Rural Education and the Rural Education Achievement Program (REAP): Overview and Policy Issues

Order Code RL33804 Rural Education and the Rural Education Achievement Program (REAP): Overview and Policy Issues Updated March 21, 2008 Jeffrey J. Kuenzi Specialist in Social Legislation Domestic Social Policy Division Rural Education and the Rural Education Achievement Program (REAP): Overview and Policy Issues Summary Advocates for rural local educational agencies (LEAs) maintain that these school districts have many advantages — for example, that rural schools are more likely to be closely connected to the community, parents, and students than is the case in larger, urban and suburban LEAs. At the same time, rural schools face a variety of challenges, both in general (such as lack of access to cultural and educational resources) and more specifically regarding current federal requirements related to the No Child Left Behind Act, or NCLBA, (such as special problems meeting the requirement for “highly qualified” teachers under NCLBA). There are many ways to define a rural school. The definition of a rural school can be based on location (e.g., distance from metropolitan areas), by size, or by population density. Targeting can also be based on how poor a rural school district is. Depending on the definition used, the number of rural LEAs can vary from 11% to more than 60% of all LEAs, and can be said to serve as few as 2% of all public school students to as many as one-quarter of all students. Rural school districts differ in important ways from their urban and suburban counterparts. Rural districts tend to have fewer minority students: while large and mid-size cities often have majority minority student populations, rural school districts tend to be predominantly white. Rural districts tend to have smaller schools. For example, high schools in rural areas have an average enrollment of about 200 students, while urban and suburban high schools average between 800 and 1,200 students. Similarly, rural schools have fewer teachers (for example, 20 teachers for the average rural high school and nearly 60 teachers for the average urban high school). Finally, rural districts are less likely to have special schools and programs. For example, nearly 10% of urban schools are charter schools, while less than 2% of rural schools are charters. One way that Congress has aimed to aid rural schools is through the Rural Education Achievement Program (REAP), which provides funds to small, rural LEAs (an enrollment of less than 600) and relatively poor rural LEAs (a child poverty rate of at least 20%). Approximately 4,000 LEAs receive funds under the Small, Rural Schools Achievement program (SRSA), and an additional 1,200 LEAs receive Rural Low-Income School (RLIS) grants. The REAP program is part of the Elementary and Secondary Education Act (ESEA), which the 110th Congress is expected to consider for reauthorization. One possible policy question involves a potential change in how rural LEAs are identified under the program. The statute specifies the use of locale codes to determine which LEAs are located in rural areas. The U.S. Department of Education (ED) has proposed changes to the determination of locale codes. If adopted, the new locale code system could eliminate some LEAs from eligibility for REAP funds (perhaps as many as 400 from the SRSA program) and add newly eligible LEAs (perhaps 35). This report will not be updated. Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 What Is a Rural School District? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Metro-Centric Locale Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Urban-Centric Locale Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Population Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Poverty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Characteristics of Rural School Districts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Ethnicity of Rural School Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Size Characteristics of Rural LEAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Special Programs in Rural LEAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 The Rural Education Achievement Program (REAP) . . . . . . . . . . . . . . . . . . . . . 17 REAP Eligibility, Grant Determination, and Use of Funds . . . . . . . . . . . . . 18 Eligibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Grant Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Use of Funds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Distribution of Certain ESEA Grants to Rural LEAs . . . . . . . . . . . . . . . . . . . . . 24 Does REAP Compensate Small, Rural LEAs for Small Formula Grant Amounts? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Are Rural LEAs at a Disadvantage for Obtaining Competitive Grants? . . . 27 Urban-Centric Locale Codes: Possible Impacts on Eligibility . . . . . . . . . . . . . . . 29 Possible Policy Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Shift in Locale Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Allocating Excess Funds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Increase Benefits to Small, Poor LEAs . . . . . . . . . . . . . . . . . . . . . . . . 34 Adjust SRSA Formula to Reduce Anomalies . . . . . . . . . . . . . . . . . . . 35 Appendix: Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 CCD Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 ED Budget Service Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 ED Grants Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 List of Figures Figure 1. Poverty Rate by Metro-Centric Locale Code . . . . . . . . . . . . . . . . . . . . . 9 Figure 2. Enrollment in LEAs in Cities, Towns, and Rural Areas . . . . . . . . . . . 10 Figure 3. Ethnic Make-Up of Urban, Suburban, Town, and Rural LEAs . . . . . . 11 Figure 4. Where Children of Various Ethnic Groups Go to School . . . . . . . . . . 12 Figure 5. Average Size of Schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 6. Average Class Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Figure 7. Average Full-Time-Equivalent (FTE) Teachers per School . . . . . . . . 14 Figure 8. Percentage of Schools That Are Charters . . . . . . . . . . . . . . . . . . . . . . 15 Figure 9. Percentage of Title I Schools That Are Schoolwide Programs . . . . . . 16 Figure 10. Percentage of Schools That Are Magnets . . . . . . . . . . . . . . . . . . . . . 17 List of Tables Table 1. Various Definitions of Rural School Districts . . . . . . . . . . . . . . . . . . . . . 5 Table 2. Comparisons of Metro-Centric and Urban-Centric Locale Codes . . . . . 7 Table 3. Appropriations for REAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Table 4. Estimating Numbers of LEAs Eligible for REAP Programs . . . . . . . . 19 Table 5. REAP State Amounts for FY2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Table 6. Comparison of Grants for SRSA-Eligible LEAs and Other LEAs . . . . 26 Table 7. Data on Competitive Grant Recipients (ED Office of Elementary and Secondary Education, or OESE, FY2003) . . . . . . . . . . . . . . . . . . . . . . 28 Table 8. Comparison of Schools Classification by Metro-Centric and by Urban-Centric Locale Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Table 9. Estimates of Amounts and Number of Grantees Under the SRSA Formula Based on Metro-Centric and Urban-Centric Locale Codes . . . . . . 31 Rural Education and the Rural Education Achievement Program (REAP): Overview and Policy Issues Introduction1 Advocates for rural schools argue that these schools have many advantages. For example, the Wisconsin Rural Policy Network Forum listed a number of advantages, including the following: ! ! ! ! Rural teachers are key members of the community and tend to know the students and their families. Rural schools have a flatter organizational structure with fewer layers than non-rural school systems, and are able to adjust or adapt relatively quickly to change. Students in rural schools tend to support one another with activities, such as peer mentoring. The schools within rural communities are very visible and strongly connected with the community.2 Rural school advocates also admit that rural schools face challenges. These include the following: ! ! ! Rural schools face extreme fiscal limitations, which result in various problems, such as limited range of curricular options and a lack of advanced placement course offerings as well as difficulties providing competitive salaries to attract and retain highly qualified teachers. Rural schools tend to have declining enrollment. Many rural schools are in sparsely populated areas, which results in several problems, such as high transportation costs and limited access to cultural and educational resources.3 In addition to these general challenges, rural local educational agencies (LEAs) may face particular problems meeting requirements of the Elementary and Secondary 1 This is an updated version of a report originally written by Richard N. Apling. 2 Wisconsin Department of Public Instruction. Summary of the Official Proceedings Wisconsin Rural Policy Network Forum, January 2004. Downloaded from [http://www. dpi.state.wi.us/rural/pdf/ri_sum.pdf] on December 5, 2006, pp. 2-3. 