LIU - FINAL SOLVING THE PUZZLE OF CHARTER SCHOOLS: A NEW FRAMEWORK FOR UNDERSTANDING AND IMPROVING CHARTER SCHOOL LEGISLATION AND PERFORMANCE Elaine Liu Contributing to the controversy over charter schools are empirical studies, such as the Center for Research on Education Outcomes National Charter School Study 2013, that find that charter schools have mixed results, with some seeing better student achievement outcomes relative to traditional public schools, and others seeing worse. This Note argues that developing a new framework to understand differences in state charter school laws based on their underlying purposes helps explain the significant variation across states in charter school performance. Based on a state-by-state analysis of charter school legislation, this Note finds that states with charter school laws generally fall into three categories: (1) “gap-closing”; (2) “libertarian-oriented”; and (3) “mixed.” The first category includes states with laws that promote charter schools focused on improving outcomes for disadvantaged and underachieving students. The second category includes states with laws that promote charter schools as a means to provide parents and local communities, rather than the government, with more choices and greater control in education. The third  J.D. Candidate 2015, Columbia Law School; A.B. 2010, Harvard University. The author would like to thank Professor James S. Liebman for his guidance. She would also like to express gratitude to her family, including her sisters Eva and Eden, for all their support throughout her life, and to the editorial staff of the Columbia Business Law Review for their assistance in preparing this Note for publication. LIU - FINAL 274 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 category encompasses those states where charter school legislation promotes both of these purposes. Furthermore, results indicate that gap-closing states generally enjoy better charter school performance outcomes than do libertarian- oriented states, and that the typical charter school law of gap- closing states differs significantly from that of libertarian- oriented states. As such, this Note proposes that states can maximize charter school student achievement by adopting features that are characteristic of gap-closing states’ charter school laws—e.g., greater restrictions on entrance into the charter school market, and greater oversight of both charter school authorizers and charter schools themselves. I. Introduction ............................................................... 275 II. Overview of the Charter School Landscape .............. 279 A. Origins and Rapid Expansion of Charter Schools ................................................................. 279 B. Review of Literature on the Inconsistency in Charter School Performance ............................... 285 C. Review of Literature on the Variation in Charter School Laws ........................................... 289 D. Review of Literature Evaluating Charter School Laws ......................................................... 291 III. Study Methodology .................................................... 295 A. Development of the Framework Based on Legislative Purpose ............................................. 295 B. Assessment of the Correlation Between Legislative Purpose and Outcomes ..................... 297 C. Analysis of Key Charter School Law Features by Legislative Purpose ........................................ 299 V. An Original Framework: Gap-Closing Versus Libertarian-Oriented States ...................................... 307 A. Determination of How to Best Categorize Legislative Purposes ............................................ 307 B. Assignment of Charter School Laws to Legislative Purpose Categories ........................... 313 V. Correlation Between Legislative Purpose and Outcomes .................................................................... 318 LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 275 A. Overall Comparison of Charter and Traditional Public School Performance .............. 318 B. Average Magnitude of Charter School Impact ... 319 VI. Key Features of Charter School Laws by Legislative Purpose .................................................... 322 A. Charter School Sector Flexibility ........................ 322 B. Charter School Autonomy ................................... 329 C. Accountability ...................................................... 334 D. Funding ................................................................ 339 E. Overall Policy Implications ................................. 342 VII. Conclusion .................................................................... 346 Appendix .............................................................................. 