For other versions of this document, see http://wikileaks.org/wiki/CRS-RL34066 ------------------------------------------------------------------------------ Order Code RL34066 English Language Acquisition Grants Under the No Child Left Behind Act: Analysis of State Grant Formula and Data Options June 29, 2007 Rebecca R. Skinner Specialist in Social Legislation Domestic Social Policy Division English Language Acquisition Grants Under the No Child Left Behind Act: Analysis of State Grant Formula and Data Options Summary The number of limited English proficient (LEP) students enrolled in K-12 education increased by 60.8% from the 1994-1995 school year to the 2004-2005 school year, while total student enrollment increased by 2.6% over the same time period. Given this tremendous growth in the LEP student population and the likelihood that Congress will consider legislation to reauthorize the Elementary and Secondary Education Act of 1965 (ESEA), as amended by the No Child Left Behind Act of 2001 (NCLBA; P.L. 107-110), during the 110th Congress, this report examines the formula used to provide grants to states under the English Language Acquisition program, authorized by Title III of the ESEA. This program provides grants to states to help ensure that LEP and recent immigrant students attain proficiency in English. Much of the debate surrounding the reauthorization of this program has focused on the data used to determine how many LEP and immigrant students are in each state, as these data are the basis upon which grants are determined. This report examines the American Community Survey (ACS) data that the U.S. Department of Education has used to calculate state grants since FY2005. It also analyzes state-reported data that could potentially be used to calculate these grants. Differences in LEP and immigrant student counts based on the different data sources are compared, revealing substantial differences in student counts for some states depending on the data source used. FY2007 grants are calculated using both the ACS and state-reported data to examine the potential differences in state grant amounts depending on the data source used. The differences in student counts that exist between the ACS and state-reported data are reflected in the differences in estimated state grant amounts, as some states would receive substantially more or less funding if state-reported data were used to calculate grants rather than the ACS data. Consideration is also given to the drawbacks of using either the ACS or state- reported data and possible alternative strategies for determining state grant awards (e.g., averaging the student counts from the ACS and state-reported data) are discussed. This report will be updated as warranted by legislative action. Contents English Language Acquisition State Grants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Data Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 American Community Survey Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 State-Reported LEP Student Counts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 State-Reported Immigrant Student Counts . . . . . . . . . . . . . . . . . . . . . . . . . . 9 ACS Data Compared with State-Reported Data . . . . . . . . . . . . . . . . . . . . . . . . . 11 Data Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Student Count Data Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Estimated FY2007 State Grants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Selecting Data on Which To Base the Distribution of Funds . . . . . . . . . . . . . . . 17 ACS Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 State-Reported Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Data Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Possible Alternatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 List of Tables Table 1. LEP Student Counts from the 2003, 2004, and 2005 American Community Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Table 2. Immigrant Student Counts from the 2003, 2004, and 2005 American Community Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Table 3. Estimated State LEP Student Counts Based on Data Available from the Common Core of Data and the NCELA: 2002-2003, 2003-2004, and 2004-2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Table 4. Immigrant Student Counts Based on State-Reported Data in Title III Biennial Reports: 2002-2003 and 2003-2004 . . . . . . . . . . . . . . . 10 Table 5. Comparison of Estimated Limited English Proficient Student Counts from the 2005 American Community Survey and 2004-2005 State-Reported Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Table 6. Comparison of Estimated Immigrant Student Counts from the 2005 American Community Survey and 2003-2004 State-Reported Data . . . . . . 14 Table 7. Estimated FY2007 State Grants Based on the 2005 American Community Survey and State-Reported Data . . . . . . . . . . . . . . . . . . . . . . . 16 English Language Acquisition Grants Under the No Child Left Behind Act: Analysis of State Grant Formula and Data Options The number of limited English proficient (LEP) students enrolled in K-12 education increased by 60.8% from the 1994-1995 school year to the 2004-2005 school year; total student enrollment increased by 2.6% over the same time period. Given this tremendous growth in the LEP student population and the likelihood that the 110th Congress will consider legislation to reauthorize the Elementary and Secondary Education Act of 1965 (ESEA), as amended by the No Child Left Behind Act of 2001 (NCLBA; P.L. 107-110), this report examines the formula used to provide grants to states under the English Language Acquisition program, authorized by Title III of the ESEA. This program provides grants to states to help ensure that LEP and recent immigrant students attain proficiency in English. Much of the debate surrounding the reauthorization of this program has focused on the data used to determine how many LEP and immigrant students are in each state, as these data are the basis upon which grants are determined. This report begins with a general overview of the English Language Acquisition program, focusing specifically on the state grant formula. This is followed by a detailed analysis of the American Community Survey (ACS) data currently used by the U.S. Department of Education (ED) to calculate these grants, as well as state- reported data that could potentially be used to calculate these grants. The third section of the report compares student counts based on ACS data and state data and examines differences in estimated FY2007 state grants if state data were used as the basis for determining the awards. The report concludes with an examination of some of the drawbacks of using either the ACS or state-reported data for determining state grants and other possible alternative strategies for calculating state grants. English Language Acquisition State Grants Title III, Part A of the ESEA authorizes formula grants to states to ensure that limited English proficient (LEP) students and immigrant children develop English proficiency.1 Prior to determining state grant allocations, statutory language provides 1 Statutory language defines a limited English proficient student to be a student (1) who is between the ages of 3 and 21, (2) who is enrolled