3 Ibid., pp. 3-5. CRS-2 Education Act (ESEA) as amended by the No Child Left Behind Act (NCLBA), such as standards of adequate yearly progress (AYP), consequences of failure to make AYP (such as providing public-school choice and supplementary educational services), and ensuring that all teachers of core academic subjects (such as math and science) are “highly qualified.”4 The Government Accountability Office (GAO) has found that rural school districts may be more likely than other districts to face problems in complying with NCLBA requirements. GAO findings include those listed below. ! ! ! ! ! Achieving NCLBA goals for large enrollments of economically disadvantaged students present more challenges for rural LEAs than for nonrural LEAs. Some rural districts lack the community resources, such as libraries and museums, which may support improved academic performance. Compared with nonrural LEAs, rural LEAs are more likely to experience problems recruiting teachers because of difficulties offering competitive salaries. Small rural districts are more likely to report that factors related to school size and geographic isolation, such as limited personnel, make it difficult to release teachers and administrators for attending conferences and training, impeding their ability to implement NCLBA requirements. Some rural districts indicated limited numbers of staff created difficulties completing NCLBA requirements, such as reporting on school progress.5 Both the U.S. Department of Education (ED) and the Congress have sought to address concerns of rural school districts. In response to the GAO report, ED pointed out that it has attempted to provide additional flexibility to rural LEAs. For example, ED allows teachers in rural LEAs “extra time — up to 3 years — to meet teacher qualification requirements,” and permits states to “use a single state test for teachers to demonstrate subject matter competency for core academic subjects.”6 Congress has enacted and funded the Rural Education Achievement Program (REAP) to help address challenges that rural LEAs face. 4 See, for example, the following CRS reports: CRS Report RL32495, Adequate Yearly Progress (AYP): Implementation of the No Child Left Behind Act, by Wayne C. Riddle; CRS Report RL30834, K-12 Teacher Quality: Issues and Legislative Action; by Jeffrey J. Kuenzi; and CRS Report RL31329, Supplemental Educational Services for Children from Low-Income Families Under ESEA Title I-A, by David P. Smole. 5 U.S. Government Accountability Office (GAO), No Child Left Behind Act Additional Assistance and Research on Effective Strategies Would Help Small Rural Districts, GAO-04-909, September 2004. (Cited hereafter as GAO Effective Strategies). 6 “Meeting Minutes of Secretary’s Rural Education Task Force,” October 14, 2005, p. 7. Downloaded from [http://www.ed.gov/nclb/freedom/local/rural/index.html#meetings] on December 5, 2006. (Cited hereafter as “Task Force Minutes”). CRS-3 What Is a Rural School District? Despite the interest and concern of Congress and others about rural school districts, determining which LEAs are rural is complex and sometimes controversial. Complexity and controversy can result because different definitions of “rural” can result in significant changes in the number of such LEAs and on the targeting of any program aimed at assisting them. Estimates of the number of rural districts vary widely. For example, according to the Secretary of Education’s task force on rural education, “forty three percent of the nation’s public schools are in rural areas” and “nearly one-third of America’s school-aged children attend public schools in these communities.”7 Presumably using a more stringent definition, the GAO found that “in the 2001-02 school year, rural districts comprised 25 percent of all school districts in the country.”8 Among possible characteristics that might be used to identify rural LEAs are location, size, population density, and poverty level. Table 1 shows numbers and characteristics of LEAs identified as rural based on these characteristics. Location A rural LEA might be defined as one located in a rural area. But what is a rural area? ED’s National Center for Education Statistics (NCES), in cooperation with the Census Bureau, has devised a typology to classify the location of LEAs and individual schools. NCES uses these “locale codes” to classify each school in an LEA based on its geographic location. NCES than categorizes each LEA based on the code or codes assigned to each school.9 Since the 1980s, NCES has used the socalled “metro-centric” locale codes, which have 8 classifications. As discussed below, NCES and the Census Bureau have recently changed codes to an “urbancentric” system with 12 classifications. Metro-Centric Locale Codes. Metro-centric locale codes are based on the physical location represented by an address that is matched against a geographic database maintained by the Census Bureau. This database is the Topographically Integrated and Geographically Encoded Referencing system, or “TIGER.”10 Metrocentric locale codes are used to classify schools and LEAs according to the following typology:11 7 Task Force Minutes, p. 5. 8 GAO Effective Strategies, p. 2. 9 The general rubric is that an LEA is assigned the locale code of the schools enrolling 50% or more of the LEA’s students. If no single code accounts for 50% or more of an LEA’s students, the LEA is assigned the code of schools accounting for the highest percentage of its students. 10 Source: NCES website at [http://nces.ed.gov/ccd/rural_locales.asp]. 11 Ibid. CRS-4 1. 2. 3. 4. 5. 6. 7. 8. Large City: A central city of a core based statistical area (CMSA) or metropolitan statistical area (MSA),12 with the city having a population greater than or equal to 250,000. Mid-Size City: A central city of a CMSA or MSA, with the city having a population less than 250,000. Urban Fringe of a Large City: Any territory within a CMSA or MSA of a Large City and defined as urban by the Census Bureau. Urban Fringe of a Mid-size City: Any territory within a CMSA or MSA of a Mid-size City and defined as urban by the Census Bureau. Large Town: An incorporated place or Census-designated place with a population greater than or equal to 25,000 and located outside a CMSA or MSA. Small Town: An incorporated place or Census-designated place with a population less than 25,000 and greater than or equal to 2,500 and located outside a CMSA or MSA. Rural, Outside MSA: Any territory designated as rural by the Census Bureau that is outside a CMSA or MSA of a Large or Mid-size City. Rural, Inside MSA: Any territory designated as rural by the Census Bureau that is within a CMSA or MSA of a Large or Mid-size City. As Table 1 shows, LEAs categorized by NCES as rural (locale codes of 7 and 8) account for over 50% of all school districts and nearly 18% of all students. Including LEAs categorized as locale 6 (small towns) accounts for over 60% of all school districts and 25% of all students. As noted above (footnote 8) NCES classifies LEA status based on the code of a majority or plurality of its schools. A more stringent definition is to classify an LEA as rural only if all of its schools are classified as rural. Table 1 shows that this definition makes a significant difference: reducing the number of LEAs defined as rural from 8,200 to 7,200 and cutting the number of students served by rural LEAs by 4 million. Reductions are also seen when this more stringent definition is applied to the 6, 7, and 8 classification. 12 According to the Census Bureau website [http://www.census.gov/population/www/ estimates/aboutmetro.html]: The United States Office of Management and Budget (OMB) defines metropolitan [urban core area with a population of 50,000 or more] and micropolitan [urban core area with a population between 10,000 and 50,000] statistical areas according to published standards that are applied to Census Bureau data. The general concept of a metropolitan or micropolitan statistical area is that of a core area containing a substantial population nucleus, together with adjacent communities having a high degree of economic and social integration with that core. . . . The term “core based statistical area” (CBSA) became effective in 2000 and refers collectively to metropolitan and micropolitan statistical areas. CRS-5 Table 1. Various Definitions of Rural School Districts Rural school district criterion Estimated number of school districts meeting criterion Number of Percentage of school school districts districts meeting meeting criterion criterion Estimated with with enrollment in poverty poverty Percentage school rate of rate of of all districts Percentage 20% or 20% or school meeting of total more more districts criterion enrollment LEA categorized as locale code of 7 or 8 8,213 51.5% 8,502,709 17.7% 2,061 25.1% All schools in LEA categorized as locale code of 7 or 8 7,222 45.3% 4,486,772 9.3% 1,858 25.7% LEA categorized as locale code of 6, 7, or 8 9,964 62.5% 12,183,312 25.3% 2,618 26.3% All schools in LEA categorized as locale code of 6, 7, or 8a 9,322 58.4% 9,398,303 19.5% 2,533 27.2% All schools in LEA meeting new locale codes of rural (41, 42, 43) 5,957 37.3% 3,206,824 6.7% 1,641 27.5% LEA with enrollment less than 600 6,579 41.0% 1,612,972 3.4% 1,336 19.5% LEA in counties with fewer than 10 persons per square mile 1,757 11.0% 844,233 1.8% 550 31.3% 3,382 24.3%a Total enrollment PK to 12 LEAs with enrollment > 0 48,093,461 15,955 Source: CRS analysis of CCD data for school year 2003-2004. a. Percentage based on number of LEAs for which poverty data are available (approximately 14,000) Urban-Centric Locale Codes. Over the last two years, NCES has worked with the Census Bureau to revise location classifications for schools and LEAs. To differentiate this new system from the old locale codes, these are termed “urbancentric” locale codes.13 NCES provides two reasons for these changes: First, improvement in geocoding permits precise location of most schools based on longitude and latitude. The second reason is changes made by the Office of 13 According to NCES, the metro-centric codes are based on metropolitan statistical areas and are coterminous with counties. The urban-centric codes rely on urbanized areas, which are densely settled geographic cores with densely settled areas surrounding them. “Meeting Minutes of Secretary’s Rural Education Task Force,” April 27, 2006, p. 9. Downloaded from [http://www.ed.gov/nclb/freedom/local/rural/index.html#meetings] on December 5, 2006. CRS-6 Management and Budget (OMB) in the definition of metropolitan and nonmetropolitan areas. NCES maintains that improved geocoding technology makes the new codes more accurate. In addition, the new codes overcome some shortcomings of the metro-centric locale codes, such as the lack of a classification for suburbs, a significant undercounting of school districts in towns, and imprecision in distinguishing rural schools in remote, isolated areas from those nearer to urban cores.14 The new urban-centric locale codes are as follows: 11. Large City: Territory inside an urbanized area and inside a principal city with population of 250,000 or more. 12. Midsize City: Territory inside an urbanized area and inside a principal city with population less than 250,000 and greater than or equal to 100,000. 