347 I. INTRODUCTION [T]here is a certain common level of education that all kids need, no matter how different they are, and . . . it would be a good thing to allow [charter] schools to be developed which had a clear mission, which could reach out to kids who wanted to be a part of that mission, [which] could achieve educational excellence for children who otherwise might be left behind . . . . — William J. Clinton, 42nd President of the United States1 Charter schools encourage educational entrepreneurs to try innovative methods. . . . These diverse, creative schools are proof that parents from all walks of life are willing to challenge the status quo if it means a better education for their children. More competition and more choices for parents and students will raise the bar for everyone. — George W. Bush, 43rd President of the United States2 1 William J. Clinton, President of the United States, Remarks at the City Academy in St. Paul, Minnesota (May 4, 2000) (transcript available at http://www.presidency.ucsb.edu/ws/?pid=58436, archived at http://perma. cc/86ZU-L76B). LIU - FINAL 276 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 Since 1991, when Minnesota became the first state to enact legislation to permit the formation of charter schools, charter schools have rapidly grown in number.3 Despite this increase in the number of charter schools nationwide, heated debate over the merits of establishing charter schools continues. Some critics have argued that charter schools are a harmful “conspiracy” by hedge fund managers and other “corporate” interests4 to destroy public education and replace it with marketized private choice.5 Contributing to the controversy are empirical studies that find that charter schools have inconsistent results, with some seeing better student achievement outcomes relative to traditional public 2 Presidential Quotes on Public Charter Schools and U.S. Education, NAT’L ALLIANCE FOR PUB. CHARTER SCH.: THE CHARTER BLOG (Feb. 9, 2013, 12:00 AM), http://www.publiccharters.org/2013/02/presidential-quotes- public-charter-schools-u-s-education/, archived at http://perma.cc/8DB5- XHCX. 3 Charter schools in 2012–2013 number more than 6,000, or 6.3% of all public schools. See The Public Charter Schools Dashboard, NAT’L ALLIANCE FOR PUB. CHARTER SCH., http://dashboard.publiccharters.org/ dashboard/schools/page/overview/year/2013, archived at http://perma.cc/ M5LY-XF94. See also Robin J. Lake, Will the Charter Movement Rest on Its Laurels or Innovate and Expand?, in HOPES, FEARS & REALITY: A BALANCED LOOK AT AMERICAN CHARTER SCHOOLS IN 2012 1 (Robin J. Lake ed., Univ. of Wash. Ctr. on Reinventing Pub. Educ. 8th ed. 2013), http://www.crpe.org/sites/default/files/pub_hfr12_may13.pdf, archived at http://perma.cc/7UE7-3QZN (“Since the first charter law was enacted in 1991, the movement has grown steadily, with 300 to 400 new charter schools added each year . . . .”). 4 Three of the largest private funders of the pro-charter school movement are the Bill and Melinda Gates Foundation, the Walton Family Foundation, and the Eli and Edythe Broad Foundation. Joanne Barkan, Got Dough? How Billionaires Rule Our Schools, DISSENT, Winter 2011, at 49. 5 See, e.g., Michael Q. McShane, Putting Charter School Conspiracy Theories to Rest, EDUCATIONNEXT (Feb. 20, 2013), http://education next.org/putting-charter-school-conspiracy-theories-to-rest/, archived at http://perma.cc/MK5R-635D (“In education policy circles, the ‘charter schools are a plan by ultra-conservatives to privatize the public school system’ is a conspiracy theory that is quite popular.”). LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 277 schools, and others seeing worse.6 This variation in performance has led some critics to argue that charter schools do not warrant their costs,7 while other education reformers have sought to better understand what factors foster the success of charter schools.8 This Note argues that developing a meaningful framework to understand the differences in state charter school laws—and their purposes in particular—helps explain the significant variation across states in charter school performance. The epigram quotations at the beginning of this Note from former Presidents William J. Clinton and George W. Bush illustrate not only the diverse political coalition that has supported the development of charter schools, but also the different emphases and goals that supporters can have, such as ensuring that the worst students still meet minimum standards, or expanding the choices that parents can have for their children. Based on a state-by-state analysis of charter school legislation, this Note finds that states with charter school laws generally fall into three categories in terms of their purpose. The first category includes states with laws that promote charter schools focused on improving outcomes for disadvantaged and underachieving students, i.e., “gap- closing states.” The second category includes states with laws that promote charter schools as a means to provide parents and local communities, rather than the government, 6 See, e.g., Julian Betts & Emily Tang, The Effect of Charter Schools on Student Achievement: A Meta-Analysis of the Literature, CTR. ON REINVENTING PUB. EDUC. 1 (Oct. 2011), http://files.eric.ed.gov/fulltext/ ED526353.pdf, archived at http://perma.cc/WMM5-72XZ (“[T]he authors find compelling evidence that charters under-perform traditional public schools in some locations, grades, and subjects, and out-perform traditional public schools in other locations, grades, and subjects.”). 7 See, e.g., Diane Ravitch, Op-Ed., The Charter Mistake, L.A. TIMES, Oct. 1, 2013, at A15. 