifting from a half-day to a full-da y kindergarten program. When combined with the initial categorical dollars that were used to reduce class sizes first in the highest poverty-concentration schools, Kenosha was able to fund dramatic class size reductions without increasing taxes, eliminating ot her programs, or increasing its average expenditure per pupil.Implementing the Class Size Reduction Policy The two Â“newÂ” revenue sources discussed in the pre vious section made it possible for the district to help fund reduced clas s sizes at all of its elementary schools, but implementing this initiative was complicated. F irst, because the excess revenues from shifting to a full-day kindergarten program wo uld be phased-in over a three-year period, the superintendent initially advised the di strict to slowly implement the class-size reduction policy, making sure that the resources we re there before putting the plan into action. However, the board voted to move ahead imme diately because it felt that when their strategy for finding the money to do so becam e public information, there could be many other claims for its useÂ—including other progr ams as well as teacher salary increases. Although class-size reduction was very p opular with teachers, the policy would be to reduce class size for just kindergarten through grade 5, so teachers of other grades could very well have had different ideas abo ut how to spend those dollars. Though the superintendent initially was skeptical a bout moving so fast, he eventually came around because he saw the wisdom of quickly im plementing a popular policyÂ—class size reductionÂ—and avoiding a lengthy debate over alternative uses of the Â“foundÂ” dollars. In order to avoid the potential delay, district an d board leaders acted quickly. In July of 1999, the Kenosha school board accepted the district proposal to fund 22 new teaching positions to reduce class sizes for the fo llowing school year. This meant that the district had just over a month to hire all of these teachers plus the normal new hires for the year. It managed to accomplish this task, hirin g 37 new teachers (22 for class-size reduction) in the one month before school began in September. This was one of the difficulties with trying to implement this initiati ve so quickly. The 22 new teachers were awarded centrallyÂ—schools had to request these extra teachers from the office of the superintendent and explain how they would be used to reduce class sizes and what space they would use. B ecause the district had staffed each elementary school on a 24 students to 1 teacher bas is, each school was required to use the new teachers for class sizes below that level. Some schools had the extra space, especially the high-poverty elementary schools that had declining enrollment. Other schools had to give up their art, music, or multipu rpose rooms, which some schools were reluctant to do. Some sacrifices had to be made to implement this policy, but the district had made the decision to reduce class size, and to do so they needed classroom space. These initial new teacher positions helped bring do wn class sizes somewhat, but district leaders had begun to make long-term plans for forma lly reducing class sizes even further, to specific target levels. Over the course of the 1999-2000 school year, dist rict leaders had more time to plan the further reduction of elementary class size s in the Kenosha School District. This time, they decided to reduce the staffing ratios fo r grades K-5 rather than centrally award
7 of 14Â“extraÂ” teachers to the sites, which had produced a variety of different but lower class sizes. This district-wide policy change made the cl ass-size reduction policy explicit for all elementary schools. As previously mentioned, the ratio that had been i n place was one teacher for every 24 students in grades K-5. In the 2000-2001 s chool year, with the new class-size reduction policy in place, the ratios became 20 to 1 for K-3 and 23 to 1 for grades 4-5. These ratios were set in part according to how many extra classrooms were available at the schools; although district leaders would have l iked to have even lower class sizes, this was the level at which they felt schools had t he appropriate extra space to accommodate the additional classrooms. The class si ze maximums also changed. Previously, the maximums were 27 for kindergarten t hrough grade 3, and 29 for grades 4-6; under the new allocation formula, maximums wer e set at 22 for kindergarten through grade 3, and 25 for grades 4-5. Further, in the past, when a class exceeded its ma ximum, an instructional aide was placed in the classroom to provide assistance. More specifically, if the class size was above 26 but not above 30, a half-time instruct ional aide was provided; a full-time aide was provided if the actual class size was over 30. But under the new, 2000-2001 class-sizereduction policy and class-size maximum s, most of the resources for instructional aides were reallocated for the additi onal classroom teachers that were necessary to keep class sizes lower. The Assistant Superintendent of Instruction stressed that the district had adopted an overall value that whenever possible, resources would be used to employ fully trained and licensed teache rs rather than instructional aides. This decision was bolstered by the fact that the same re search that