13. Small City: Territory inside an urbanized area and inside a principal city with population less than 100,000. 21. Large Suburb: Territory outside a principal city and inside an urbanized area with population of 250,000 or more. 22. Midsize Suburb: Territory outside a principal city and inside an urbanized area with population less than 250,000 and greater than or equal to 100,000. 23. Small Suburb: Territory outside a principal city and inside an urbanized area with population less than 100,000. 31. Fringe Town: Territory inside an urban cluster that is less than or equal to 10 miles from an urbanized area. 32. Distant Town: Territory inside an urban cluster that is more than 10 miles and less than or equal to 35 miles from an urbanized area. 33. Remote Town: Territory inside an urban cluster that is more than 35 miles from an urbanized area. 41. Fringe Rural: Census-defined rural territory that is less than or equal to 5 miles from an urbanized area, as well as rural territory that is less than or equal to 2.5 miles from an urban cluster. 42. Distant Rural: Census-defined rural territory that is more than 5 miles but less than or equal to 25 miles from an urbanized area, as well as rural territory that is more than 2.5 miles but less than or equal to 10 miles from an urban cluster. 43. Remote Rural: Census-defined rural territory that is more than 25 miles from an urbanized area and is also more than 10 miles from an urban cluster. 14 See [http://nces.ed.gov/ccd/rural_locales.asp]. CRS-7 Table 2 shows comparable codes under the 2 classification systems. Table 2. Comparisons of Metro-Centric and Urban-Centric Locale Codes Corresponding Categories Metro-Centric Urban-Centric City Suburb Town Rural 1, 2 3, 4 5, 6 7, 8 11, 12, 13 21, 22, 23 31, 32, 33 41, 42, 43 Source: NCES website [http://nces.ed.gov/ccd/rural_locales.asp]. Table 1 shows that the new codes result in significantly fewer LEAs classified as rural, assuming that the stricter “all schools” rule is used.15 A comparison of LEAs with all schools classified as rural under the metro-centric system (7 and 8) with those with all rural schools classified as rural under the urban-centric system (41, 42, and 43) shows a reduction of more than 1,200 LEAs classified as rural, and a reduction in the number of students served in rural LEAs of over 1 million. Size Another criterion for identifying rural school districts is enrollment size. There are many small LEAs in the United States. The median LEA size (i.e., the enrollment encompassing 50% of all LEAs) is 880 students.16 Ten percent of all LEAs have 100 students or fewer. As discussed below, Congress has defined a small LEA as one having fewer than 600 students. As Table 1 shows, approximately 6,500 LEAs (or more than 40% of LEAs reporting some enrollment in 2003-2004) fit this criterion, and these LEAs enroll about 1.6 million students (about 3.4% of all public school students in 2003-2004). On average, these LEAs have 1 or 2 schools and about 20 teachers. Many of these small LEAs are in midwest and southwestern states as well as in California. For example, Texas alone has over 500 LEAs with fewer than 600 students, about 45% of all its LEAs and (nearly 9% of the total of such districts in the nation). On the other hand, several states, such as South Carolina, West Virginia, Florida, Alabama, and Delaware, have only a handful of LEAs that meet the 600 student criterion. 15 The current CCD data provide urban-centric codes for individual schools but not for LEAs. 16 Percentages are based on LEAs reporting some enrollment in school year 2003-2004 (approximately 16,000 LEAs). CRS-8 Not all small school districts are located in rural areas. Of the 6,579 LEAs with fewer than 600 students, 4,700 have locale codes of 7 or 8. At the same time, nearly 900 have locale codes of 1 (i.e., are classified as within an MSA or CMSA).17 Population Density Population density is another way to define rural school districts. One measure of population density is total population divided by total land area in square miles to produce data on persons per square mile. Unfortunately, these data are readily available only by county. Since many states have multiple LEAs per county18 as well as LEAs that are located in multiple counties, merging LEA data with county data on people per square mile gives only a rough approximation of how densely or sparsely populated LEAs are. These county-level data suggest that there are many LEAs in relatively sparsely populated areas. For example, 50% of all LEAs are in counties with fewer than 90 persons per square mile; 25% are in counties with fewer than 30 persons per square mile; and 10% are in counties with fewer than 10 persons per square mile, which is the criterion used in the SRSA program (as discussed below). As one would expect, many rural LEAs are in sparsely populated counties. Ninety percent of rural LEAs outside metropolitan areas (locale code 7) are in counties with fewer than 90 people per square mile. Nearly 25% of these rural LEAs are in counties with less than 10 persons per square mile.19 Table 1 shows that LEAs in counties with less than 10 persons per square mile account for about 10% of all LEAs and about 2% of all students. As one would expect, most of the LEAs in low-density counties are in western states. For example, nearly 90% of Alaska’s LEAs are in counties with fewer than 10 people per square mile. This is true for 85% of Wyoming’s LEAs and for two-thirds of North Dakota’s and Nevada’s LEAs. On the other hand, most states east of the Mississippi River have less than 2% of their LEAs in counties that are this sparsely populated, and 19 states have no LEAs in counties with population density less than 10 persons per square mile. Poverty Although measures of poverty are not indicators of whether an LEA is rural, as we shall see, Congress has used a measure of child poverty as a means of targeting funds to some rural LEAs. The only source of data on school-age poverty (ages 5 to 17 Presumably, many of these are charter schools classified as LEAs. The CCD LEA data base does not provide information on whether an LEA is a charter school. 18 For example, the website for Lake County, Illinois lists 44 school districts; see [http://www.epodunk.com/cgi-bin/localList.php?local=6328&locTGroup=School_ districts&direction=down&sec=0&qty=44]. 19 Not all LEAs classified as rural are in sparsely populated counties. For example, nearly one-fourth of rural LEAs within metropolitan areas (locale code 8) are in counties with 600 people or more per square mile. CRS-9 17) for LEAs is the Small Area Income and Poverty Estimates (SAIPE) from the Census Bureau.20 The average poverty rate21 is 14.2% for all LEAs in the CCD that are also included in the SAIPE data base. About 25% of all LEAs have poverty rates of 20% or more (the standard Congress used to target funds to LEAs with relatively high poverty rates). This finding — that about 25% of LEAs have poverty rates of at least 20% — holds for most categorizations of rural LEAs discussed above and included in Table 1 (see last column). The exceptions are LEAs in sparsely populated counties, which tend to have higher poverty rates (31.3% are at or above this criterion), and small LEAs, which tend to have lower poverty rates (19.5% are at or above the 20% poverty level).22 Figure 1 shows how poverty rates vary by locale code. Large city school districts tend to have the highest poverty rates. LEAs in mid-size cities, towns, and rural areas outside urban areas have similar rates. LEAs in fringe cities and in rural areas within urban areas tend to have lower rates. Figure 1. Poverty Rate by Metro-Centric Locale Code 25% 20% 15% 10% 5% 0% 1. large city 2. mid- 3. fringe 4. fringe size city large city mid-city 5. large tow n 6. small tow n 7. rural outside urban areas 8. rural inside urban areas Source: CRS analysis of ED data. There are, of course, regional differences in poverty rates. Schools in the South and West tend to have higher rates than LEAs in New England and Mid-Atlantic states. For example, the overall poverty rate among LEAs in New England states is 8.4%, and the average rate for rural LEAs outside urban areas in New England is 20 See [http://www.census.gov/hhes/www/saipe/index.html]. These estimates for income year 2003 have been merged into the CCD LEA data set. There are about 14,000 LEAs that are in both data bases. 21 Poverty rate is calculated by dividing estimated number of children 5 to 17 years old in poverty by total number of children 5 to 17 years old. 22 The average poverty rate for LEAs in counties with fewer than 10 persons per square mile is 17.3%. The average poverty rate for LEAs with fewer than 600 students is 15.3%. CRS-10 10.3%. Overall school poverty in southern states is 21.0%, and the average rate for southern rural LEAs outside urban areas in 22.5%. Characteristics of Rural School Districts This section compares selected characteristics of rural LEAs to LEAs in other categories based on one definition of rural and non-rural status: LEAs’ metro-centric locale codes (codes 7 and 8 for rural LEAs). As noted above and as shown in Figure 2, rural school districts — as determined based on locale codes — enrolled about 8.5 million children in school year 2003-2004. This represents about 18% of the 48 million public school children in that school year. At the same time, nearly 75% of all public school children attend schools in large, mid-size, and fringe cities.23 Figure 2. Enrollment in LEAs in Cities, Towns, and Rural Areas (based on metro-centric locale codes) 25,000,000 20,000,000 15,000,000 10,000,000 5,000,000 Large and Mid- Fringe cities Size Cities Towns Rural Source: CRS analysis of ED data. 23 These data are based on the total enrollment variable in the CCD LEA data base. Recall that NCES uses a decision rule based on the plurality of schools’ locale codes to assign locale codes to LEAs. Tabulation of enrollments from the CCD school data base for the same school year (2003-2004) results in higher rural school enrollments (about 10 million or about 21% of the 48 million public school children. Much of the redistribution comes from schools and LEAs with the locale code of 8 (rural within an urban area). Presumably, a number of LEAs have individual schools located