8 See, e.g., Valerie Strauss, Why Charter Schools Need Better Oversight, WASH. POST ANSWER SHEET (Sept. 5, 2013, 11:00 AM), http://www.washingtonpost.com/blogs/answer-sheet/wp/2013/09/05/why- charter-schools-need-better-oversight/, archived at http://perma.cc/SW9T- UHNP. LIU - FINAL 278 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 with more choices and greater control in education, i.e., “libertarian-oriented states.” Finally, the third category encompasses those states where charter school legislation promotes both of these purposes, i.e., “mixed states.” Furthermore, results indicate that gap-closing states generally enjoy better charter school performance outcomes than do libertarian-oriented states, and that the typical charter school law of gap-closing states differs significantly from that of libertarian-oriented states. As such, this Note proposes that states can maximize charter school student achievement by adopting features that are characteristic of gap-closing states’ charter school laws. Key steps include greater restrictions on entrance into the charter school market (e.g., limiting for-profit education management organizations) and greater oversight of charter school authorizers and charter schools themselves (e.g., giving oversight bodies the authority to impose sanctions based both on student performance and financial outcomes). Part II examines the current status of charter schools in the United States and reviews the literature evaluating state charter school laws. Part III explains the methodology this Note uses to analyze state charter school laws. Part IV provides an original framework for distinguishing different charter laws based on their underlying purposes. Part V shows how this framework helps explain the variation among states in charter school performance. Part VI identifies the key features of charter school laws that are associated with gap-closing, libertarian-oriented, and mixed purposes and proposes steps to improve charter school performance through adoption of the crucial features of gap- closing charter school laws. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 279 II. OVERVIEW OF THE CHARTER SCHOOL LANDSCAPE A. Origins and Rapid Expansion of Charter Schools Whether created from scratch (“start-ups”) or converted from traditional public or private schools (“conversions”),9 charter schools are usefully defined as “publicly funded elementary or secondary schools that have been freed from some of the rules, regulations, and statutes that apply to other public schools, in exchange for some type of accountability for producing certain results, which are set forth in each charter school’s charter.”10 Areas over which state laws tend to provide charter schools with increased autonomy (i.e., exemption from regulation to some degree) include curriculum, staffing, and budget and resource allocation decisions. Autonomy is not complete by any means, however. For example, charter schools often must comply with regulations regarding class size, graduation, bilingual education, special education, health, safety, and civil rights.11 Charter school laws also typically allow families to choose whether to attend these schools, exempting them from rules requiring students to attend one 9 Joe Nathan, Heat and Light in the Charter School Movement, 79 THE PHI DELTA KAPPAN 499, 500 (1998); see also Sara Mead & Andrew J. Rotherham, A Sum Greater Than the Parts: What States Can Teach Each Other About Charter Schooling, EDUC. SECTOR 7–8 (Sept. 2007), http:// www.educationsector.org/publications/sum-greater-parts-what-states-can- teach-each-other-about-charter-schooling, archived at http://perma.cc/ YVY4-YQMW; Ron Zimmer & Richard Buddin, Getting Inside the Black Box: Examining How the Operation of Charter Schools Affects Performance, 82 PEABODY J. EDUC. 231, 233 (2007). 10 MORLEY D. GLICKEN, SOCIAL WORK IN THE 21ST CENTURY: AN INTRODUCTION TO SOCIAL WELFARE, SOCIAL ISSUES, AND THE PROFESSION 161 (2d ed. 2011); see also Charter School, K12 ACADEMICS, http://www.k12academics.com/alternative-education/charter-school#.VKj9 sisViOM, archived at http://perma.cc/BGD7-YFDR (last visited Jan. 14, 2015). 11 Lance D. Fusarelli, Charter Schools: Implications for Teachers and Administrators, 76 CLEARING HOUSE 20, 21 (2002). LIU - FINAL 280 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 assigned school or another.12 In most states, charter schools must employ a lottery to select students when applications to enroll exceed the school’s capacity.13 Charter schools across the nation can vary widely in their missions, program structure, and other characteristics.14 For instance, among the most controversial type are “virtual” charter schools or “cyberschools,” which offer online, non- classroom-based instruction to students.15 Some states, such as Delaware, Louisiana, and New York, prohibit these schools entirely.16 The concept of charter schools originates from New England educator Ray Budde’s proposal, more than twenty years ago, that local school boards give small groups of teachers “charters,” or contracts, to experiment with new 12 Nathan, supra note 9, at 500. 13 Robert Bifulco & Helen F. Ladd, The Impacts of Charter Schools on Student Achievement: Evidence from North Carolina, 1 EDUC. FIN. & POL’Y 50, 51 (2006). 