documented the effectiveness of small classes also showed that add ing an instructional aide to a larger class provided no effect on student performance (Ac hilles, 1999). Indeed, many of the elementary schools that received Title I funds elim inated instructional aides and used their Title I funds to hire teachers instead. In Ke nosha, it takes the resources for about two full-time aides for a school to hire a licensed teacher. For the low-performing schools in Kenosha, class s ize targets were set even lower than at the other elementary schools. Kenosha defin ed low-performing by student performance on the Wisconsin Knowledge and Concepts Examination, a statewide testing system using the Terra Nova commercial test All schools with an average WSAS score below the state average were categorized as Â“ lowperforming.Â” Even before the district changed its staffing allocation formula, t hese schools, because they had the highest poverty concentrations, had access to all o f the previously mentioned categorical program dollars that could be used to help fund cla ss-size reduction: Title I, Title VI, P-5 and SAGE. The district class-size goal for low-performing, h ighpoverty schools was set at 15, with a maximum of 18. This maximum was set beca use schools receiving Title VI funds, the federal class-size reduction program, ha d to reduce class sizes to 18 or lower. Thus, the district class-size goal for the low perf orming schools is lower than the federal requirement. The extra funding from SAGE, Title I, Title VI and P-5 that these schools received was used to enable the district to reduce class sizes to 15 in many of its low-performing schools. However, because there are more low-performing elementary schools in the Kenosha School District than resourc es to reduce class sizes to those target levels, the district focused on reducing cla ss size to these low levels at the lowest performing schools first. As of late 2000, the district had made significant progress in reaching its lower class size goals. Of the 24 elementary schools in t he district, one had class sizes of 15 or lower, eight more had class sizes of 18 or lower, 1 1 had class sizes of 20 or lower, and
8 of 14four had class sizes of 22 or lower. Thus, all elem entary school class sizes were below the maximum of 22 for grades K-3, and well below th e maximum of 25 for grades 4-5. Further, while only one highpoverty school had an average class size of 15 or lower, all elementary schools at which the majority of student s are low-income had class sizes of 20 or lower. For many elementary schools, the combination of th e new, lower staffing ratio and the reallocation of categorical funds has meant a dramatic reduction in class size and a feeling of optimism about the subsequent effect o n student achievement. For example, at Wilson Elementary School, where 86 percent of th e students qualify for free or reduced-price lunch, class sizes have been reduced from an average of 21.9 in 1998-99 to an average of 16.4 in 2000-01. In order to reduc e class sizes that far, the school needed a total of nine more classroom teachers in 2 000-01 than it had in 1998-99. The reduction in class size was possible because of the change in the staffing formula and the allocation of categorical funds for class-size redu ction. Enrollment growth does not explain any of the new teachers for 2000-2001; Wils on had only 10 more students in that year than in 1998-99. The school received three add itional teachers for 2000-01 because of the new staffing allocation formula. The other s ix teachers were paid for with categorical funds: one with other district funds fo r classsize reduction, two with federal Title VI funds, 2.5 with SAGE funds, and 0.5 from T itle I. The district funds for class-size reduction were the result of a board dec ision that 30 new teachers would be hired for the 2000-2001 school year; Wilson got one of those teachers. It is also interesting to note that this had formerly been a P -5 school, but they traded in that status in order to qualify for SAGE since they were able t o hire more teachers with those funds. By concentrating many of its resources on class-si ze reduction, Wilson was able to lower class sizes to an average of 16 students. This is noteworthy because it is within the range of 13-17 identified by the Tennessee STAR study as being especially effective at raising student achievement levels and producing other favorable outcomes (Finn & Achilles, 1999; Grissmer, 1999). Class sizes have also been significantly reduced a t Vernon Elementary School, where 36 percent of students qualify for the free a nd reduced-price lunch program. Although this is a significantly lower percentage t han at Wilson, Vernon still receives substantial categorical funding. As was true at Wil son, it is this funding in combination with the district's lower staffing allocation formu la that has allowed Vernon to reduce class sizes from an average of 27 in 1998-99 to an average of 18 in 2000-01. Enrollment has stayed virtually the same over this time period ; 468 students attended Vernon in 2000-01 as compared to 472 in 1998-99. Therefore, n one of the 10 additional teachers on staff at Vernon in 2000-01 can be explained by e nrollment growth. Instead, four of the teachers were a result of the change in the sta ffing ratio from 24 to 1 to 20 to 1; one was paid for using the additional district funds fo r classsize reduction; one was paid for with federal Title VI funds; three were paid for wi th Title I; and one with P-5 funds. By concentrating its funding sources on class-size red uction, Vernon was able to reduce class sizes by a remarkable nine students per class room.Initiatives Supporting Class Size Reduction In addition to getting class sizes down to desired levels, part of the implementation process is to provide teachers with professional development that will help them teach their small classes more effectivel y. Leaders in Kenosha knew that research has shown that teachers can help produce a dditional achievement gains when professional development in small-class instruction is provided (McRobbie, Finn &