in these rural areas, but the plurality of these schools receive a non-rural code. The point to remember, as discussed earlier in this report, is that analyses may differ significantly, depending on how one defines rural LEAs. CRS-11 Ethnicity of Rural School Children The ethnicity of school districts differ substantially, as Figure 3 shows. Large and mid-size LEAs tend to be majority-minority. For example, in school year 20032004 whites made up only about 20% of the enrollment in large city districts and less than 50% of enrollment in mid-size school districts. On the other hand, whites made up large proportions of the enrollment in small towns (about 70%) and in rural school districts (about 80%). Figure 3. Ethnic Make-Up of Urban, Suburban, Town, and Rural LEAs (based on metro-centric locale codes) 100% 90% 80% 70% Indian 60% Asian 50% Hispanic 40% Black White 30% 20% 10% 0% 1. large city 2. mid- 3. fringe 4. fringe size city large city mid-city 5. large town 6. small town 7. rural outside urban areas 8. rural inside urban areas Source: CRS analysis of ED data. Another view of ethnic make-up of school districts is to ask where children from various groups go to school (see Figure 4). Whites tend to be enrolled in suburban and rural school districts (45% and 27%). Black, Hispanic, and Asian-American students are much more likely to be enrolled in cities and suburbs and less likely to be enrolled in rural LEAs (11% of Blacks, 8% of Hispanics, and 5% of AsianAmericans are enrolled in rural school districts). Indian students are more likely to attend schools in rural areas (40% enroll in such schools) and less likely to enroll in urban or suburban districts. CRS-12 Figure 4. Where Children of Various Ethnic Groups Go to School 60% Cities Suburbs Towns Rural Areas 50% 40% 30% 20% 10% 0% White Black His panic As ian Indian Ethnic Group Source: CRS analysis of ED data. Size Characteristics of Rural LEAs As noted earlier, one characteristic of many rural LEAs is small size. For example, many have fewer than 600 students. As Figure 5 shows, not only are many LEAs small but their schools are also relatively small. For example, while high schools in large and mid-size cities average more than 1,000 students, rural high schools outside urban areas (locale code 7) average about 200 students. Similarly, urban and suburban middle schools and primary schools tend to be substantially larger than rural schools — especially rural schools outside urban areas. Figure 5. Average Size of Schools 1400 primary middle high other 1200 1000 800 600 400 200 0 1. large city 2. mid-size city 3. fringe large city 4. fringe mid-city 5. large town 6. small town 7. rural outside urban areas 8. rural inside urban areas CRS-13 Source: CRS analysis of ED data. Class size, however, does not appear to vary substantially across LEA location (Figure 6). Class size in urban and suburban LEAs tends to be a bit over 15 children per class; rural classes tend to be at or a bit below 15 children per class. Figure 6. Average Class Size 25 primary middle high 20 15 10 5 0 1. large city 2. mid-size city 3. fringe large city Source: CRS analysis of ED data. 4. fringe mid-city 5. large town 6. small town 7. rural outside urban areas 8. rural inside urban areas CRS-14 At least part of the reason for this is that city and suburban LEAs tend to have more teachers (Figure 7). For example, high schools in large and mid-size cities have an average of about 60 full-time equivalent (FTE) teachers. Rural schools, being smaller, tend to have few teachers. For example, high schools in rural school districts outside urban areas average about 20 teachers. At least some of these differences in numbers of teachers and similarities in class size probably reflect school funding formulas based to some degree on enrollments. Figure 7. Average Full-Time-Equivalent (FTE) Teachers per School 80 70 primary middle high 60 50 40 30 20 10 0 1. large city 2. mid-size city 3. fringe large city 4. fringe mid- 5. large tow n 6. small tow n 7. rural 8. rural inside city outside urban urban areas areas Source: CRS analysis of ED data. Special Programs in Rural LEAs In recent years, Congress has supported a number of reforms aimed at increasing flexibility and improving education. The CCD schools data base provides information on three of these reforms by identifying which schools are charter schools, which are schoolwide projects under the ESEA Title I-A program, and which are magnet schools. CCD data indicate that all 3 of these reforms tend to be concentrated in urban and suburban school districts. Public charter schools are public elementary or secondary schools that are “exempted from certain rules and regulations otherwise applicable to public schools, in exchange for a commitment toward attaining positive results in meeting state content and performance standards in accordance with the terms and conditions of a charter granted by an authorized public chartering agency.”24 Figure 8 shows that, in school year 2003-2004, charter schools accounted for 9% of schools in large cities and 5% of schools in mid-size cities while accounting for less than 2% of schools in 24 Summary, CRS Report RL31128, Funding for Public Charter School Facilities: Federal Policy Under the ESEA, by David P. Smole. CRS-15 rural districts outside urban areas and about 2% of rural schools within urban areas. Similarly, in 2003-2004, nearly 80% of all charter schools were located in LEAs in large or mid-size cities or fringe cities. Only 15% were located in rural LEAs. Figure 8. Percentage of Schools That Are Charters 10% 9% 8% 7% 6% 5% 4% 3% 2% 1% 0% 1. large city 2. mid-size city 3. fringe large city 4. fringe mid- 5. large tow n 6. small tow n city 8. rural 7. rural inside urban outside areas urban areas Source: CRS analysis of ED data. The schoolwide programs provision under the Title I-A of ESEA permits a school with a relatively high percentage of pupils from low-income families (40%) to use funds from Title I-A and from certain other ESEA programs to serve all children in the school. (Schools with lower rates of pupils from low-income families must target Title I-A funds on services for Title I-A eligible children.)25 Figure 9 shows that schoolwide programs are more prevalent in large urban LEAs (where 80% of Title I schools are schoolwide programs) and LEAs in mid-size cities (where about 70% are schoolwide programs) than in rural LEAs (less than 50% of Title I schools are schoolwide programs). 25 For further information, see CRS Report RL31487, Education for the Disadvantaged: Overview of ESEA Title I-A Amendments Under the No Child Left Behind Act, by Wayne C. Riddle. CRS-16 Figure 9. Percentage of Title I Schools That Are Schoolwide Programs 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1. large city 2. mid-size city 3. fringe large city 4. fringe mid- 5. large tow n 6. small tow n city 8. rural 7. rural inside urban outside areas urban areas Source: CRS analysis of ED data. Magnet schools can be “defined as public elementary or secondary schools or education centers that offer special curricula capable of attracting substantial numbers of students with different racial backgrounds.”26 For example, a magnet school might feature certain academic subjects, such as a science magnet school or a fine arts magnet or might concentrate on particular careers, such as a health careers magnet school or a magnet focusing on aero-space careers. As Figure 10 shows, magnet schools tend to be more common in urban areas (at least in part because of the role these schools have played in desegregation plans). Magnets make up about 12% of schools in large urban school districts but only about 1% of rural schools. Similarly, nearly 90% of all magnet schools are in large and mid-size cities and fringe cities. 26 CRS Report RL33506, School Choice Under the ESEA: Programs and Requirements, by David P. Smole, pp. 15-16. CRS-17 Figure 10. Percentage of Schools That Are Magnets 14% 12% 10% 8% 6% 4% 2% 0% 1. large city 2. mid-size city 3. fringe large city 4. fringe mid-city 5. large tow n 6. small tow n 7. rural outside urban areas 8. rural inside urban areas Source: CRS analysis of ED data. The Rural Education Achievement Program (REAP) As part of the No Child Left Behind Act (NCLBA), Congress created the Rural Education Achievement Program (REAP-Title VI-B of the ESEA) to address “the unique needs of rural school districts” (§6202). These needs, according to the statute, include the lack of “personnel and resources needed to compete effectively for Federal competitive grants” and “formula grant allocations in amounts too small to be effective in meeting their intended purposes.”27 REAP authorizes 2 programs: the Small, Rural School Achievement Program (SRSA-subpart 1), which focuses on LEAs with less than 600 students and the Rural and Low-Income School Program (RLIS-subpart 2), which focuses on larger rural LEAs with relatively high poverty rates (at least 20% of children from families below the poverty line). REAP authorized $300 million for FY2002 and “such sums” for FY2003FY2007. Funds are to be divided equally between the two programs. Table 3 shows the appropriations for the program. Appropriations have grown modestly, except for the FY2006 amount, which was subject to the 1 percent across-the-board cut for most domestic discretionary programs required by P.L. 109-148. Overall, appropriations for FY2008 represent about a 6% increase over FY2002, the first year of program funding. 