or is preparing to enroll in an elementary or secondary school, (3) who was not born in the United States or whose native language is a language other than English, who is a Native American or Alaska Native, who is a native of the outlying areas, who comes from an environment where a language other than English has had an impact on the student's level of English language proficiency, or is a (continued...) CRS-2 for several reservations of funds. These include reservations for national activities, for schools serving Native American and Alaska Native students, and for the outlying areas.2 After reserving the required funds, grants to the 50 states, the District of Columbia, and Puerto Rico are determined based on the state's proportional share of LEP students and immigrant students relative to the U.S. population of LEP students and immigrant students.3 These shares are then weighted, with a higher weight (0.8) being assigned to the state's population of LEP students and a lower weight (0.2) being assigned to the state's population of recent immigrant students. No state can receive a grant less than $500,000. The grant to Puerto Rico cannot exceed 0.5% of the total available for state distribution. Data Availability In determining the number of LEP and immigrant students in an individual state and in the United States, statutory language directs ED to use "the more accurate" of (1) data available from the American Community Survey (ACS), or (2) the number of children being assessed for English proficiency as required under Title I of the ESEA.4 In practice, ED has been using the ACS data to make state allocations since FY2005. Title III grants for a specific fiscal year have been based on ACS data from two years prior. For example, FY2007 grants are based on the 2005 ACS data. According to testimony provided by Cornelia Ashby, Director of Education, Workforce, and Income Security Issues at the Government Accountability Office (GAO), ED has not used state data because it believes the state data are incomplete.5 1 (...continued) migratory student whose native language is not English and who comes from an environment where English is not the dominant language, and (4) whose difficulties in speaking, reading, writing, or understanding English may prevent the student from reaching the proficient level on state assessments required under Title I, succeeding in classrooms where English is the language of instruction, or participating fully in society (Section 9101). Statutory language defines an immigrant student as an individual 3 to 21 years old who was not born in any state and has not been attending a school in the United States for more than three full academic years (Section 3301). These latter students are referred to as immigrant or recent immigrant students throughout this report. 2 Through FY2005, a reservation was also made to provide continuation grants for competitive grants awarded prior to the enactment of the NCLBA. 3 For the purposes of this report, the term "state" includes the District of Columbia and the Commonwealth of Puerto Rico. 4 More specifically, Section 1111(b)(7) requires states to assess the English language skills of students with limited English proficiency on an annual basis. 5 Testimony provided by Cornelia M. Ashby to the House of Representatives, Committee on Education and Labor, Subcommittee on Early Childhood, Elementary and Secondary Education. (March 23, 2007). Impact of NCLB on English Language Learners. (Hereafter referred to as Ashby testimony.) CRS-3 For example, ED noted that for the 2004-2005 school year, not every state provided data, and some data included only partial student counts.6 American Community Survey Data ED obtains the relevant ACS data from the U.S. Census Bureau. The LEP student count is based on the population aged 5 to 21 who reported speaking a language other than English at home and speaking English less than "very well." The number of immigrant students is based on the number of individuals aged 3 to 21 who reported entering the United States during the two years prior to the survey or during the survey year. For example, for the 2005 ACS, individuals entering the country in 2003 or later were counted as recent immigrant students. According to GAO, these questions were developed for the 1980 census to "obtain information needed about current language use and limited English language proficiency as a result of legislation such as the Civil Rights Act of 1964, the Bilingual Education Act, and the Voting Rights Act."7 These questions have not been modified since their inception. Thus, they were not designed specifically for Title III purposes. Tables 1 and 2 provide the ACS counts for LEP and immigrant students, respectively, used by ED in determining state grant amounts for FY2005, FY2006, and FY2007. Few states had relatively stable LEP student counts from 2002 to 2003 and 2003 to 2004 (Table 1). It was more common for states to experience substantial increases or decreases in their number or percentage of LEP students across the various ACS administrations. For example, the number of LEP students in Arizona declined by over 16,000 students from 2002 to 2003 but increased by nearly 21,000 students from 2004 to 2005, while the number of LEP students in Texas decreased by almost 58,000 students from 2002 to 2003 and increased by almost 25,000 students from 2003 to 2004. In other states, the change in the estimated number of LEP students may have been relatively small, but because some states serve small numbers of LEP students, changes in their student counts can result in large percentage changes. For example, the number of LEP students in Arkansas increased by about 8,000 students from 2002 to 2003 and decreased by about 5,000 students from 2003 to 2004, resulting in percentage changes in student counts of 59.9% and 21.6%, respectively. Although the number of immigrant students identified through the ACS is smaller then the number of LEP students, similar patterns were found in the immigrant student counts across years (Table 2). 6 U.S. Government Accountability Office. (2006). No Child Left Behind Act: Education's Data Improvement Efforts Could Strengthen the Basis for Distributing Title III Funds (GAO-07-140). Available online at [http://www.gao.gov]. (Hereafter referred to as GAO, Basis for Distributing Title III Funds.) 