Archived Information

1. For a discussion of the history of the charter school movement, see Budde, R., Education by Charter: Restructuring Schools and School Districts. Andover, MA: The Regional Laboratory for Educational Improvement of the Northeast Islands, 1989. Nathan J., Charter Schools: Creating Hope and Opportunity for American Education. San Francisco: Jossey-Bass, 1996. Sautter, R.C., "Charter Schools: A New Breed of Public Schools," Policy Briefs, Report 2, Oak Brook, IL: North Central Regional Educational Laboratory, 1993.

2. Defining what is or is not a charter school is complex. Cities such as Milwaukee, and public school districts, such as those in New York, have established charter-like schools if one judges by the freedom these schools have from regulations and the choice that students have to attend these schools. The Study, however, defines charter schools as schools established within the provisions of state charter school laws.  There exists no consensus definition of "charter schools" or "charter school legislation." The Study examines schools created under state policies that intend to: (1) allow the creation of schools by means that depart from the previously-established process of starting a school and/or (2) allow schools to operate in a fashion that departs from established practices, often in combination with a performance-based contract. We have opted to exclude some states with legislation or policies that may share charter-like characteristics but which pre-date the introduction of the charter concept (e.g., Oregon). In addition, we have excluded single state-sponsored specialty schools (e.g., state schools for the arts, or schools for low-incidence special education students) even if such schools operate pursuant to the terms of a state-granted charter or charter-like contract. Finally, we have excluded Puerto Rico's "Community Schools" initiative.

3. It is difficult to fix a precise figure for the number of charter schools across the country. New schools begin operating at different times during the year, states define and count charter schools differently, and not all schools that are granted charters begin operation on the date proposed to the state. In addition, some charter schools granted a charter may never become operational, and charter schools that began operation can have their charters revoked. Subsequent reports from the Study will document the number and type of charter schools which go out of existence and report on the reasons for closure (including nonrenewal of a charter and revocation of a charter before the term of its charter has expired). As of January 1997, three operating charters were closed. Study staff identified the number of charter schools in operation by contacting the person responsible for charter schools in state departments of education and requesting information on operating charter schools using our definition, and consulting all available published sources, including charter school directories. The 252 and 428 charter schools reported in Exhibit 2 are schools that were delivering instruction to students as of January 1, 1996 and as of January 1, 1997, respectively. This count excludes branches of the same school operating in different locations under one charter as is the case for several charters in Arizona and California. Ten states had operational charter schools as of January 1996: Arizona, California, Colorado, Georgia, Hawaii, Massachusetts, Michigan, Minnesota, New Mexico, and Wisconsin. The analysis in subsequent chapters is based on charter schools in these ten states. The 15 additional charter-law states and the District of Columbia did not have operational charter schools as of January 1, 1996. Our second annual report will include schools in operation as of January 1, 1997.

4. Kolderie, T., The States Will Have to Withdraw the Exclusive. St. Paul, MN: Center for Policy Studies, 1990, and The Essentials of the 'Charter School' Strategy. St. Paul, MN: Center for Policy Studies, 1994.

5. American Federation of Teachers, Charter School Laws: Do They Measure Up? Washington, DC: The American Federation of Teachers, 1996.

6. Public Law 103-382.

7. Two Minnesota private nonsectarian schools have been converted to public charter schools. However, the Minnesota State Attorney General has issued an opinion that casts doubt on whether other private schools can be permitted to convert unless the developers have first tried to establish a newly created charter school and have been denied. This opinion is currently being challenged.

8. The Wisconsin law allows for the conversion of private schools only in Milwaukee.

9. See note 6.

10. The California State Board of Education began waiving the cap in February 1996 and has granted waivers to at least 16 charters in excess of the cap at the time this report was drafted.

11. Describing a state's approach to charter schools includes case law, regulatory context, and agency actions as well as the charter legislation. Other researchers and commentators have suggested that the states can be categorized into such categories as how much "autonomy" they provide charter schools. See Bierlein, L., and Mulholland, L., Comparing Charter School Laws: The Issue of Autonomy. Policy Brief, Morrison Institute for Public Policy, School of Public Affairs, Arizona State University, September, 1994; Buechler, M., Charter Schools: Legislation and Results After Four Years. Bloomington, IN: Indiana Education Policy Center, January, 1996; Millot, D., Autonomy, Accountability, and the Values of Public Education: A Comparative Assessment of Charter School Statutes Leading to Model Legislation. Santa Monica, CA: RAND, December, 1994.