27 20 U.S.C. §7341a(1) and (2). CRS-18 Table 3. Appropriations for REAP Fiscal Year 2002 2003 2004 2005 2006 2007 2008 Appropriation % Change (rounded to nearest from Prior $000) Year $162,500,000 $167,653,00 3.2% $167,831,000 0.1% $170,624,000 1.7% $168,919,000 -1.0% $168,919,000 0.0% $171,854,000 1.7% REAP Eligibility, Grant Determination, and Use of Funds Eligibility. As noted above, rural LEAs can be defined in various ways, and Congress has chosen combinations of rural definitions to determine LEA eligibility, which differ for the two programs. An LEA is eligible for the SRSA program if all schools served by the LEA have a metro-centric locale code of 7 or 828 and either its average daily attendance (ADA) is less than 600 or the county or counties in which the LEA is located has a population density of fewer than 10 people per square mile. An LEA is eligible for the RLIS program if all its schools have locale codes of 6, 7, or 829 and at least 20% of the children the LEA serves are from families below the poverty line. Finally, an LEA that receives a grant under the SRSA program is not eligible for RLIS funding. Table 4 shows how these criteria interact to produce estimates of LEAs eligible for the SRSA and RLIS programs. As the table illustrates, compared with determination by locale alone, combining eligibility criteria significantly reduces the number of LEAs that are eligible for assistance. In the case of the SRSA program (as noted below), actual grants for eligible LEAs can be reduced or even eliminated depending on funds eligible LEAs receive under offsetting ESEA formula grant programs. 28 The Secretary of Education may waive the locale code requirement (but not the ADA/density requirement) based on a state government agency’s determination that the LEA is located in a rural area. (§6211(b)(2)) See U.S. Department of Education, “Guidance on the Rural Education Achievement Program (REAP),” June 2003, Appendix A-5 and Appendix A-6. (Cited hereafter as “ED REAP Guidance”.) 29 The statute does not provide the Secretary with waiver authority of the locale code requirement for the RLIS program. CRS-19 Table 4. Estimating Numbers of LEAs Eligible for REAP Programs Small Rural School Achievement Program Eligibility Locale Code of 7 or 8 and enrollment less than 600 students or in county with less than 10 persons per square mile 7,222 4,653 5,088 Rural Low-Income Schools Program Eligibility Locale Code of 6, 7, or 8 and school-age poverty at least 20% and not eligible for SRSA program 9,322 2,533 1,292 Source: CRS analysis of CCD Data. Grant Determination. Amounts that LEAs receive and aggregate state amounts are determined differently under the two programs. Under the SRSA program, an initial amount is calculated for each eligible LEA as follows: To a base grant of $20,000 an additional amount is added based on the number of students over 50 times $100; however, no initial amount may exceed $60,000. The following are some examples of initial amount calculations: ! ! ! ! LEAs with 50 students or fewer have initial amounts of $20,000. An LEA with 51 students has an initial amount of $20,100 ($20,000 plus $100 times 51-50). An LEA with 449 students has an initial amount of $59,900 ($20,000 plus $100 times 449-50). LEAs with 450 to 599 students have initial grants of $60,000 (for example, the calculation based on 451 students would be $20,000 plus $100 times 451-50=$60,100, which exceeds the maximum, so the initial amount is $60,000). As noted above, the SRSA program aims to supplement ESEA grant programs. As a result, an LEA’s final grant is based on adjusting its initial amount by the total amount it received from the following ESEA grant programs in the prior fiscal year: ! ! ! ! LEA subgrants under the Teacher and Principal Training and Recruiting Fund (Subpart 2 of Title II), LEA technology grants (Section 2412(a)(2)(A) of Title II), LEA grants under the Safe and Drug-Free Schools and Communities program (Section 4114), and Innovative Programs under the Promoting Informed Parental Choice and Innovative Programs (Part A of title V). CRS-20 As a result of this “off-set” provision, an LEA receiving a total of $60,000 or more from these 4 ESEA programs would not receive any additional funds under the SRSA program.30 State amounts for the SRSA program are the sum of amounts allocated to LEAs. (See the first column of grants in Table 5.) The current SRSA statutory formula does not provide a mechanism under which all currently appropriated funds could be allocated. For example in FY2004, the formula alone (with the $20,000 minimum and $60,000 maximum together with the off-set provision) would have allotted about $67 million; while the appropriation for that fiscal year was about $83 million for the SRSA program. The statute does not provide details on how to deal with this situation, i.e., how to ratably increase grants.31 ED does allocate all funds appropriated and has implemented a procedure to do this, which, among other things, appears to maintain the minimum and maximum formula amounts. For example, ED does not ratably increase an initial grant above $60,000. Unlike the SRSA program, grants are first made to states under the RLIS program. Also unlike the SRSA program, the statute instructs the Secretary to reserve funds from the total RLIS appropriation for Bureau of Indian Affairs (BIA) schools (½%) and for outlying areas (½%).32 The remainder is allotted to states based on each state’s share of students attending schools in eligible LEAs. Thus, for example, a state with 1% of students attending schools in eligible LEAs in that state would receive 1% of funds remaining after reserving BIA and outlying area funds. (See the second column of grants in Table 5 for FY2008 estimated state amounts.) States then award grants to eligible LEAs either competitively or based on a formula.33 Note that this procedure makes it impossible to estimate individual LEA grants at the national level (unlike the SRSA program). 30 Approximately 200 SRSA-eligible LEAs receive no SRSA funding because the amount of funding they receive from the off-setting ESEA programs equals or exceeds their initial grant amounts. However, as noted below, these LEAs are eligible for some flexibility in using funds under these 4 off-setting programs. See the discussion of uses of funds below. 31 The statute does generally provide for the situation in which funds are initially insufficient, and subsequently additional funds become available (20 U.S.C. §7345a(b)(3)). 32 The outlying areas receiving RLIS grants are American Samoa, Guam, the Northern Mariana Islands, and the Virgin Islands. 33 A state may use a formula based on the proportion of students in average daily attendance (ADA) in eligible LEAs or an alternative formula, as approved by the Secretary, that results in serving “equal or greater concentrations of children from families with incomes below the poverty line, relative to the concentrations that would be served” if the ADA formula were used (§6221(b)((2)(C). CRS-21 Table 5. REAP State Amounts for FY2008 (FY2008 estimates, rounded to the nearest $000) State or Entity Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware District of Columbia Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Small Rural School Achievement Program 0 128,000 2,158,000 1,236,000 6,116,000 2,054,000 1,194,000 0 0 0 30,000 0 906,000 5,942,000 276,000 4,532,000 3,830,000 175,000 66,000 1,685,000 0 1,405,000 2,770,000 2,985,000 55,000 5,333,000 4,996,000 4,339,000 86,000 1,279,000 1,973,000 589,000 1,859,000 813,000 661,000 2,259,000 7,093,000 1,487,000 312,000 63,000 0 907,000 Rural, Low-Income Schools 5,870,000 0 1,188,000 3,703,000 1,266,000 223,000 0 107,000 0 1,474,000 7,385,000 0 21,000 834,000 302,000 0 122,000 5,815,000 6,005,000 2,164,000 0 124,000 946,000 117,000 7,257,000 3,047,000 152,000 87,000 0 816,000 0 2,139,000 1,571,000 4,718,000 50,000 1,870,000 4,793,000 542,000 493,000 0 3,799,000 46,000 CRS-22 State or Entity Tennessee Texas Utah Vermont Virginia Washington West Virginia Wisconsin Wyoming American Samoa Guam Northern Mariana Islands Puerto Rico Virgin Islands Freely Associated States Indian set-aside Other (non-State allocations) Small Rural School Achievement Program 130,000 8,565,000 264,000 0 50,000 2,103,000 0 3,213,000 10,000 0 0 0 0 0 0 0 0 Rural, Low-Income Schools 2,807,000 7,643,000 0 0 773,000 1,019,000 3,607,000 73,000 0 81,000 175,000 55,000 0 118,000 0 430,000 100,000 Source: ED Budget Service. Note: Totals may differ slightly due to rounding. Table 5 shows a number of patterns in the distribution of grants under the two REAP programs. In a number of cases, states receive funds under one program but not under the other. For example, Alabama receives no SRSA funding but does receive RLIS grants. This is because practically none of Alabama’s 130 LEAs have enrollments less than 600. This is also true for other southeastern states, which tend to have larger consolidated or countywide LEAs and few or no small LEAs. On the other hand, Alabama has about 60 LEAs for which all schools have metro-centric locale codes of 6, 7, or 8 and poverty rates of at least 20%. Thus Alabama receives a substantial grant under the RLIS program, as do other southeastern states. On the other hand, some states receive little, if any, RLIS funding. One reason is that some states have very few high poverty LEAs. For example, Iowa, which receives no RLIS funding, has only 2 LEAs with poverty rates of 20% or more. Another reason is that some states have many LEAs that are eligible for both programs but can only be eligible for SRSA grants, as required under the statute. For example, South Dakota, which received more than 10 times the funding under the SRSA than under RLIS, has nearly 85% of its LEAs that are eligible for both programs and thus can receive grants only under the SRSA program. Finally, there are several states that receive little or no funds from either program. In FY2008, CRS-23 Hawaii, Maryland, and Vermont receive no REAP funding.34 Hawaii is a single statewide LEA. Maryland’s LEAs do not appear to meet eligibility criteria for either program. Vermont LEAs may not apply for the SRSA program because grants are considered too small.35 Use of Funds. Recipients of SRSA grants may use funds for activities authorized by several ESEA programs: ! ! ! ! ! Improving Basic Programs Operated by Local Educational Agencies (Part A of Title I), Teacher and Principal Training and Recruiting Fund and Enhancing Education Through Technology (Part A or D of Title II), Language Instruction for Limited English Proficient and Immigrant Students (Title III), Safe and Drug-Free Schools and Communities and 21st Century Community Learning Centers (Part A or B of Title IV), and Innovative Programs (Part A of Title V). In addition, all LEAs that are eligible for SRSA grants (whether or not they receive grants because off-setting ESEA funding exceeds initial grant calculations) have the flexibility to use funds from the off-setting ESEA programs for any activities authorized by the above ESEA programs.36 ED provides the following example of use of funds under REAP-Flex: “[A]n LEA may use funds under the Safe and Drug-Free Schools Program (Title IV, Part A) to incorporate technology into its early reading program — an authorized local activity under the Educational Technology State Grant (Title II, Part D).”37 The GAO also found that flexibility under the SRSA program allowed small, rural LEAs to redirect funds to crucial NCLBA needs. “[I]n one rural state contacted, officials reported that many of their districts used Safe and Drug-Free School 34 This was also the case for the District of Columbia (all schools in an urban area). 