7 Ibid., p. 10. CRS-4 Table 1. LEP Student Counts from the 2003, 2004, and 2005 American Community Surveys Change from Change from ACS 2003 ACS 2004 ACS 2005 2003 to 2004 2004 to 2005 (FY2005 (FY2006 (FY2007 State Grants) Grants) Grants) Number Percent Number Percent Alabama 15,225 14,970 18,745 -255 -1.7% 3,775 25.2% Alaska 5,500 5,090 4,225 -410 -7.5% -865 -17.0% Arizona 117,530 101,140 121,895 -16,390 -13.9% 20,755 20.5% Arkansas 13,635 21,800 17,095 8,165 59.9% -4,705 -21.6% California 1,050,180 1,075,825 1,097,205 25,645 2.4% 21,380 2.0% Colorado 66,865 60,430 61,675 -6,435 -9.6% 1,245 2.1% Connecticut 28,080 33,020 33,165 4,940 17.6% 145 0.4% Delaware 6,030 7,015 8,355 985 16.3% 1,340 19.1% District of Columbia 5,835 2,950 3,490 -2,885 -49.4% 540 18.3% Florida 231,710 235,830 234,505 4,120 1.8% -1,325 -0.6% Georgia 93,155 78,495 85,275 -14,660 -15.7% 6,780 8.6% Hawaii 10,565 12,945 14,230 2,380 22.5% 1,285 9.9% Idaho 12,485 12,550 9,860 65 0.5% -2,690 -21.4% Illinois 176,630 182,210 182,730 5,580 3.2% 520 0.3% Indiana 57,500 70,380 40,740 12,880 22.4% -29,640 -42.1% Iowa 17,370 12,900 16,015 -4,470 -25.7% 3,115 24.1% Kansas 15,965 17,160 21,115 1,195 7.5% 3,955 23.0% Kentucky 16,565 17,580 17,160 1,015 6.1% -420 -2.4% Louisiana 18,740 15,235 14,165 -3,505 -18.7% -1,070 -7.0% Maine 2,590 3,865 3,535 1,275 49.2% -330 -8.5% Maryland 38,640 39,900 47,550 1,260 3.3% 7,650 19.2% Massachusetts 77,685 59,785 64,815 -17,900 -23.0% 5,030 8.4% Michigan 72,320 49,255 62,675 -23,065 -31.9% 13,420 27.2% Minnesota 44,530 48,180 39,575 3,650 8.2% -8,605 -17.9% Mississippi 7,410 4,775 7,870 -2,635 -35.6% 3,095 64.8% Missouri 28,600 19,950 21,765 -8,650 -30.2% 1,815 9.1% Montana 1,515 2,920 2,185 1,405 92.7% -735 -25.2% Nebraska 14,100 12,460 14,935 -1,640 -11.6% 2,475 19.9% Nevada 48,730 58,010 38,540 9,280 19.0% -19,470 -33.6% New Hampshire 5,905 5,195 5,000 -710 -12.0% -195 -3.8% New Jersey 121,360 100,680 107,955 -20,680 -17.0% 7,275 7.2% New Mexico 40,205 27,690 28,805 -12,515 -31.1% 1,115 4.0% New York 388,795 332,065 275,230 -56,730 -14.6% -56,835 -17.1% North Carolina 65,600 73,710 70,970 8,110 12.4% -2,740 -3.7% North Dakota 2,190 2,095 1,700 -95 -4.3% -395 -18.9% Ohio 42,860 48,885 48,005 6,025 14.1% -880 -1.8% Oklahoma 31,570 20,575 21,085 -10,995 -34.8% 510 2.5% Oregon 37,755 43,100 49,910 5,345 14.2% 6,810 15.8% Pennsylvania 61,600 75,935 74,245 14,335 23.3% -1,690 -2.2% Rhode Island 17,865 11,875 12,130 -5,990 -33.5% 255 2.1% South Carolina 16,155 15,525 22,940 -630 -3.9% 7,415 47.8% South Dakota 4,055 2,855 4,065 -1,200 -29.6% 1,210 42.4% Tennessee 25,595 33,180 28,635 7,585 29.6% -4,545 -13.7% Texas 603,105 545,330 570,145 -57,775 -9.6% 24,815 4.6% Utah 19,215 20,590 21,050 1,375 7.2% 460 2.2% Vermont 1,585 1,140 1,900 -445 -28.1% 760 66.7% Virginia 53,935 52,640 57,440 -1,295 -2.4% 4,800 9.1% Washington 58,840 59,350 78,270 510 0.9% 18,920 31.9% West Virginia 2,465 2,320 3,250 -145 -5.9% 930 40.1% Wisconsin 44,275 39,665 38,855 -4,610 -10.4% -810 -2.0% CRS-5 Change from Change from ACS 2003 ACS 2004 ACS 2005 2003 to 2004 2004 to 2005 (FY2005 (FY2006 (FY2007 State Grants) Grants) Grants) Number Percent Number Percent Wyoming 1,780 1,885 2,130 105 5.9% 245 13.0% Total 3,942,395 3,792,910 3,828,805 -149,485 -3.8% 35,895 0.9% Source: Table prepared by CRS, June 2007, based on data provided by the U.S. Department of Education (ED), Budget Service. Note: The American Community Survey (ACS) is administered by the U.S. Census Bureau. The Census Bureau provides ED with specific data runs from the most recent ACS to enable ED to calculate Title III grants. Table 2. Immigrant Student Counts from the 2003, 2004, and 2005 American Community Survey Change from Change from ACS 2003 ACS 2004 ACS 2005 2003 to 2004 2004 to 2005 (FY2005 (FY2006 (FY2007 State Grants) Grants) Grants) Number Percent Number Percent Alabama 10,500 10,195 7,710 -305 -2.9% -2,485 -24.4% Alaska 1,705 2,415 965 710 41.6% -1,450 -60.0% Arizona 20,670 35,400 35,660 14,730 71.3% 260 0.7% Arkansas 3,485 6,545 4,680 3,060 87.8% -1,865 -28.5% California 238,495 229,805 251,275 -8,690 -3.6% 21,470 9.3% Colorado 18,920 14,840 16,835 -4,080 -21.6% 1,995 13.4% Connecticut 10,255 10,725 10,670 470 4.6% -55 -0.5% Delaware 1,525 2,520 2,495 995 65.2% -25 -1.0% District of Columbia 2,125 1,665 1,285 -460 -21.6% -380 -22.8% Florida 105,365 100,595 93,535 -4,770 -4.5% -7,060 -7.0% Georgia 21,285 25,045 36,945 3,760 17.7% 11,900 47.5% Hawaii 3,635 5,145 6,645 1,510 41.5% 1,500 29.2% Idaho 5,730 3,360 5,010 -2,370 -41.4% 1,650 49.1% Illinois 36,390 43,520 35,965 7,130 19.6% -7,555 -17.4% Indiana 8,270 12,940 11,985 4,670 56.5% -955 -7.4% Iowa 7,755 2,910 4,150 -4,845 -62.5% 1,240 42.6% Kansas 4,890 4,305 6,035 -585 -12.0% 1,730 40.2% Kentucky 4,160 6,965 5,275 2,805 67.4% -1,690 -24.3% Louisiana 9,955 3,105 3,185 -6,850 -68.8% 80 2.6% Maine 735 1,000 995 265 36.1% -5 -0.5% Maryland 18,895 18,755 26,765 -140 -0.7% 8,010 42.7% Massachusetts 19,355 17,520 23,935 -1,835 -9.5% 6,415 36.6% Michigan 27,330 18,330 20,640 -9,000 -32.9% 2,310 12.6% Minnesota 12,340 7,180 14,420 -5,160 -41.8% 7,240 100.8% Mississippi 1,350 1,035 2,695 -315 -23.3% 1,660 160.4% Missouri 10,585 4,300 7,315 -6,285 -59.4% 3,015 70.1% Montana 440 980 465 540 122.7% -515 -52.6% Nebraska 4,390 4,280 4,130 -110 -2.5% -150 -3.5% Nevada 10,410 9,690 9,445 -720 -6.9% -245 -2.5% New Hampshire 3,235 1,255 1,155 -1,980 -61.2% -100 -8.0% New Jersey 53,080 31,035 38,670 -22,045 -41.5% 7,635 24.6% New Mexico 5,800 3,900 5,720 -1,900 -32.8% 1,820 46.7% New York 75,560 87,320 83,310 11,760 15.6% -4,010 -4.6% North Carolina 20,495 25,145 27,890 4,650 22.7% 2,745 10.9% North Dakota 695 770 415 75 10.8% -355 -46.1% Ohio 13,805 14,070 13,525 265 1.9% -545 -3.9% Oklahoma 10,450 9,740 5,935 -710 -6.8% -3,805 -39.1% Oregon 7,900 10,845 10,925 2,945 37.3% 80 0.7% CRS-6 Change from Change from ACS 2003 ACS 2004 ACS 2005 2003 to 2004 2004 to 2005 (FY2005 (FY2006 (FY2007 State Grants) Grants) Grants) Number Percent Number Percent Pennsylvania 15,835 13,545 16,150 -2,290 -14.5% 2,605 19.2% Rhode Island 2,570 3,420 4,610 850 33.1% 1,190 34.8% South Carolina 6,195 4,080 11,865 -2,115 -34.1% 7,785 190.8% South Dakota 380 790 1,835 410 107.9% 1,045 132.3% Tennessee 13,740 10,160 9,800 -3,580 -26.1% -360 -3.5% Texas 106,445 126,650 130,990 20,205 19.0% 4,340 3.4% Utah 5,695 8,155 7,410 2,460 43.2% -745 -9.1% Vermont 870 300 645 -570 -65.5% 345 115.0% Virginia 25,800 24,835 25,835 -965 -3.7% 1,000 4.0% Washington 14,835 21,350 24,375 6,515 43.9% 3,025 14.2% West Virginia 2,845 235 200 -2,610 -91.7% -35 -14.9% Wisconsin 8,880 9,320 8,805 440 5.0% -515 -5.5% Wyoming 310 765 1,085 455 146.8% 320 41.8% Total 1,016,365 1,012,755 1,082,260 -3,610 -0.4% 69,505 6.9% Source: Table prepared by CRS, June 2007, based on data provided by the U.S. Department of Education (ED), Budget Service. Note: The American Community Survey (ACS) is administered by the U.S. Census Bureau. The Census Bureau provides ED with specific data runs from the most recent ACS to enable ED to calculate Title III grants. State-Reported LEP Student Counts There are several potential sources of state-reported data and three types of state LEP student counts: (1) total number of LEP students, (2) number of LEP students receiving services (Title III or non-Title III), and (3) number of LEP students being served in Title III. The National Clearinghouse for English Language Acquisition and Language Instruction Educational Programs (NCELA) uses data provided by states to ED in their Consolidated State Performance Reports (CSPRs) to produce an annual state-by-state count of the number of LEP students enrolled.8 The CSPRs collect data on the total number of students identified as LEP. The most recent data available are for the 2004-2005 school year. Missing data and data discrepancies were resolved by NCELA through telephone calls to the relevant states. The Common Core of Data (CCD) collects data on the total number of students receiving LEP services.9 This is not limited to Title III services only. Rather, it includes students served in appropriate programs of language assistance. This is somewhat different than the data available from NCELA, as the CCD count does not include students who are identified as LEP but are not receiving services.10 The most recent data available from the CCD are for the 2004-2005 school year. 8 Information about how NCELA produces LEP student counts was provided by NCELA staff members, Dr. Judith Wilde and Suzanne Abdelrahim. 