9 of 14Harman, 1998; Bearl, 1998; U.S. Department of Educa tion, 1998). The Assistant Superintendent of Instruction stressed the importan ce of this aspect of implementation, and has continued to seek School Board support for substantial investments in professional development in order to help realize t he potential of better instruction in the smaller classes. Title VI, the federal funding sour ce for class-size reduction, allowed districts to use 15 percent of its funding for prof essional development in 1999-2000; although district administrators thought this would be a good idea, the board voted to use all the funds to pay for teachers in the first year. However, for the 2000-2001 school year, districts were allowed to use up to 25 percen t of their Title VI funds for professional development, and this time the board a greed, in part because the funding had increased and they were able to hire almost as many teachers and have money leftover for professional development. The district was able to allocate $137,000, or approximately 25 percent of the federal money they receive for comprehensive school reform, toward professional development programs de signed to improve instructional practices in small classrooms. This money made it p ossible for 700 of the district's 1600 teachers to participate in such professional develo pment activities as Everyday Mathematics training. Further, the district has continued to encourage a ll schools to adopt some type of Â“whole schoolÂ” educational strategy to accompany bo th the above professional development and small class sizes. For example, for the 2000-2001 school year, the district provided $299,000 in grants to schools wit hout P5 or Title I funds to help them afford various school reforms, including comprehens ive school reform models. This district's commitment to raising student achievemen t scores is apparent through the creative use of all of their funds to make changes that they believe will boost student achievement.Monitoring Results Although the superintendent and other district lea ders have made class-size reduction a priority in Kenosha elementary schools, they recognize that it is just one reform to be used in conjunction with other reforms for a common purpose: to boost student achievement. For that reason, student achie vement scores are carefully measured and reported in order to track the success of these reforms. In 2000-01, students at Kenosha elementary schools took the Iowa Test of Ba sic Skills (ITBS) in grades 2, 3, and 5; the Wisconsin 3rd grade reading test, and th e Wisconsin Knowledge and Concepts Examination (WKCE) in 4th grade. The district placed the most e mphasis on the 3rd grade reading test and WKCE. Each school ha d benchmarks that they were working toward, which were set in terms of the perc ent of students at proficiency level and the percent advanced. Awards were given to scho ols that increased these percentages by five percent annually. Student achievement growth has been substantial at many Kenosha Elementary schools, including Wilson, one of the schools used as an example in the last section. In addition to reduced class sizes, Wilson has used th e principles of the Marva Collins school design for the past few years, and has more recently adopted Direct Instruction for reading, and Core Knowledge as a curriculum gui de. In the 1997-98 school year, only 17 percent of Wilson students were considered proficient on the third grade reading test. In 1999-2000, 51 percent were at or above pro ficiencyÂ—a dramatic improvement of 34 percentage points. Columbus Elementary School, a school with 58 perce nt of students who qualified for free or reduced-price lunch in 2000-01, has als o made great strides. Funded by their
10 of 14categorical dollars, Columbus has had reduced class sizes for three years. The school chose class-size reduction rather than a comprehens ive school reform model, and is now listed in the high achievement category for the WKC E in Math. The state average score for percent proficient in math is 52, and 68 percen t of Columbus students scored at or above proficiency (Barth, Haycock, Jackson, Mora, R uiz, Robinson, & Wilkins, 1999). This is the best indication so far that the policy to reduce class sizes in Kenosha will boost student achievement in all schools, as it has at Columbus Elementary. In addition to Columbus and Wilson, student achievement has ris en at many other Kenosha schools, in part due to initiatives like class-size reductio n and comprehensive school reform models. Conclusion Leaders in the Kenosha School District in Wisconsi n managed to significantly reduce