35 The GAO study, which selected the most rural states for its study “based on the percentage of their school districts in rural communities, the percentage of their students attending schools in rural communities, and the average distance between the school district in the state and the nearest metropolitan statistical area as a measure of geographic isolation” (GAO Effective Strategies, p. 3). Vermont met these criteria and was included in GAO’s study. 36 In its guidance on REAP, ED refers to alternative use of funds as “REAP-Flex” and differentiates this flexibility from other ESEA flexibility as follows: REAP-Flex does not involve a transfer of funds from one program to another. Rather, REAP-Flex gives an LEA broader authority in spending “applicable funding” for alternative uses under selected federal programs. On the other hand, when an LEA transfers funds from one program to another under the transferability authority in section 6123, the transferred funds increase the allocation of the receiving program and are subject to all of the rules and requirements of the receiving program. ED REAP Guidance, (section II-B-1). 37 “ED REAP Guidance,” section II-B-5. CRS-24 Program funds to support their technology initiatives, which, in turn, helped with implementing some of the provisions of NCLBA.”38 RLIS grant recipients may use funds for the following purposes: ! ! ! ! ! ! ! Teacher recruitment and retention, including the use of signing bonuses and other financial incentives; Teacher professional development, including programs that train teachers to utilize technology to improve teaching and to train special needs teachers; Educational technology, including software and hardware, as described in part D of Title II (Enhancing Education Through Technology); Parental involvement activities; Activities authorized under the Safe and Drug-Free Schools program under part A of Title IV; Activities authorized under part A of Title I; and Activities authorized under Title III (Language Instruction for Limited English Proficient and Immigrant Students).39 The GAO reported other uses of REAP funds to help meet costs associated with NCLBA requirements, including ! ! ! ! 86% of responding rural superintendents reported spending REAP funds on student and teacher technology needs; 66% reported using REAP funds for NCLBA supplementary services for students; 94% said they used these funds for professional development related to helping teachers meet NCLBA highly qualified teacher requirements; and 60% used REAP funds for student remedial services to prepare them for annual assessments.40 Distribution of Certain ESEA Grants to Rural LEAs As noted above, two purposes of the REAP program are to compensate rural school districts because they are at a disadvantage in obtaining competitive grants from the Department of Education, and ED formula grants are often too small to have an impact. Certainly conventional wisdom would support these contentions. With respect to competitive grants, rural schools — especially small rural schools — tend to have fewer administrative staff, who are generally thought to be key to writing grant proposals and obtaining competitive grants. Regarding formula grants, since 38 GAO Effective Strategies, p. 35. 39 States may reserve no more than 5% of RLIS funds for state administration and technical assistance (§6222(b)). 40 Ibid., p. 34 CRS-25 many grants are distributed to states and to LEAs based on factors related to size (e.g., school-age population and school-age poverty), smaller LEAs receive smaller grants. The grant application process and the subsequent federal reporting requirements can reduce or eliminate the value of formula grants for some smaller LEAs. Data on formula and competitive grants do not completely support these conventional notions, however. Does REAP Compensate Small, Rural LEAs for Small Formula Grant Amounts? As noted above, the REAP aims to compensate rural school districts because they often receive formula grants that are “too small to be effective in meeting” the intended goals of the formula grant programs.41 Table 6 examines this proposition for LEAs eligible for the SRSA program.42 The table presents median grants for the four ESEA programs that off-set the final SRSA grants (see discussion above on SRSA grant determination). Clearly SRSA-targeted LEAs tend to receive smaller grants than do other LEAs. For example, the median grants for ESEA Title II-A (Teacher and Principal Training and Recruiting Fund and Enhancing Education Through Technology) for SRSA-eligible LEAs is less than 50% of the median grant for other LEAs ($14,300 vs. $30,600). Similarly, the median total of all four grants for SRSA-eligible LEAs is $19,700 compared with $42,300 for other LEAs. At the same time, SRSA-eligible grants are somewhat larger on a per student basis. For example, the combined median per-student grant for the four ESEA programs is $89 compared with a combined median per-student grant of $60 for all other LEAs. 41 42 §6202(2). LEA grants are not available at the national level for the RLIS programs because funds are allocated to states by formula, and no national data are available on states’ distribution of RLIS grants to LEAs. CRS-26 Table 6. Comparison of Grants for SRSA-Eligible LEAs and Other LEAs (grants rounded to nearest $100) SRSA-eligible LEAs Grant Median grant All but SRSAeligible LEAs Median Median grant grant per per Median student grant student ESEA Title II-A 14,300 63 30,600 44 ESEA Title III-D 1,500 6 2,100 4 ESEA Title IV-A 1,800 8 4,100 5 ESEA Title V 2,100 8 3,900 6 Total of 4 ESEA grants 19,700 89 42,300 60 Final SRSA grant 19,000 92 NA NA Final SRSA + ESEA grants total 42,400 181 NA NA Source: CRS analysis of ED Budget Service data. Note: Median. Table 6 shows that the median SRSA grant (for FY2004) was $19,000 for the nearly 4,000 LEAs that received grants.43 The median per-pupil grant was $92. In part, because of the SRSA off-setting requirement, SRSA grant totals and per-pupil amounts can vary widely. Maximum grants were, of course $60,000. However, grants ranged as low as $39. Per student grants also ranged widely, from less than $1.00 per student to $19,000 per student. Per-pupil amounts for some western LEAs were substantial because these LEAs qualified for the minimum grant of $20,000 and are very small, in a few cases having less than 10 students. Finally, Table 6 shows that the SRSA program does indeed compensate small, rural school districts for relatively small ESEA grants. The median combination of the four ESEA grants and the SRSA grant was $42,400, roughly the same as the median combined ESEA grants for other LEAs. On a per-pupil basis, SRSA grantees fared better: $181 vs. $60. In addition, the SRSA flexibility provision permits small, rural LEAs to concentrate funds from the off-setting programs and the SRSA grant on one or a few activities authorized by these programs; whereas, other LEAs must confine their use of these ESEA funds to activities authorized by the individual programs. 43 Although eligible for SRSA grants, 223 LEAs received no grant because the sum of the four off-setting ESEA formula grant programs equaled or exceeded their initial SRSA amount. CRS-27 Are Rural LEAs at a Disadvantage for Obtaining Competitive Grants? In FY2003, according to the ED database on discretionary and formula grant awards,44 the Office of Elementary and Secondary Education (OESE) awarded 500 discretionary (i.e., competitive) grants.45 Of these, about 60% went to LEAs for which data are available from the CCD. The remainder went to other entities, such as institutions of higher education. Overall, OESE competitive grants for FY2003 totaled approximately $78 million, of which about $51 million went to LEAs represented in the CCD database. Three grant competitions accounted for nearly 80% of the grants and 67% of the funds: ! ! ! 21st Century Community Learning Centers/After School Learning Centers, Fund for the Improvement of Education/Smaller Learning Communities, and Improving Literacy Through School Libraries. Table 7 shows that, on several measures, rural LEAs (locale codes 7 and 8) fared well in obtaining OESE grants. These LEAs received slightly more than onethird of the LEA grants (34.6%) and of the overall LEA funding (34.2%). In addition, average grants were similar for large urban LEAs ($158,000) and for rural LEAs (locale 7: $161,000 and locale 8: $149,000). However, when comparing the number of LEAs that received grants, disparities are evident. About 2% of all LEAs and about 4% of large urban LEAs received OESE grants in FY2003 but only about 1% of rural LEAs received grants. 