9 The CCD uses the term English language learners (ELL) rather the LEP. The term LEP was used until the 2001-2002 school year. For consistency, the term LEP is used throughout this report. 10 For example, students may not receive LEP services if their parents do not want them to participate. CRS-7 A third source of data is the biennial report on Title III performance. The first report was published by ED in 2005 and covered 2002-2004. It included data on the number of LEP students served in Title III programs during the 2002-2003 and 2003- 2004 school years. It did not, however, report on the total number of LEP students in the state or the total number of LEP students receiving Title III and non-Title III services. According to staff at NCELA, the second biennial report, expected later this year, will include counts of both the number of LEP students served in Title III programs and the total number of LEP students in the state. They also indicated that much of the data included in the forthcoming second biennial report has been drawn from the annual CSPRs. LEP student counts from the NCELA data and CCD data are compared for the 2002-2003, 2003-2004, and 2004-2005 school years (Table 3). The NCELA data produced using the CSPRs are more complete than the data available from the CCD, as state data are missing for several states in the CCD data.11 As the NCELA LEP student count is, in theory, a more comprehensive count than the CCD LEP student count, it would be expected that the NCELA counts would be higher than the CCD counts but not substantially higher, as it is not expected that many parents would choose for their children not to receive services. The data on Table 3 do not consistently support these theories. For example, for the 2004-2005 school year, the NCELA count is actually lower than the CCD count in 18 states. For Indiana, Louisiana, Oklahoma, Vermont, and West Virginia, the NCELA count is at least 20% lower than the CCD count. Although it was expected that the NCELA count might be somewhat higher than the CCD count, in Florida, Mississippi, New Hampshire, North Dakota, Rhode Island, South Carolina, South Dakota, Vermont, and Wisconsin, the NCELA count is at least 20% higher than the CCD count. Similar issues exist with the data from the 2002-2003 and 2003-2004 school years. It is the exception, rather than the rule, that the NCELA data and the CCD data match or differ by a relatively small number of students. In addition, differences between the NCELA data and the CCD data change for some states from year to year with the NCELA count being higher in some years, the CCD count being higher in other years, and the magnitude of the differences between the counts changing from year to year. This raises questions about how states are conducting LEP student counts, whether these counts are being conducted consistently within a state and from year to year, and which students are actually being included in the counts. 11 For several states, data are either not available, or data were missing for more than 20% of schools or districts in a state, so the data were not publicly reported. CRS-8 Table 3. Estimated State LEP Student Counts Based on Data Available from the Common Core of Data and the NCELA: 2002-2003, 2003-2004, and 2004-2005 A B C D E F G H I J 2002-2003 School Year 2003-2004 School Year 2004-2005 School Year Difference Difference Difference CCD NCELA (Col C - CCD NCELA (Col F - CCD NCELA (Col I - State Data Data Col B) Data Data Col E) Data Data Col H) Alabama 10,568 10,566 -2 10,825 13,312 2,487 14,801 15,295 494 Alaska 16,378 20,272 3,894 19,877 21,533 1,656 21,533 20,140 -1,393 Arizona 143,744 149,354 5,610 155,840 144,145 -11,695 194,171 155,789 -38,382 Arkansas 15,146 14,838 -308 17,174 15,581 -1,593 18,647 17,384 -1,263 California 1,599,542 1,599,542 0 1,598,366 1,598,535 169 1,585,647 1,591,525 5,878 Colorado 86,128 86,129 1 97,043 91,751 -5,292 90,372 90,391 19 Connecticut 22,651 22,547 -104 25,959 25,867 -92 27,931 27,580 -351 Delaware 3,449 3,523 74 3,956 4,246 290 4,858 5,094 236 District of Columbia 5,798 5,363 -435 5,727 5,201 -526 5,657 4,771 -886 Florida 203,712 292,077 88,365 196,037 282,066 86,029 214,562 299,346 84,784 Georgia 70,464 59,840 -10,624 65,876 59,126 -6,750 60,334 50,381 -9,953 Hawaii 12,853 12,853 0 12,850 12,850 0 17,017 18,376 1,359 Idaho 18,747 19,753 1,006 19,649 20,541 892 20,987 17,649 -3,338 Illinois 168,727 169,414 687 ! 161,700 ! ! 192,764 ! Indiana 42,629 22,584 -20,045 42,632 28,741 -13,891 51,212 31,956 -19,256 Iowa 13,961 13,961 0 15,238 15,238 0 14,606 14,421 -185 Kansas 17,942 25,006 7,064 22,399 25,504 3,105 26,041 23,512 -2,529 Kentucky 6,343 6,017 -326 8,446 8,446 0 10,471 11,181 710 Louisiana 11,108 6,854 -4,254 12,175 7,546 -4,629 12,979 7,990 -4,989 Maine 2,632 3,006 374 2,852 3,179 327 2,868 2,896 28 Maryland 27,311 27,422 111 27,695 27,849 154 21,709 24,811 3,102 Massachusetts 51,622 51,622 0 49,297 49,297 0 49,773 49,923 150 Michigan ! 60,479 ! 62,025 62,265 240 62,778 64,345 1,567 Minnesota 51,275 52,244 969 53,507 54,878 1,371 56,976 56,829 -147 Mississippi 2,250 2,916 666 2,916 4,681 1,765 3,365 4,152 787 Missouri 13,121 13,121 0 14,855 14,855 0 ! 15,403 ! Montana 6,642 7,043 401 6,668 6,948 280 6,716 6,911 195 Nebraska 13,803 13,803 0 15,586 15,586 0 16,124 16,124 0 Nevada 58,753 53,492 -5,261 69,896 58,753 -11,143 71,557 72,117 560 New Hampshire 3,270 3,270 0 2,755 2,755 0 2,569 3,235 666 New Jersey 57,548 57,245 -303 58,349 66,451 8,102 ! 61,287 ! New Mexico 65,317 65,317 0 54,528 54,528 0 62,386 70,926 8,540 New York 178,909 302,961 124,052 ! 191,992 ! ! 203,583 ! North Carolina 59,849 60,149 300 60,967 70,937 9,970 68,381 70,288 1,907 North Dakota 883 6,176 5,293 1,638 6,500 4,862 2,033 4,749 2,716 Ohio 25,782 20,778 -5,004 23,368 23,302 -66 27,499 25,518 -1,981 Oklahoma 40,192 36,508 -3,684 40,042 33,266 -6,776 44,454 33,508 -10,946 Oregon 52,331 52,588 257 64,618 61,695 -2,923 64,676 59,908 -4,768 Pennsylvania ! 38,288 ! ! 41,606 ! ! 39,847 ! Rhode Island 10,087 11,600 1,513 9,723 9,645 -78 9,001 10,921 1,920 South Carolina 7,467 8,239 772 10,653 12,653 2,000 12,528 15,396 2,868 South Dakota 4,524 3,361 -1,163 4,477 3,433 -1,044 4,194 5,847 1,653 Tennessee ! 14,953 ! ! 19,352 ! ! 