12. Hereafter, all citations of numbers refer to information the Study gathered, using a telephone survey, from all charter schools that met our definition and were established as of January 1, 1996. At that time, the total number of charter schools was 252. Surveys were completed for 89 percent of these schools. Insofar as possible, information about the remaining eleven percent will be included in next year's report, which will also provide data on charter schools that began operating after January 1, 1996. The total number of charter schools used in the exhibits in this report varies somewhat because some schools did not answer several questions on the survey.

13. Of the 252 charter schools, we have school size data on 223 schools.

14. If we were to exhibit all the analysis and comparisons in the text, the report would be too burdensome to read. Consequently, we will generally display data in the text showing a ten-state comparison and show data in Appendix C to display state-by-state comparisons. We will use endnotes to indicate differences in conclusions that might be drawn if we were to examine the state-by-state data instead of a ten-state base.

15. Pooling the data in this fashion is a form of weighting the data according to the number of charter schools and public schools in the state. Other weighting methods could be used with somewhat different results. Throughout this report we will use the pooling approach, but point out other results that may come about by other ways to weight the data when these differences are important.  Appendix C presents detailed data on various statistics for each of the ten states that had operational charter schools by January 1, 1996. Within-state data not presented in the body of the text are presented in Appendix C.

16. The state-by-state data in Exhibit C-1 of Appendix C shows that for all states with a significant number of charter schools, a much higher proportion of charter schools have less than 200 students than the proportion for all public schools in their states.  All Minnesota charter schools have less than 200 students, and about half of California charter schools are small. In other states, between 50 and 100 percent of charter schools are small compared to other public schools.

17. In Exhibit 8, schools have been classified as elementary, secondary, K-12, and ungraded/other. All schools having grade levels up to grade 8 are called "elementary" schools; all schools having grade levels above 8 (but none below eight) are called "secondary" schools; schools spanning most elementary and secondary grades are called "elementary and secondary" schools; and schools not meeting these definitions are called "other" schools.

18. Of the 22,252 public schools in the 10 charter states, school size data were not available for 588 schools (2.6%). Source: U.S. Department of Education, National Center for Education Statistics, Common Core of Data Survey, 1993-94.

19. Source: U.S. Department of Education, National Center for Education Statistics, Common Core of Data Survey, 1993-94.

20. Of the 225 responding charter schools, the telephone survey provided information about grade-level configuration for 223 schools.

21. See Exhibit C-3, Appendix C for an elaboration of these data. Exhibit C-4 shows the percentage of schools that have less than 200 students for charter schools and for all public schools for each state in the ten-state base.

22. Exhibit C-5, Appendix C shows the percentage of schools that are newly created or pre-existing by school level. These data indicate that school in the K-12 and other grade level configurations are most likely to be newly created schools and that pre-existing private schools are most likely to be primary or early-grade elementary schools.

23. The five racial categories are those used by the Census Bureau in gathering data on individuals.  The Study used these categories to allow comparisons to national data. In descriptions of particular schools and their students, we will refer to students in the way that their schools refer to them.

24. Of the 225 schools that respond to the telephone survey, 11 schools did not have reliable reports for student data. Therefore, the data in the exhibit refers to 214 charter schools.

25. Source: U.S. Department of Education, National Center for Education Statistics, Common Core of Data Survey, 1993-94

26. The small number of charter schools make exact comparisons to all public schools risky for a variety of additional reasons. For example, the total number of schools that were operational in California as of January 1996 was 92. However, nine of these schools did not respond to our 1996 survey. Of the 83 responding schools, three did not have reliable data on racial composition. The total enrollment for charter schools was collected by summing the enrollment and the number of students in each racial category for the 80 schools rather than for the 92 charter schools that were operational in 1996. Similarly, for Arizona, we report on 38 of the 47 charter schools; for Colorado, 22 of the 24; for Massachusetts, 13 of the 15; and for Michigan, 38 of the 43. In the remaining states, we report data for all charter schools (Georgia, 3, Hawaii, 2, New Mexico, 4, Minnesota, 17, and Wisconsin, 5).

27. Only one charter school had students who were predominantly of Asian origin. Seventy-five percent of the almost 200 students are of Asian origin at this elementary school located in an inner city.