44 45 See [http://www.ed.gov/fund/data/award/grntawd.html]. The database contains grants authorized under other statutes; however, this analysis was limited to OESE competitive grants, which are generally authorized under ESEA. CRS-28 Table 7. Data on Competitive Grant Recipients (ED Office of Elementary and Secondary Education, or OESE, FY2003) 1. Large City Total OESE discretionary grants 36 46 57 Percentage of discretionary grants 10.9% 13.9% 17.2% Average grant $158,000 $142,000 $136,000 Total grant amount 6. Small Town 7. Rural, Outside MSA 8. Rural, Inside MSA Totals (grants that can be classified by locale code) 3 42 80 31 331 0.9% 12.7% 24.2% 9.4% 100.0% $149,000 $164,000 $188,000 $161,000 $149,000 $155,000 4. Urban 3. Urban Fringe of a 2. Mid-Size Fringe of a Mid-Size 5. Large City Large City City Town 36 10.9% $5,702,000 $6,532,000 $7,754,000 $5,355,000 $491,000 $7,887,000 $12,909,000 $4,610,000 $51,240,000 Percent of total grant amount 11.1% 12.7% 15.1% 10.5% 1.0% 15.4% 25.2% 9.0% 100.0% Percent of all LEAs receiving grants 4.4% 4.8% 2.2% 2.3% 2.3% 2.4% 1.4% 1.1% 2.1% Source: CRS analysis of ED grants database at [http://www.ed.gov/fund/data/award/grntawd.html]. Note: Average and total grants rounded to the nearest $000. CRS-29 Urban-Centric Locale Codes: Possible Impacts on Eligibility As noted above, eligibility for the two REAP programs is based, in part, on the locale codes of LEAs’ schools. In addition to other requirements, all schools in an LEA must have metro-centric locale codes of 7 or 8 for the LEA to be eligible for the SRSA program and 6, 7, or 8 for an LEA to be eligible for the RLIS program. As discussed above, NCES and the Census Bureau have devised a new set of urbancentric codes, which are said to more accurately depict a school’s geographic location. Currently, both sets of locale codes are available and will be available for perhaps two years. Thereafter NCES will only make available the more recent urbancentric locale codes.46 Therefore the Congress may wish to consider how the use of the new codes might impact REAP eligibility and funds distribution. Table 8 compares the number of schools classified according to metro-centric locale codes as located in cities, urban fringes, towns, and rural areas with those classified according to urban-centric locale codes as located in these areas. Clearly there is a great deal of overlap. For example, nearly 92% of all schools classified as rural under the metro-centric system are also classified as rural under the urbancentric system. At the same time, some schools are classified differently under the two systems. For example, about 8% of schools classified as rural under the metrocentric system would not be rural under the urban-centric system. Table 8. Comparison of Schools Classification by Metro-Centric and by Urban-Centric Locale Codes Urban-centric locale codes (new codes) City Suburb Towns Rural (codes 11- (codes 21- (codes 31- (codes 4113) 23) 33) 43) Metrocentric locale codes (old codes) Totals Cities (codes 1 and 2) 96.0% 2.4% 0.4% 1.2% 100.0% 24,895 Urban fringe (codes 3 and 4) 1.4% 80.7% 13.4% 4.5% 100.0% 31,268 Towns (codes 5 and 6) 0.2% 0.4% 87.0% 12.4% 100.0% 9,628 Rural (codes 7 and 8) 0.5% 2.0% 5.7% 91.8% 100.0% 29,497 Source: CRS analysis of CCD data. 46 Estimated number of schools Telephone conversation with John Sietsema of NCES, November 14, 2006. CRS-30 What impact would a change to the urban-centric locale codes have on the distribution of REAP funds? Table 9 shows estimated SRSA state totals and estimated numbers of LEA grant recipients based on metro-centric and urban-centric locale codes.47 All other formula factors (e.g., enrollment and county density) were the same for both sets of estimates. All formula factors (except for the urban-centric codes, which came from the CCD database) came from an ED Budget Service data base used to determine LEA SRSA grants.48 The total allocated under both scenarios ($83.2 million) is the total of FY2004 SRSA grants for LEAs for which complete data were available.49 Table 9 shows some substantial changes, in both the estimated total funds states would receive and the number of LEAs receiving SRSA grants. Estimated dollar differences range from an increase of $1.3 million (Oklahoma) to a loss of $1.2 million (California). Estimated percentage changes range from a gain of nearly 40% (North Dakota and South Dakota) to a loss of more than 60% (Massachusetts).50 Overall, an estimate of 386 fewer LEAs would receive grants based on the urban-centric locale code criterion. This estimate includes 421 LEAs estimated to lose grants based on the urban-centric codes and 35 LEAs that would receive grants if the urban-centric codes were used. Virtually all of the LEAs estimated to lose funding are classified as metro-centric code 7 (a reduction of 186 LEAs or 6% in this category) or metro-centric code 8 (a reduction of 208 LEAs or 23% in this category).51 Recall that the latter code designated rural LEAs within metropolitan areas. The urban-centric locale coding systems, which takes into account distance from metropolitan areas, apparently does not classify some of these LEAs as rural. 47 RLIS allocations are not estimated because it is uncertain which urban-centric code or codes should be used to substitute for metro-centric code 6. 48 This approach may underestimate the impact of the new codes because it does not allow for LEAs that were not eligible based on the Budget Service data to be deemed eligible. It only permits estimating numbers that would no longer be eligible and resulting reallocation. In some cases, states gain in the estimated number of LEAs receiving grants. This is because there are additional funds to allocate (because some LEAs are no longer eligible). As a result, some LEAs now receive funding when they would not under the implementation of current law. 49 The amount allocated to LEAs for FY2004 was $83.5 million. This total is slightly higher than the amount allocated in Table 9 because complete data were not available for 15 LEAs that received FY2004 grants. 50 Larger estimated percentage changes occur for Wyoming (more than a 100% increase in the state total) and for Delaware (a 100% loss), although these changes are from very small original totals. 51 The remainder are LEAs that were eligible because of state alternative rural definitions. These LEAs perhaps would remain eligible if their alternative definitions were accepted by ED. CRS-31 Table 9. Estimates of Amounts and Number of Grantees Under the SRSA Formula Based on Metro-Centric and Urban-Centric Locale Codes (estimated grants rounded to nearest $000; calculations may differ slightly due to rounding) State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware District of Columbia Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Estimated total Estimated total grants based on grants based on current law current law Percentage (metro-centric (urban-centric Dollar dollar locale codes) locale codes) difference difference $0 $0 $0 0% 171,000 216,000 45,000 26% 1,693,000 1,665,000 -28,000 -2% 1,058,000 1,069,000 11,000 1% 5,198,000 3,985,000 -1,213,000 -23% 2,020,000 2,280,000 259,000 13% 1,124,000 951,000 -172,000 -15% 59,000 0 -59,000 -100% 0 0 0 0% 0 0 0 0% 21,000 23,000 3,000 13% 0 0 0 0% 799,000 918,000 119,000 15% 5,882,000 5,392,000 -490,000 -8% 264,000 218,000 -47,000 -18% 4,743,000 4,991,000 248,000 5% 3,641,000 4,130,000 489,000 13% 165,000 84,000 -81,000 -49% 44,000 27,000 -17,000 -39% 1,797,000 1,913,000 116,000 6% 0 0 0 0% 1,170,000 432,000 -738,000 -63% Estimated number of LEA grantees based on current law (metro-centric locale codes) 0 13 78 51 285 86 35 2 0 0 2 0 44 249 11 168 159 7 2 114 0 37 Estimated number of LEA grantees based on current law (urban-centric locale codes) 0 15 69 49 210 84 27 0 0 0 2 0 42 206 8 163 156 4 1 105 0 14 Change in estimated number of grantees 0 2 -9 -2 -75 -2 -8 -2 0 0 0 0 -2 -43 -3 -5 -3 -3 -1 -9 0 -23 CRS-32 State Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Washington Estimated total Estimated total grants based on grants based on current law current law Percentage (metro-centric (urban-centric Dollar dollar locale codes) locale codes) difference difference 2,319,000 2,129,000 -190,000 -8% 2,750,000 3,051,000 301,000 11% 57,000 56,000 -1,000 -1% 5,152,000 5,620,000 468,000 9% 5,342,000 5,783,000 441,000 8% 7,257,000 7,376,000 119,000 2% 200,000 209,000 9,000 4% 969,000 730,000 -239,000 -25% 1,991,000 958,000 -1,033,000 -52% 300,000 391,000 91,000 30% 1,304,000 1,278,000 -26,000 -2% 606,000 294,000 -312,000 -51% 683,000 944,000 260,000 38% 1,589,000 941,000 -648,000 -41% 6,945,000 8,274,000 1,329,000 19% 1,242,000 1,333,000 91,000 7% 164,000 183,000 19,000 12% 83,000 93,000 10,000 12% 0 0 0 0% 881,000 1,205,000 324,000 37% 126,000 131,000 5,000 4% 8,100,000 8,826,000 725,000 9% 148,000 103,000 -45,000 -30% 0 0 0 0% 39,000 45,000 6,000 16% 2,017,000 2,106,000 89,000 4% Estimated number of LEA grantees based on current law (metro-centric locale codes) 119 122 4 243 332 320 8 59 61 27 77 20 86 44 346 71 9 3 0 70 4 375 7 0 2 110 Estimated number of LEA grantees based on current law (urban-centric locale codes) 102 120 3 227 320 295 8 44 28 28 69 11 92 28 346 64 9 3 0 73 4 357 5 0 2 98 Change in estimated number of grantees -17 -2 -1 -16 -12 -25 0 -15 -33 1 -8 -9 6 -16 0 -7 0 0 0 3 0 -18 -2 0 0 -12 CRS-33 State West Virginia Wisconsin Wyoming Totals Estimated total Estimated total grants based on grants based on current law current law Percentage (metro-centric (urban-centric Dollar dollar locale codes) locale codes) difference difference 0 0 0 0% 3,101,000 2,853,000 -248,000 -8% 7,000 16,000 9,000 137% 83,221,000 83,221,000 0 0% Estimated number of LEA grantees based on current law (metro-centric locale codes) 0 118 2 3,982 Estimated number of LEA grantees based on current law (urban-centric locale codes) 0 103 2 3,596 Change in estimated number of grantees 0 -15 0 -386 Source: CRS analysis based on CCD data and ED Budget data. Note: These are estimated grant totals only. In addition to other limitations, much of the data that would be used to calculate actual grants are not yet available. These estimates are provided solely to assist in comparisons of the relative impact of alternative formulas in the legislative process. They are not intended to predict specific amounts states will receive. CRS-34 Possible Policy Issues Shift in Locale Codes. One policy issue is the possible shift to the urbancentric locale codes in determining eligibility for REAP grants. As discussed above, replacing metro-centric codes with the newer, arguably more accurate urban-centric codes will remove some LEAs from eligibility and add others. As a result, some LEAs and states will lose funding, others will gain. Unless there are significant increases in REAP funding (unlikely given current budget constraints), any formula change will be controversial because there will be “winners” and “losers.” This, in turn, means that there are no easy policy alternatives. One possible option would be to mandate the continued use of the metro-centric codes. This approach has the obvious advantage of ensuring that LEAs are not eliminated from the program and that funds are not shifted from state to