class size in the majority of their elementa ry schoolsÂ—by creative resource reallocation and deployment of all the revenues mad e possible by student demographic characteristics and the state's school finance syst em. In the 2000-01 school year, more than one-third of Kenosha's elementary schools had class sizes of 18 or lower, and all 24 schools had class sizes at or below 22. Although in dividual schools in this district had begun to reduce class size school-by-school by real locating Title I and P-5 funds and using Title VI and SAGE, the district-wide change t o a lower elementary school staffing allocation formula made sure that class sizes were reduced at every elementary school, and to lower levels than the schools could have aff orded with just categorical dollar reallocation. The major revenue source for this expensive policy was excess revenues derived from the combination of growing enrollment and the shift from a half-day to a full-day kindergarten. For every new student, the marginal c ost of educating that student was approximately $3000 but the district received an ex tra $7000 via the state school finance formula, or an excess of average over marginal cost s of $4000 per child. The total combined district enrollment growth from these two phenomenaÂ—natural growth and kindergarten expansionÂ—was about 500 students a yea r. This produced excess revenues of nearly $2,000,000 (500 students times $4000/stud ent), which was sufficient to hire 40 additional teachers at an individual cost of $50,00 0 in salary and benefits. This quite ingenious way to fund smaller class sizes, combined with additional dollars from selected categorical programsÂ—federal Title I and T itle VI, and state SAGE and P5Â—allowed for even lower classes in the highest po verty, lowest performing schools, reaching the level of 18 or lower in nine of the di strict's 24 elementary schools. District leaders hope that the positive results from Columbu s Elementary School, which has had lower class sizes in place for three years, will be replicated in elementary schools district-wide. This case shows how important it is for district l eaders who want to make changes using reallocated dollars to have full know ledge of the district budget and how that budget is derived. In Kenosha, district leader s decided to reallocate categorical dollars to class-size reduction, but they needed an additional source of funding to reduce class sizes to target levels. They were able to fin d that additional funding source because they understood the principle that Â“newÂ” students, whether from natural enrollment growth or the shift from half-day to full-day kinde rgarten, could produce Â“newÂ” dollars because of the excess of average over marginal cost s. These changes enabled the district to provide full-day kindergarten and reduce class s izes in all elementary schools, initiatives that research shows are particularly po werful in helping students from
11 of 14low-income backgrounds learn to read and do mathema tics in the early elementary grades (McRobbie, Finn, & Harman, 1998; Slavin and Madden, 1996; Slavin, Karweit and Madden, 1989).NoteThis article was prepared for the Consortium for Po licy Research in Education, Wisconsin Center for Education Research, University of Wisconsin-Madison. The research reported in this paper was supported by a grant from the U.S. Department of Education, Office of Educational Research and Impro vement, National Institute on Educational Governance, Finance, Policy-Making and Management, to the Consortium for Policy Research in Education (CPRE) and the Wis consin Center for Education Research, School of Education, University of Wiscon sin-Madison (Grant No. OERI-R3086A60003). The opinions expressed are those of the authors and do not necessarily reflect the view of the National Instit ute on Educational Governance, Finance, Policy-Making and Management, Office of Ed ucational Research and Improvement, U.S. Department of Education, the inst itutional partners of CPRE, or the Wisconsin Center for Education Research.ReferencesAchilles, C. 1999. Let's put kids first finally: Getting class size ri ght Thousand Oaks, CA: Corwin Press.Barth, P., Haycock, K., Jackson, H., Mora, K., Ruiz P., Robinson, S., & Wilkins, A. (1999). Dispelling the myth: High-poverty schools exceeding expectations Washington, D.C.: Education Trust.J. & Achilles, C. (1999). Tennessee's class size st udy: Findings, implications, misconceptions. Educational Evaluation and Policy Analysis 21(2), 97-109. Grissmer, D. (Ed.). (1999). Class Size: Issues and new findings [Special Issue]. Educational Evaluation and Policy Analysis 21(2). McRobbie, J., Finn, J., & Harman, P. (1998). Â“Class -size reduction: Lessons learned from experience.Â” Policy Brief, WestEd, August 1998 Molnar, A., Smith, P., Zahorik, J., Palmer, A., Hal bach, A., & Ehrle, K. (1999). Â“Evaluating the SAGE Program: A Pilot Program in Ta rgeted Pupil-Teacher Reduction in Wisconsin.Â” Available online: http://www.asu.edu /educ/epsl/. Odden, A., & Archibald, S. (2000). Reallocating resources: How to boost student achievement without asking for more. Thousand Oaks, CA: Corwin Press. Odden, A., & Picus, L. O. (2000). School finance: A policy perspective (2nd ed.). New York: McGraw-Hill.Slavin, R., Karweit, N. & Madden, N. (1989). Effective programs for students at risk Newton, MA: Allyn & Bacon.