28. Source: U.S. Department of Education, National Center for Education Statistics, Common Core of Data Survey, 1993-94

29. Although not subscribing to the opinion that charter school admission policies are inequitable, other researchers and commentators have also reported this issue as a common concern among charter school critics. Becker, H. J., Parent Involvement Contracts in California's Charter Schools. Los Alamitos, CA: Southwest Regional Laboratory (WestEd), 1995. Buechler, M., Charter Schools: Legislation and Results after Four Years. Bloomington IN: Indiana Education Policy Center, January, 1996. Jacobson, L., "Under the Microscope: As Charter Schools Flourish, the Big Question for Researchers is: Do They Work?" Education Week, 6 November, 1996. O'Neil, J., "New Options, Old Concerns," Educational Leadership, Vol. 54, No. 2 (October 1996). Yamashiro, K. and L. Carlos, ?Issues at a Glance: More on Charter Schools,? San Francisco, CA: WestEd, 1996.

30. The extent to which individual charter schools serve "representative" student populations is difficult to define. What is the relevant population of students to which the charter school students should be compared? Charter schools may draw students from multiple districts or attendance areas. For example, 15 percent of the charter schools are independent study or home-based schools, with 70 percent being newly created. These schools, including those that use distance-learning technologies, enroll students who live beyond conventionally defined attendance areas. Also, comparing the racial composition of charter school students to public school students within existing attendance areas or districts may be misleading because of the small size of most charter schools. When such schools are within the boundaries of a large district, comparisons about representativeness may not be meaningful unless the district only has one predominant racial group. The Study's preliminary site visits to a sample of predominately white schools suggest that these schools reflect the geographic area within which they are located, where geographic area was loosely defined to encompass access to the school with existing or supplied transportation.

31. More subtle processes of selecting students, however, may be at work. Intensive field research in subsequent years should allow us to probe deeper into selection processes. For example, we will want to ask, in situations where it is possible, whether charter schools actively seek out students from diverse ethnic or racial backgrounds. The research team documented several cases where the schools do reach out actively, but we cannot report definitive data at this time.

32. Source: To Assure the Free Appropriate Public Education of All Children with Disabilities: Eighteenth Annual Report to Congress of the Implementation of the Individuals With Disabilities Education Act. Washington, DC: U.S. Department of Education, 1996.

33. U.S. Department of Education, Office of Bilingual Education and Minority Languages Affairs, Summary of Bilingual Education State Education Agency Program Survey of States' Limited English Proficient Persons and Available Educational Services 1993-94, Development Associates, Inc., 1995. As was the case for students with disabilities, data on LEP students are not included in the U.S. Department of Education's Common Core of Data Survey. Consequently, the comparisons indicated in the text are preliminary.

34. U.S. Department of Education, Office of Bilingual Education and Minority Languages Affairs, Summary of Bilingual Education State Education Agency Program Survey of States' Limited English Proficient Persons and Available Educational Services 1993-94, Development Associates, Inc., 1995.

35. Source: Office of Bilingual Education and Minority Languages Affairs.

36. Source: U.S. Department of Agriculture, Food and Consumer Services, Program Information Division, 1994.

37. The exhibit is based on the 207 charter schools that provided information on Title I eligibility--92 percent of the 225 schools that responded to the survey and 82 percent of the operational charter schools.

38. Four charter schools that were eligible for Title I did not provide data on receipt of Title I funding; data on these schools are excluded from the final column of this exhibit.

39. We were unable to obtain data on Title I eligibility for all public schools within the ten charter states.

40. Charter schools' eligibility for Title I can be complicated; eligibility depends on (1) whether the charter school is part of a school district or independent from any district; (2) the poverty rate of the charter school; and (3) the poverty cut-off that determines schools' eligibility Title I in the district.

41. As with any self-reported data, the answers to the telephone survey must be treated cautiously. Since usually only one individual at the school responded to the telephone interview, the individual's responses may not represent the consensus at a school. To guard against these and other sources of inaccuracy, we will present only findings from the survey that are consistent with the interview and observational information gathered in the field.

42. The 42 charter schools in the first-year fieldwork sample cover a range of the diverse conditions under which charter schools were started and implemented. The sample includes schools from five states in which charters were operational as of the 1994-95 school year--California, Colorado, Michigan, Minnesota, and Wisconsin; newly created charter schools as well as public and private pre-existing schools that were converted to charter status (about two-thirds of the fieldwork sample are newly created schools); schools spanning a wide variety of grade-level configurations and school sizes; and schools with racial diversity similar to that described in chapter III for the total population of charter schools. The Study's first-year fieldwork did not include New Mexico. This state passed a charter law in 1993 which permitted the creation of five charter schools. New Mexico's charter schools can only be schools that convert from pre-existing public schools; they can only be sponsored and approved by school districts of which they remain a legal arm, and state regulations can be relaxed only through a waiver process. At the time, the sample was drawn, it was not clear whether the state's charter schools had been in full operation for a minimum of one year.