19,355 ! Texas 630,686 630,148 -538 661,052 660,707 -345 684,583 684,007 -576 Utah 43,299 46,342 3,043 49,556 46,521 -3,035 45,027 56,319 11,292 Vermont 1,057 1,052 -5 1,992 1,017 -975 1,990 1,393 -597 Virginia 49,845 49,840 -5 60,301 60,306 5 66,970 67,933 963 Washington 70,431 66,038 -4,393 58,523 69,323 10,800 75,103 75,678 575 West Virginia 1,281 2,103 822 1,477 1,594 117 1,774 1,236 -538 Wisconsin 25,764 34,203 8,439 26,424 35,770 9,346 26,616 35,871 9,255 Wyoming 3,519 3,206 -313 3,475 3,429 -46 3,593 3,742 149 Total 4,029,340 4,340,006 310,666 3,829,284 4,317,002 487,718 3,887,069 4,459,603 572,534 CRS-9 Source: Table prepared by CRS, June 2007. The "NCELA data" were provided by the National Clearinghouse on English Language Acquisition and Language Instruction Educational Programs based on an analysis of data reported by states on Consolidated State Performance Reports. The "CCD data" were collected through the Common Core of Data by the National Center for Education Statistics (NCES) at the U.S. Department of Education and reported in a series of annual reports (NCES 2005-314, NCES 2006-307, and NCES 2007-309). Note: While a total is shown for the CCD data for the 2004-2005 school year, a total was not included in the NCES report, as data were missing for more than 15% of all schools or districts nationally. !: Data were not available or data were missing from more than 20% of schools or districts within a state. Although not shown in Table 3, data from the NCELA and CCD from 2003- 2004 were compared with data reported in the biennial report for 2003-2004. As previously discussed, the NCELA data, in theory, provide the most comprehensive count of LEP students, including all identified LEP students. The CCD data, in theory, provide the second most comprehensive count of LEP students by including all students receiving LEP services. The biennial report data, in theory, are the least comprehensive of the three sources of data as they include only the number of LEP students receiving Title III services. A brief examination of data for Alabama, Alaska, Arizona, Arkansas, and California revealed that the biennial LEP student count was higher than the CCD count in Alabama and Arizona. It was also higher than the NCELA count in Arizona. This raises further questions about how LEP students are being counted at the state level and whether state counts are a reliable basis upon which to make state grant allocations. State-Reported Immigrant Student Counts Data sources for immigrant student counts are more limited than those available for LEP students. The primary source of this information is the Title III biennial report. States are required to report on the number of immigrant students enrolled and the number of immigrant students served in Title III programs. The most recent biennial report includes these counts for the 2002-2003 and 2003-2004 school years. Neither NCELA nor the CCD produces immigrant student counts. Table 4 provides the immigrant student counts from the biennial report for the 2002-2003 and 2003- 2004 school years, the most recent years for which data are available. CRS-10 Table 4. Immigrant Student Counts Based on State-Reported Data in Title III Biennial Reports: 2002-2003 and 2003-2004 A B C D E Number of Number of Immigrant Children Immigrant Children Difference in the and Youth During and Youth During Number of Immigrant 2002-2003 2003-2004 Children and Youth Percent State School Year School Year (Col C - Col B) Change Alabama 5,355 4,166 -1,189 -22.2% Alaska 1,818 1,163 -655 -36.0% Arizona 40,721 34,074 -6,647 -16.3% Arkansas 4,626 4,696 70 1.5% California 254,450 269,939 15,489 6.1% Colorado 10,486 15,642 5,156 49.2% Connecticut 14,977 16,398 1,421 9.5% Delaware 1,665 1,327 -338 -20.3% District of Columbia 1,631 1,376 -255 -15.6% Florida 169,819 158,168 -11,651 -6.9% Georgia 38,919 40,150 1,231 3.2% Hawaii 4,678 5,242 564 12.1% Idaho ! 1,440 ! ! Illinois 61,139 65,629 4,490 7.3% Indiana 10,686 11,130 444 4.2% Iowa 3,925 3,284 -641 -16.3% Kansas 9,184 7,924 -1,260 -13.7% Kentucky 3,397 5,199 1,802 53.0% Louisiana 3,848 3,683 -165 -4.3% Maine 1,129 1,280 151 13.4% Maryland 18,237 18,156 -81 -0.4% Massachusetts 21,395 25,740 4,345 20.3% Michigan 12,236 12,530 294 2.4% Minnesota 15,414 16,236 822 5.3% Mississippi 952 1,316 364 38.2% Missouri 8,020 7,518 -502 -6.3% Montana 273 348 75 27.5% Nebraska 5,698 5,635 -63 -1.1% Nevada 12,565 16,479 3,914 31.2% New Hampshire 1,991 1,200 -791 -39.7% New Jersey 54,185 45,814 -8,371 -15.4% New Mexico 9,631 8,132 -1,499 -15.6% New York 123,948 116,822 -7,126 -5.7% North Carolina 31,183 29,232 -1,951 -6.3% North Dakota 1,007 1,009 2 0.2% Ohio 12,389 11,687 -702 -5.7% Oklahoma 9,466 7,622 -1,844 -19.5% Oregon 7,730 7,455 -275 -3.6% Pennsylvania 15,519 16,138 619 4.0% Rhode Island 3,322 2,900 -422 -12.7% South Carolina 6,254 6,716 462 7.4% South Dakota 909 1,020 111 12.2% Tennessee 19,569 16,325 -3,244 -16.6% Texas 121,064 116,818 -4,246 -3.5% Utah 14,195 17,145 2,950 20.8% Vermont 598 567 -31 -5.2% Virginia 23,432 21,440 -1,992 -8.5% Washington 21,196 24,997 3,801 17.9% CRS-11 A B C D E Number of Number of Immigrant Children Immigrant Children Difference in the and Youth During and Youth During Number of Immigrant 2002-2003 2003-2004 Children and Youth Percent State School Year School Year (Col C - Col B) Change West Virginia 178 175 -3 -1.7% Wisconsin 7,548 6,608 -940 -12.5% Wyoming 191 191 0 0.0% Total 1,222,748 1,215,881 -6,867 -0.6% Source: Table prepared by CRS, June 2007, based on data provided by states in their Title III biennial reports (Biennial Evaluation Report to Congress on the Implementation of the State Formula Grant Program, 2002-2004). !: Data either not reported or not available. ACS Data Compared with State-Reported Data This section makes direct comparisons between the most recent ACS data and the most recent state data. It begins with a discussion of data limitations in making these comparisons. This discussion is followed by a detailed analysis of differences in LEP and immigrant student counts between the two types of data. The section concludes with an analysis of estimated FY2007 state grants using both data sources and how these grants would differ based on the underlying data used for the calculation. Data Limitations It should be noted that comparisons of student counts have been conducted with state data from two different school years. As the LEP student count accounts for 80% of a state's total grant amount, it was important to have the most recent data available and, if possible, to be using a school year comparable to the year in which the ACS data were collected. The LEP student counts produced by NCELA for the 2004-2005 school year met both these criteria. And, unlike the CCD data for the same school year, they presumably include a more nearly complete count of total LEP student enrollment, as the data were confirmed with state officials as needed. Although one of the primary purposes of this request was to compare the ACS data with state data, it should be noted that the NCELA and ACS data are collected from different respondents using different questions. Again, the NCELA data include the population of students identified as LEP, while the ACS is a sample survey conducted with native and non-native English speaking individuals. As previously mentioned, data sources for immigrant student counts are more limited, so the latest available data were from the 2003-2004 school year. Thus, in addition to the aforementioned problems of collecting data from different respondents using different questions, the state-reported immigrant data were taken from a different year than the 2005 ACS data. These caveats must be taken into account when examining student counts and estimated state grants based on these data. The estimated grant amounts discussed CRS-12 below are only rough estimates of what states might receive if state data were relied upon to make grants. Student Count Data Comparisons As shown in Tables 5 and 6, there are some substantial differences in LEP and immigrant student counts when the 2005 ACS data are compared with the 2004-2005 LEP student counts available from NCELA and the 