12 of 14 Slavin, R., Madden, N., Dolan, L., & Wasik, B. (199 6). Every child, every school: Success for all. Newbury Park, CA: Corwin.About the AuthorsAllan OddenCPRE/WCER 1025 West Johnson Street, #653Madison WI 53706-1796Voice: 608 263 4260 Fax: 608 263 6448 Email: email@example.comDr. Odden is a Professor of Educational Administrat ion at the University of Wisconsin-Madison, and Co-Director of the Consortiu m for Policy Research in Education (CPRE). His areas of concentration are ed ucation policy, teacher compensation and school finance. Sarah Archibald is a Researcher in the CPRE offices of the Wiscons in Center for Educational Research; her areas of specialty are ed ucation policy, school finance and effective resource use.Copyright 2001 by the Education Policy Analysis ArchivesThe World Wide Web address for the Education Policy Analysis Archives is epaa.asu.edu General questions about appropriateness of topics o r particular articles may be addressed to the Editor, Gene V Glass, firstname.lastname@example.org or reach him at College of Education, Arizona State University, Tempe, AZ 8 5287-0211. (602-965-9644). The Commentary Editor is Casey D. C obb: email@example.com .EPAA Editorial Board Michael W. Apple University of Wisconsin Greg Camilli Rutgers University John Covaleskie Northern Michigan University Alan Davis University of Colorado, Denver Sherman Dorn University of South Florida Mark E. Fetler California Commission on Teacher Credentialing Richard Garlikov firstname.lastname@example.org Thomas F. Green Syracuse University Alison I. Griffith York University Arlen Gullickson Western Michigan University Ernest R. House University of Colorado Aimee Howley Ohio University
13 of 14 Craig B. Howley Appalachia Educational Laboratory William Hunter University of Calgary Daniel Kalls Ume University Benjamin Levin University of Manitoba Thomas Mauhs-Pugh Green Mountain College Dewayne Matthews Western Interstate Commission for HigherEducation William McInerney Purdue University Mary McKeown-Moak MGT of America (Austin, TX) Les McLean University of Toronto Susan Bobbitt Nolen University of Washington Anne L. Pemberton email@example.com Hugh G. Petrie SUNY Buffalo Richard C. Richardson New York University Anthony G. Rud Jr. Purdue University Dennis Sayers Ann Leavenworth Centerfor Accelerated Learning Jay D. Scribner University of Texas at Austin Michael Scriven firstname.lastname@example.org Robert E. Stake University of IllinoisÂ—UC Robert Stonehill U.S. Department of Education David D. Williams Brigham Young UniversityEPAA Spanish Language Editorial BoardAssociate Editor for Spanish Language Roberto Rodrguez Gmez Universidad Nacional Autnoma de Mxico email@example.com Adrin Acosta (Mxico) Universidad de Guadalajaraadrianacosta@compuserve.com J. Flix Angulo Rasco (Spain) Universidad de Cdizfelix.firstname.lastname@example.org Teresa Bracho (Mxico) Centro de Investigacin y DocenciaEconmica-CIDEbracho dis1.cide.mx Alejandro Canales (Mxico) Universidad Nacional Autnoma deMxicocanalesa@servidor.unam.mx Ursula Casanova (U.S.A.) Arizona State Universitycasanova@asu.edu Jos Contreras Domingo Universitat de Barcelona Jose.Contreras@doe.d5.ub.es Erwin Epstein (U.S.A.) Loyola University of ChicagoEepstein@luc.edu Josu Gonzlez (U.S.A.) Arizona State Universityjosue@asu.edu Rollin Kent (Mxico)Departamento de InvestigacinEducativa-DIE/CINVESTAV Mara Beatriz Luce (Brazil)Universidad Federal de Rio Grande do Sul-UFRGS
14 of 14 email@example.com firstname.lastname@example.org email@example.comJavier Mendoza Rojas (Mxico)Universidad Nacional Autnoma deMxicojaviermr@servidor.unam.mxMarcela Mollis (Argentina)Universidad de Buenos Airesmmollis@filo.uba.ar Humberto Muoz Garca (Mxico) Universidad Nacional Autnoma deMxicohumberto@servidor.unam.mxAngel Ignacio Prez Gmez (Spain)Universidad de Mlagaaiperez@uma.es Daniel Schugurensky (Argentina-Canad)OISE/UT, Canadadschugurensky@oise.utoronto.ca Simon Schwartzman (Brazil)Fundao Instituto Brasileiro e Geografiae Estatstica firstname.lastname@example.org Jurjo Torres Santom (Spain)Universidad de A Coruajurjo@udc.es Carlos Alberto Torres (U.S.A.)University of California, Los Angelestorres@gseisucla.edu