43. We consider this classification to be quite preliminary for several reasons.  First, the phone interviewers spoke to one person at the school for the purpose of completing the survey. It is possible that different founding members would disagree as to the most important reason or reasons for starting the charter, in which case we would not be obtaining an accurate answer. Although different respondents at the same school often spoke of additional related reasons for founding the school, most interviewees at field sites who were either involved in the founding or had explicit knowledge of that process seemed to agree on the most important reason. Nineteen of the 225 schools in the phone sample did not provide answers that could be coded. It is possible that some of these cases were the result of inconsistent views at the school. Second, since many schools have a variety of reasons for becoming a charter school, asking for the single most important reason could unintentionally force an answer that is contrary to the reality. However, Appendix D presents data tables parallel to Exhibits 19 and 20 that show the multiple responses which phone respondents gave to list their most important reasons; the patterns in these tables are virtually identical to those in the exhibits in this chapter. Third, the categories themselves are abstract categories that the Study derived from the telephone answers and the field experience. When asked why a charter school was started, school staff, parents, and community members cited many specific reasons that arose from their unique context. More in-depth site visits will be needed to develop a richer or more valid classification of the reasons why schools were founded, but at this early stage of the Study the coded responses from the telephone interviews were confirmed by the field visits.

44. When multiple responses are counted, 81 percent of newly created schools cited realizing an educational vision.

45. Of the 225 charter schools that responded to the survey, 206 responded to this question in ways that could be coded.

46. When multiple responses are counted, 61 percent of new schools cited realizing an educational vision.

47. The telephone survey also asked respondents to indicate which curriculum and instructional strategies from a list of educational approaches were major features at the school. The answers to these closed-ended questions also indicated that charter schools are implementing a broad range of curriculum and instructional approaches. The responses from newly created and pre-existing schools show no significant difference in the percentage of charter schools which say they are currently implementing the various approaches from that list.

48. Of the 225 charter schools that responded to the survey, 206 responded to this question in ways that could be coded.

49. As reported in chapter III, at least five charter schools serve predominantly Native American students.

50. When multiple-responses are counted, 23 schools cited financial reasons, and 44 percent were pre-existing private schools.

51. In our fieldwork, we found that schools vary in the number of parent volunteer hours required and in their sanctions (or lack of) for parents who fail to volunteer the required hours. We will explore this issue in greater depth in future fieldwork.

52. When multiple-responses are counted, 21 schools cited parent involvement as an important reason for founding their charter school and 61 percent of these were newly created charter schools.

53. When multiple-responses are counted, 21 schools cited attracting students as an important reason. Forty-three percent of these charter schools were private pre-existing schools.

54. The statistical test for the Analysis of Variance (ANOVA) uses the F ratio, which is obtained by dividing the between groups mean square by the within groups mean square. The significance level is obtained by comparing the calculated F value to the F distribution. The last column refers to the significance level for one degree of freedom.

55. Factor Analysis is a statistical procedure used to identify latent dimensions underlying related groups of variables. The factors are hypothetical variables (or underlying dimensions) that represent the minimum number of hypothetical variables that explain the variation across the answers given by respondents to the barriers questions. The numbers shown in Exhibit 24 are similar to correlations between the barrier questions (in which the respondents' scores of 1 to 5 are assumed to be an equal interval scale for the sake of this exploratory analysis) and "factor scores." A factor score is the result of fitting the intercorrelations among the data to best clarify the empirical patterns revealed in the data. (In this case, we used a varimax rotation.) The numbers in the table have no intrinsic substantive meaning, but a number close to one (or -1) is a "high correlation" between the respondents' answers on each particular barrier question and the underlying dimension of difficulty.

56. To amplify and explore further the material presented in Exhibit 25, Appendix D shows the results of using the factor scores from the analysis whose results are presented in Exhibit 24. The comparison of the mean factor scores for the three clusters is generally in agreement with the conclusions suggested by Exhibit 25, except that the factor scores show that more newly created charter schools have significantly more difficulty from resource limitations than do pre-existing schools.

[Summary]

[Appendix A]