2003-2004 immigrant student counts available from the biennial report. With respect to LEP student counts, for example, the state data indicate that there are almost 500,000 more LEP students in California than indicated by the ACS data. Arizona, Colorado, Florida, Nevada, New Mexico, Texas, and Utah each reported at least 25,000 more LEP students than accounted for by the 2005 ACS. At the same time, if state LEP student counts were used instead of ACS LEP student counts, Georgia, New Jersey, New York, and Pennsylvania would have their student counts reduced by 25,000 students or more. If the same data are examined based on the percentage change in student counts if state data were used instead of ACS data, Alaska would experience the largest percentage increase in LEP students (376.7%), followed by Montana (216.3%), North Dakota (179.4%), Utah (167.5%), and New Mexico (146.2%). Although California would experience the largest increase in the number of LEP students, this change would result in a 45.1% increase in LEP student enrollment.12 West Virginia would experience the largest percentage decrease in enrollment (62.0%), followed by Mississippi (47.2%), Ohio (46.8%), and Pennsylvania (46.3%). If state data were used in lieu of ACS data for immigrant counts, Florida would experience the largest increase in the number of immigrant students, followed by New York, Illinois, and California (Table 6). Overall, increases in immigrant student counts would range from 16 students in Arkansas to 64,633 students in Florida. The largest decreases in the number of immigrant students would occur in Texas, followed by Maryland and Michigan. Overall, decreases in the number of immigrant students would range from 12 students in Pennsylvania to 14,100 students in Texas. In terms of percentage change in student counts, the largest increase in immigrant students would occur in North Dakota (143.1%), followed by Utah (131.4%) and Illinois (82.5%). The greatest decreases would be experienced by Wyoming (82.4%), followed by Idaho (71.3%) and Mississippi (51.2%). 12 The percentage change in the number of students is calculated relative to the number of LEP students identified on the ACS. As California has the largest number of students based on the ACS counts, having the largest increase in the number of students based on the state data is not a large enough increase relative to California's initial LEP student count to result in the largest percentage increase among the states. CRS-13 Table 5. Comparison of Estimated Limited English Proficient Student Counts from the 2005 American Community Survey and 2004-2005 State-Reported Data 2005 2004-2005 Difference Percent State ACS State Data (State - ACS) Difference Alabama 18,745 15,295 -3,450 -18.4% Alaska 4,225 20,140 15,915 376.7% Arizona 121,895 155,789 33,894 27.8% Arkansas 17,095 17,384 289 1.7% California 1,097,205 1,591,525 494,320 45.1% Colorado 61,675 90,391 28,716 46.6% Connecticut 33,165 27,580 -5,585 -16.8% Delaware 8,355 5,094 -3,261 -39.0% District of Columbia 3,490 4,771 1,281 36.7% Florida 234,505 299,346 64,841 27.7% Georgia 85,275 50,381 -34,894 -40.9% Hawaii 14,230 18,376 4,146 29.1% Idaho 9,860 17,649 7,789 79.0% Illinois 182,730 192,764 10,034 5.5% Indiana 40,740 31,956 -8,784 -21.6% Iowa 16,015 14,421 -1,594 -10.0% Kansas 21,115 23,512 2,397 11.4% Kentucky 17,160 11,181 -5,979 -34.8% Louisiana 14,165 7,990 -6,175 -43.6% Maine 3,535 2,896 -639 -18.1% Maryland 47,550 24,811 -22,739 -47.8% Massachusetts 64,815 49,923 -14,892 -23.0% Michigan 62,675 64,345 1,670 2.7% Minnesota 39,575 56,829 17,254 43.6% Mississippi 7,870 4,152 -3,718 -47.2% Missouri 21,765 15,403 -6,362 -29.2% Montana 2,185 6,911 4,726 216.3% Nebraska 14,935 16,124 1,189 8.0% Nevada 38,540 72,117 33,577 87.1% New Hampshire 5,000 3,235 -1,765 -35.3% New Jersey 107,955 61,287 -46,668 -43.2% New Mexico 28,805 70,926 42,121 146.2% New York 275,230 203,583 -71,647 -26.0% North Carolina 70,970 70,288 -682 -1.0% North Dakota 1,700 4,749 3,049 179.4% Ohio 48,005 25,518 -22,487 -46.8% Oklahoma 21,085 33,508 12,423 58.9% Oregon 49,910 59,908 9,998 20.0% Pennsylvania 74,245 39,847 -34,398 -46.3% Rhode Island 12,130 10,921 -1,209 -10.0% South Carolina 22,940 15,396 -7,544 -32.9% South Dakota 4,065 5,847 1,782 43.8% Tennessee 28,635 19,355 -9,280 -32.4% Texas 570,145 684,007 113,862 20.0% Utah 21,050 56,319 35,269 167.5% Vermont 1,900 1,393 -507 -26.7% Virginia 57,440 67,933 10,493 18.3% Washington 78,270 75,678 -2,592 -3.3% West Virginia 3,250 1,236 -2,014 -62.0% Wisconsin 38,855 35,871 -2,984 -7.7% CRS-14 2005 2004-2005 Difference Percent State ACS State Data (State - ACS) Difference Wyoming 2,130 3,742 1,612 75.7% Total 3,828,805 4,459,603 630,798 16.5% Source: Table prepared by CRS, June 2007, based on data provided by the U.S. Department of Education (ED), Budget Service. The 2004-2005 student counts were provided by the National Clearinghouse for English Language Acquisition and Language Instruction Educational Programs (NCELA), based on an analysis of data reported by states on their Consolidated State Performance Reports. Note: The American Community Survey (ACS) is administered by the U.S. Census Bureau. The Census Bureau provides ED with specific data runs from the most recent ACS to enable ED to calculate Title III grants. Table 6. Comparison of Estimated Immigrant Student Counts from the 2005 American Community Survey and 2003-2004 State-Reported Data 2005 2003-2004 Difference Percent State ACS State Data (State - ACS) Difference Alabama 7,710 4,166 -3,544 -46.0% Alaska 965 1,163 198 20.5% Arizona 35,660 34,074 -1,586 -4.4% Arkansas 4,680 4,696 16 0.3% California 251,275 269,939 18,664 7.4% Colorado 16,835 15,642 -1,193 -7.1% Connecticut 10,670 16,398 5,728 53.7% Delaware 2,495 1,327 -1,168 -46.8% District of Columbia 1,285 1,376 91 7.1% Florida 93,535 158,168 64,633 69.1% Georgia 36,945 40,150 3,205 8.7% Hawaii 6,645 5,242 -1,403 -21.1% Idaho 5,010 1,440 -3,570 -71.3% Illinois 35,965 65,629 29,664 82.5% Indiana 11,985 11,130 -855 -7.1% Iowa 4,150 3,284 -866 -20.9% Kansas 6,035 7,924 1,889 31.3% Kentucky 5,275 5,199 -76 -1.4% Louisiana 3,185 3,683 498 15.6% Maine 995 1,280 285 28.6% Maryland 26,765 18,156 -8,609 -32.2% Massachusetts 23,935 25,740 1,805 7.5% Michigan 20,640 12,530 -8,110 -39.3% Minnesota 14,420 16,236 1,816 12.6% Mississippi 2,695 1,316 -1,379 -51.2% Missouri 7,315 7,518 203 2.8% Montana 465 348 -117 -25.2% Nebraska 4,130 5,635 1,505 36.4% Nevada 9,445 16,479 7,034 74.5% New Hampshire 1,155 1,200 45 3.9% New Jersey 38,670 45,814 7,144 18.5% New Mexico 5,720 8,132 2,412 42.2% New York 83,310 116,822 33,512 40.2% North Carolina 27,890 29,232 1,342 4.8% CRS-15 2005 2003-2004 Difference Percent State ACS State Data (State - ACS) Difference North Dakota 415 1,009 594 143.1% Ohio 13,525 11,687 -1,838 -13.6% Oklahoma 5,935 7,622 1,687 28.4% Oregon 10,925 7,455 -3,470 -31.8% Pennsylvania 16,150 16,138 -12 -0.1% Rhode Island 4,610 2,900 -1,710 -37.1% South Carolina 11,865 6,716 -5,149 -43.4% South Dakota 1,835 1,020 -815 -44.4% Tennessee 9,800 16,325 6,525 66.6% Texas 130,990 116,818 -14,172 -10.8% Utah 7,410 17,145 9,735 131.4% Vermont 645 567 -78 -12.1% Virginia 25,835 21,440 -4,395 -17.0% Washington 24,375 24,997 622 2.6% West Virginia 200 175 -25 -12.5% Wisconsin 8,805 6,608 -2,197 -25.0% Wyoming 1,085 191 -894 -82.4% Total 1,082,260 1,215,881 133,621 12.3% Source: Table prepared by CRS, June 2007, based on data provided by the U.S. Department of Education (ED), Budget Service. The 2003-2004 immigrant student counts are based on data provided by states in their Title III biennial reports (Biennial Evaluation Report to Congress on the Implementation of the State Formula Grant Program, 2002-2004.) Note: The American Community Survey (ACS) is administered by the U.S. Census Bureau. The Census Bureau provides ED with specific data runs from the most recent ACS to enable ED to calculate Title III grants. Estimated FY2007 State Grants An analysis of the differences in estimated FY2007 grant amounts if the aforementioned state data, rather than the 2005 ACS data, were used to determine grant amounts revealed that grant amounts would change in most states, in some cases increasing or decreasing by substantial amounts. For example, if state data, rather than ACS data, had been used to calculate the FY2007 grant amounts, California's grant amount would have increased by $34.2 million or 20.2% (Table 7). Other states that would have experienced substantial increases in their FY2007 grant amounts include Florida ($7.8 million or 19.2%), Utah ($4.4 million or 123.7%), and New Mexico ($4.3 million or 100.2%). One interesting trend to note is the general reduction in state grant amounts that would occur in Northeast states (e.g., Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, and Rhode Island), if state data were used to calculate grant amounts. Overall, increases would have ranged from $20,000 in South Dakota to $34.2 million in California. The largest loss of funds would have been experienced by New York ($10.9 million or 24.3%), followed by New Jersey ($7.0 million or 38.5%), Georgia ($5.7 million or 37.5%), and Pennsylvania ($5.4 million or 47.2%). Overall, the decreases would have ranged from $12,000 in Kansas to $10.9 million in New York. With respect to the percentage change in grant amount, Alaska would have received the largest increase (262.1%), and Mississippi would have experienced the largest decrease (55.2%). CRS-16 Table 7. Estimated FY2007 State Grants Based on the 2005 American Community Survey and State-Reported Data A B C D E $ Difference Between FY2007 Estimated Grants Based on FY2007 Estimated State Data Versus Grants FY2007 Grants ACS Data Percent State (2005 ACS) (State Data) (Col C - Col B) Difference Alabama 3,277,000 2,103,000 -1,174,000 -35.8% Alaska 651,000 2,356,000 1,705,000 261.9% Arizona 19,664,000 20,623,000 959,000 4.9% Arkansas 2,721,000 2,387,000 -334,000 -12.3% California 169,058,000 203,216,000 34,158,000 20.2% Colorado 9,812,000 11,571,000 1,759,000 17.9% Connecticut 5,460,000 4,641,000 -819,000 -15.0% Delaware 1,354,000 695,000 -659,000 -48.7% District of Columbia 593,000 663,000 70,000 11.8% Florida 40,669,000 48,478,000 7,809,000 19.2% Georgia 15,123,000 9,451,000 -5,672,000 -37.5% Hawaii 2,578,000 2,549,000 -29,000 -1.1% Idaho 1,833,000 2,105,000 272,000 14.8% Illinois 27,485,000 27,758,000 273,000 1.0% Indiana 6,580,000 4,626,000 -1,954,000 -29.7% Iowa 2,523,000 1,921,000 -602,000 -23.9% Kansas 3,390,000 3,378,000 -12,000 -0.4% Kentucky 2,797,000 1,743,000 -1,054,000 -37.7% Louisiana 2,176,000 1,243,000 -933,000 -42.9% Maine 566,000 500,000 -66,000 -11.7% Maryland 9,135,000 4,500,000 -4,635,000 -50.7% Massachusetts 11,022,000 8,024,000 -2,998,000 -27.2% Michigan 10,373,000 8,370,000 -2,003,000 -19.3% Minnesota 6,708,000 7,886,000 1,178,000 17.6% Mississippi 1,314,000 589,000 -725,000 -55.2% Missouri 3,619,000 2,435,000 -1,184,000 -32.7% Montana 500,000 804,000 304,000 60.8% Nebraska 2,382,000 2,336,000 -46,000 -1.9% Nevada 6,009,000 9,614,000 3,605,000 60.0% New Hampshire 772,000 500,000 -272,000 -35.2% New Jersey 18,222,000 11,208,000 -7,014,000 -38.5% New Mexico 4,338,000 8,684,000 4,346,000 100.2% New York 44,717,000 33,853,000 -10,864,000 -24.3% North Carolina 12,261,000 10,628,000 -1,633,000 -13.3% North Dakota 500,000 626,000 126,000 25.2% Ohio 7,685,000 3,961,000 -3,724,000 -48.5% Oklahoma 3,375,000 4,464,000 1,089,000 32.3% Oregon 7,633,000 7,391,000 -242,000 -3.2% Pennsylvania 11,343,000 5,984,000 -5,359,000 -47.2% Puerto Rico 3,086,000 3,086,000 0 0.0% Rhode Island 2,078,000 1,495,000 -583,000 -28.1% South Carolina 4,288,000 2,358,000 -1,930,000 -45.0% South Dakota 729,000 749,000 20,000 2.7% Tennessee 4,781,000 3,717,000 -1,064,000 -22.3% Texas 87,896,000 87,415,000 -481,000 -0.5% Utah 3,538,000 7,916,000 4,378,000 123.7% CRS-17 A B C D E $ Difference Between FY2007 Estimated Grants Based on FY2007 Estimated State Data Versus Grants FY2007 Grants ACS Data Percent State (2005 ACS) (State Data) (Col C - Col B) Difference Vermont 500,000 500,000 0 0.0% Virginia 10,295,000 9,621,000 -674,000 -6.5% Washington 12,795,000 10,824,000 -1,971,000 -15.4% West Virginia 500,000 500,000 0 0.0% Wisconsin 5,976,000 4,630,000 -1,346,000 -22.5% Wyoming 500,000 500,000 0 0.0% Total 617,177,000 617,177,000 0 0.0% Source: Table prepared by CRS, June 2007. Estimated FY2007 state grants based on the American Community Survey (ACS) were calculated by the U.S. Department of Education (ED), Budget Service. Estimated FY2007 state grants based on state-reported data were calculated by CRS using 2004-2005 limited English proficient student (LEP) counts available from the National Clearinghouse for English Language Acquisition and Language Instruction Educational Programs, and 2003-2004 immigrant student count data available based on data provided by states in their Title III biennial reports (Biennial Evaluation Report to Congress on the Implementation of the State Formula Grant Program, 2002-2004.) Note: All data sources used to make these calculations were the most recent data sources available. State-reported data were used from two different years because the LEP student counts account for 80% of a state's grant and more recent data were available for LEP student counts than for immigrant student counts. In addition, the use of the 2004-2005 LEP student count data was more comparable to the 2005 ACS data than the 2003-2004 LEP student count data would have been. Details may not add to totals due to rounding. Notice: These are estimated grants only. In addition to other limitations, much of the data which would be needed to calculate final grants are not yet available. These estimates are provided solely to assist in comparisons of the relative impact of alternative formulas and funding levels in the legislative process. They are not intended to predict specific amounts states will receive. Selecting Data on Which To Base the Distribution of Funds The use of either the ACS or state data for calculating Title III state grants has drawbacks. This section examines the methodological issues associated with using either the ACS data or state data as the basis for distributing state grants. It also examines issues specific to counting LEP students and counting recent immigrant students. The section concludes with a brief discussion of the requirement that ED use the most accurate of these data sources to allocate state grants. ACS Data As previously discussed, the ACS data measure factors that are not necessarily related to student enrollment, and there may be data problems due to the subjective nature of the questions