state. It has the disadvantage of continuing the use of rural definitions that may be inferior to other, available definitions, and, as a result, allocating funds to LEAs that may not need as much assistance as “truly rural” LEAs that are in greater need of assistance. Another possible option would be to hold harmless those eliminated LEAs for a period of time (perhaps at a decreasing percentage of their prior year grants) so they can adjust to the funding loss. Although softening the blow to these LEAs, it would result in lower grants (assuming level or near-level funding) to other, remaining LEAs as funds are distributed among the 2 groups already served and the newly eligible LEAs. Allocating Excess Funds. As discussed above, the current SRSA formula does not permit all currently appropriated funds to be allocated to LEAs. In part, this is because SRSA grants are capped at $60,000. The act does not specify how to deal with this situation. As a result, ED has had to make policy on how these excess funds should be distributed. Apparently to adhere to the statute, the ED “ratable increase”52 procedure maintains both the $60,000 cap and $20,000 floor for the SRSA grants and ratably increases grants falling between these two requirements. The statute could be amended to reflect ED’s current procedures. This would ensure that ED continues to follow this procedure in the future. Alternatively, the statute could be amended to provide a different policy for dealing with additional appropriations. For example, the statute could specify a ratable increase procedure under which the minimum and maximum grants could be ratably increased along with all other grants. Presumably, this approach would slightly reduce LEAs’ grants that fall between the minimum and maximum grants. Increase Benefits to Small, Poor LEAs. As discussed above, LEAs that are eligible for the SRSA program (based, in part, on enrollment below 600) are not eligible for grants under the RLIS program (which targets rural LEAs with relatively high poverty rates). Since it can be argued that these LEAs are triply disadvantaged: being rural, small, and poor, a possible change in the statute could recognize this by allowing small, poor rural LEAs to benefit from both programs. This would add 52 Ratably increasing grants means increasing grants in proportion to the relationship between each LEA’s initial grant and the total excess funds to be distributed. CRS-35 about 1,000 LEAs to the RLIS eligibility list and redistribute RLIS state grants by increasing grants to states with large numbers of small, poor LEAs and reducing grants to states with few small LEAs (mostly states in the Southeast). If further targeting were desired, a higher poverty threshold could be set for small, poor LEAs. For example, a poverty rate of 30% or greater would add less than 200 LEAs to the RLIS-eligibility pool. Adjust SRSA Formula to Reduce Anomalies. The SRSA formula has resulted in some quirks, which might be addressed by formula modifications. For example, the minimum grant of $20,000 results in some very large per-pupil grants. While the median per-pupil grant is about $90, a few LEAs receive per-pupil grants as high as $19,000. This results because they have only one or a few students.53 One approach for reducing this result would be to limit LEA participation to LEAs with a minimum total enrollment. If minimum enrollment were set at 10,54 about 100 LEAs would be eliminated. Another seeming anomaly occurs when LEAs have off-setting program amounts that are just a few dollars less than their final SRSA grant. For example, some LEAs receive grants as low as $39. A solution to this problem would be to eliminate final grants that are deemed to be below a size to be effective. Alternatively, grants deemed too small on a per-pupil basis could be eliminated. (Presumably some LEAs take this into account by not applying for grants after a year in which they receive a minimal amount). For example, about 350 LEAs have per-pupil grants of less than $30, nearly 200 LEAs have per-pupil grants for less than $20, and about 75 LEAs have per-pupil grants of less than $10. These funds could then be distributed to other LEAs to enhance their grants. Another problem occurs when LEAs eligible for the SRSA program have offsetting grants larger than their initial grant. While these LEAs can still use the REAP Flex provision, they receive no additional REAP funds. One alternative to this situation would be to calculate the SRSA initial grants without the minimum and maximum grants of $20,000 and $60,000, subtract the off-setting grant amounts, then apply the minimum and maximum grant amounts. This would reduce the number of LEAs that are eligible but receive no funding. (About 200 LEAs currently experience this.) A final concern that some states have is that, unlike the RLIS program, states receive no state administration funding under the SRSA program, despite having to provide ED with much of the data used to allocate funds (such as off-setting program grant amounts). This could be addressed by reserving 2% (or some other percent) of the appropriation for the SRSA program for state administration. These funds could be distributed to states based on their proportion of students enrolled in SRSA eligible LEAs for the prior year. Of course, this would reduce funds going to small, rural LEAs by the percentage reserved for state administration. 53 According the CCD data, 5 states (Arizona, Maine, Minnesota, Montana, and Nebraska) report at least one LEA with one student. 54 This is a standard used in the ESEA Title I-A program, which has an eligibility threshold of 10 children living in poor families in order for LEAs to receive Title I-A funds. CRS-36 Appendix: Data Sources Data for this report came from a variety of sources: the Common Core of Data (CCD) collected and made available by the National Center for Education Statistics (NCES) at the U.S. Department of Education (ED), REAP allocation spreadsheets from the ED Budget Service, the ED database on discretionary and formula grant awards55, and the Small Area Income and Poverty Estimates (SAIPE) and county population density data both from the U.S. Census Bureau.56 CCD Data. NCES annually collects data on all public schools and public school districts, which it provides through the CCD database. Among the CCD’s purposes is”to provide basic information and descriptive statistics on public elementary and secondary schools and schooling in general.”57 State educational agencies (SEA) are mainly responsibility for providing CCD data on schools and school districts to NCES. Two CCD data files were used in this report: the Public School Universe and the Local Educational Agency (School District) Universe. Both data files were pared down to include only open schools and operating school districts. In addition, only schools and school districts in the 50 United States and the District of Columbia were included. (The CCD includes other schools and school districts, which are not eligible for REAP funding, such as Department of Defense Schools.) The Public School Universe file was used to determine locale code eligibility for school districts. Although the Local Educational Agency Universe file classified school districts by locale code, the algorithm NCES used to attribute codes differed from the requirement in the REAP program. REAP requires that all schools have certain locale codes (7 or 8 or 6, 7, or 8). This file was also used to determine ethnic and racial characteristics of rural and non-rural schools. In addition, the Public School Universe file was used to determine which schools had special characteristics; for example, how many charter schools and magnet schools are in rural areas. The Local Educational Agency Universe file was used to determine the number of school districts that would be considered as rural under various definitions. To do this, this file was merged with other databases. For example, it was merged with the SAIPE database to determine LEA poverty rates, which are necessary to determine eligibility for the RLIS program. The LEA Universe file was merged with data on ED grants to analyze how many rural and non-rural school districts received ED competitive grants. The file was also merged with Census Bureau data on countylevel population density. 55 See [http://www.ed.gov/fund/data/award/grntawd.html]. 56 See [http://www.census.gov/population/www/censusdata/density.html]. 57 See [http://nces.ed.gov/ccd/aboutCCD.asp]. Files for the 2003-2004 school year were used because they are the only files containing school data on both the metro-centric and urban-centric locale codes. CRS-37 ED Budget Service Data. The requirements of REAP eligibility necessitate that ED to collect data from the states that are not otherwise available. Most notably, for the SRSA program, ED must collect data on local ESEA grants for the off-setting programs and information necessary to evaluate states’ requests for waivers of the locale code criterion. These formula factors cannot be obtained from any other source. As a result, Budget Service data58 had to be used to estimate the impacts of changing locale codes from metro-centric to urban-centric (while holding constant all other eligibility criterion). The Budget Service data had to be merged with CCD data to include the urban-centric codes. ED Grants Data. ED provides information on grants awarded by fiscal year. Data on competitive grants made by the Office of Elementary and Secondary Education (OESE) for FY2003 were merged with the CCD Local Educational Agency Universe file. The grants awards data base contains data on 500 OESE awards for that fiscal year. Of those awards, 337 were made to LEAs in the CCD file. Other grantees included institutions of higher education and state educational agencies. Merging data sets reduces the number of cases for which there are useable data. For example, the CCD Local Educational Agency Universe file has data on about 17,800 school districts; however, only about 16,000 appeared to have enrolled children in school year 2003-2004. When the CCD data are merged with the Budget Service REAP data set, about 15,600 school districts have useable data. 58 Excel spreadsheet obtained from the ED Budget Service, March 2005.