and the reliance on self-reported data. For example, respondents to the ACS may not want to report that they speak English less than "very well," as this may be perceived as a socially undesirable response. Although CRS-18 these problems are consistent across states, the Census Bureau found some inconsistency in responses to these questions during its reinterview process to examine data quality.13 In addition, there is no research available that demonstrates how accurately the ACS data represent the population of LEP students.14 If these data continue to be used as the basis for distributing state grants, developing a better understanding of this relationship may be critical. The estimates of the number of students who are recent immigrants are also based on self-reported data. However, the question used to make this determination is more objective than the question used to determine whether a student is LEP, as it asks for factual information. Thus, responses to this question may be more consistent than the questions used to determine LEP.15 State-Reported Data State-reported data also have several problems that could complicate their use as the basis for determining state grants. In responding to a GAO study examining the Title III formula, ED indicated that state data were missing or incomplete for several states.16 ED also noted that states did not necessarily assess all LEP students, which could result in the number of students identified as being LEP exceeding the number of students assessed annually for English language proficiency (as required by Title I of the ESEA).17 In addition, ED noted that states may have provided inconsistent data because the instructions to states for providing this information did not include definitions of the data to be included. GAO found that the aforementioned instructions were sufficiently vague as to allow multiple interpretations of the instructions, and reported that ED had indicated that it would clarify the instructions for the 2006-2007 Consolidated State Performance Report (due in December 2007). ED was also in the process of providing feedback to states on the data provided on the 2003-2004 and 2004-2005 CSPRs and expected that this would lead to improved state data for subsequent school years. Until these data are reported by states, however, it is not possible to know how complete these data will be and whether additional followup, such as the efforts conducted by NCELA regarding the 2004-2005 data, will be needed to produce final counts that could be used as the basis for determining state grants. These issues are further complicated due to the different methodologies used to identify which students are LEP students, as there is no standard methodology by which students with limited English proficiency are identified. Screening instruments used to identify LEP students vary by state and even within states. Even 13 Schneider, P. (2004). Census 2000 Testing, Experimentation, and Evaluation Program (Topic Report No. 12, TR-12). Washington, DC: U.S. Census Bureau. Available online at [http://www.census.gov/pred/www/rpts/TR12.pdf]. 14 GAO, Basis for Distributing Title III Funds. 15 Ibid. 16 Ibid. 17 As previously noted, ED must use the most accurate of either the ACS data or the number of children being assessed for English proficiency as required under Title I. CRS-19 among states using similar methods, the states may differ in their interpretation of the results. States may also differ in how they determine which students to screen for LEP. Although most states use home language surveys to determine what language is spoken at home, some states may also use strategies such as classroom observations to identify students for screening. There are also problems with immigrant student counts reported by states. GAO found that state officials question the reliability of the data they collect, as schools and school districts may not be permitted to ask students directly whether they are immigrant students.18 Rather, some states and districts rely on information about a student's place of birth and date of entry into the school system to determine whether a student is a recent immigrant. These determinations may be further complicated if a student has no prior school documentation, so there is no way to determine based on student records whether the student previously attended another school in the United States and for how long. It should be noted, however, that if grants are determined on the basis of state- reported data, possibly through the CSPRs, a perverse incentive may be created for states to over-report the number of LEP and immigrant students to gain additional federal funds. As the funds available for this program are limited to a specific appropriation amount, if some states inflate their number of eligible students, other states legitimately serving Title III eligible students may receive less funding than they should receive. Data Accuracy Although statutory language permits ED to choose the most accurate of the ACS or state data, ED told GAO that is has not yet established criteria or a methodology for determining which of these data sources is the most accurate.19 According to the GAO report, "Education officials state that as the state data improve and become complete, complex analysis will be needed to determine the relative accuracy of these data and the ACS data."20 Thus, until this analysis is completed by ED or another organization, it may be difficult to determine whether the use of ACS data or state data will result in a grant distribution that most accurately reflects the number of LEP and recent immigrant students by state. 18 Ibid. 19 Ibid. 20 Ibid., p. 13. CRS-20 Possible Alternatives Given the drawbacks in using either the ACS or state-reported data as the basis for determining state grants, other alternatives could be considered. As previously mentioned, no research demonstrates that the ACS data accurately reflect the actual LEP student population. One option may be to require the National Academy of Sciences (NAS) to conduct a study to examine the methodology used to produce the ACS data on which Title III state grants are currently based, the availability of alternative indicators, and the reliability of the data. NAS was required to conduct a similar study of the Small Area Income and Poverty Estimates (SAIPE) data used for Title I purposes (Improving America's Schools Act, P.L. 103-382, Title I, Section 1124(c)(4)). In addition or alternatively, developing a formula based on both the ACS data and the state-reported data could be considered, possibly averaging the student counts from each. This strategy could be used on a long-term basis or as a means of transitioning from the use of ACS data to state data to determine state grants. Using both types of data simultaneously, however, would require state data and ACS data to be available from comparable years. A third alternative, specifically designed to reduce the volatility of the ACS data, would be to average the LEP student counts produced by the ACS for the last two or three available years, and do the same for the ACS immigrant student counts. ------------------------------------------------------------------------------ For other versions of this document, see http://wikileaks.org/wiki/CRS-RL34066