14 See generally Theola Labbé, Six Charter Schools Opening with Unique Outlooks, WASH. POST THE BREAKING NEWS BLOG (Aug. 24, 2006), http://www.washingtonpost.com/wp-dyn/content/article/2006/08/23/AR200 6082300689.html, archived at http://perma.cc/ADA6-P9F3 (describing new charter schools with a diverse range of programs, from a half-day program serving parents with limited English-speaking skills and preschool children to a school serving children and adults with special needs); see also Jeffrey R. Henig et al., The Influence of Founder Type on Charter School Structures and Operations, 111 AM. J. EDUC. 487, 488 (2005) (“‘[C]harter school’ is an umbrella term that can apply to a wide range of organizations differing in mission, background, and behavioral tendencies.”). Henig et al. propose categorizing the various kinds of charter schools based on “founder type,” distinguishing between “mission- oriented” schools, which are “assumed to set in a direction more in line with a purposive, collective, and philanthropic mission,” and “market- oriented” schools, which are “assumed to be more oriented toward markets.” Id. at 487–89. 15 Mead & Rotherham, supra note 9, at 9. 16 See The Essential Guide to Charter School Law: Charter School Laws Across the States 2012, CTR. FOR EDUC. REFORM 23, 39, 59 (Apr. 2012), http://edreform.com/wpcontent/uploads/2012/04/CER_2012_Charter _Laws.pdf, archived at http://perma.cc/YVF4-YVBE [hereinafter 2012 CER RANKINGS]. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 281 educational approaches.17 Albert Shanker, then-president of the American Federation of Teachers, further refined and publicized this idea.18 These proposals eventually led to the first charter school legislation in Minnesota in 1991. Charter schools have rapidly expanded since then and now educate more than 2.3 million students, or nearly five percent of all public school students.19 Regions with a particularly high proportion of public school students in charter schools are California, Florida, and the District of Columbia.20 For example, District of Columbia charter schools educate forty- three percent of all public school students in the nation’s capital.21 In addition, the Obama administration has strongly encouraged the establishment of charter schools in recent years through its Race to the Top program, which privileges states with liberal charter laws in scoring grant proposals.22 Currently, only seven states—Kentucky, 17 Nathan, supra note 9, at 500. 18 Id.; see also Stan Karp, Charter Schools and the Future of Public Education, N.J. EDUC. ASSOC. REV., Mar. 2013, at 11. Despite being an early advocate for the charter school movement, Shanker later “largely repudiated a major reform he had helped launch” due to concerns that the movement undercut teachers’ unions and resembled his “two central nemeses [of] . . . school vouchers on the right and ‘community control’ on the left.” RICHARD D. KAHLENBERG, TOUGH LIBERAL: ALBERT SHANKER AND THE BATTLES OVER SCHOOLS, UNIONS, RACE, AND DEMOCRACY 318 (2009). 19 Charter school student enrollment increased by eighty percent in the last five years alone. A Growing Movement: America’s Largest Charter School Communities, NAT’L ALLIANCE FOR PUB. CHARTER SCH. 2 (Dec. 2013), http://www.publiccharters.org/wp-content/uploads/2014/01/2013- Market-Share-Report-Report_20131210T133315.pdf, archived at http:// perma.cc/W6DW-NVT9. 20 See id. at 6. 21 See id. 22 Jon Christensen, Jacqueline Meijer-Irons & Robin J. Lake, The Charter Landscape, 2004–2009, in HOPES, FEARS & REALITY: A BALANCED LOOK AT AMERICAN CHARTER SCHOOLS IN 2009, at 4 (Robin J. Lake ed., Univ. Wash. Ctr. on Reinventing Pub. Educ. 5th ed. 2010), http://www.crpe.org/sites/default/files/pub_hfr12_may13.pdf, archived at http://perma.cc/7UE7-3QZN; Don Parker-Burgard, Limits on Charter Schools an Obstacle to Race to the Top Funds, DISTRICT ADMIN., http://www.districtadministration.com/article/limits-charter-schools- LIU - FINAL 282 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 Montana, Nebraska, North Dakota, South Dakota, Vermont, and West Virginia—lack legislation permitting the operation of charter schools.23 If trends continue, charter schools will increase their importance as a tool of educational reform throughout the United States in the coming years and increase their share of the public school “market” relative to traditional public schools. Recent changes in Chicago illustrate this shift; shortly after deciding to close an unprecedented fifty traditional public schools based on analyses indicating that they were underutilized, Chicago Public Schools invited applications for new charter schools to open in neighborhoods with overcrowded schools.24 Interest in charter schools as a potentially powerful means of improving education and a desire to rapidly replicate high-quality educational models have helped spur obstacle-race-top-funds, archived at http://perma.cc/S8QG-8XNL; Paul Manna & Laura L. Ryan, Competitive Grants and Educational Federalism: President Obama’s Race to the Top Program in Theory and Practice, 41 PUBLIUS: J. FEDERALISM 522, 527 (2011) (“[O]ne specific criterion awarded up to forty points for states that had created conditions to promote the development of high-performing public charter schools and other innovative schools. The weighting clearly favored charter schools, though, because if a state had many policies to promote different alternatives except charter schools then the most it could earn on this criterion was eight points.”). 23 The Last Eight States Without Charter Laws, CTR. FOR EDUC. REFORM (Jan. 2013), http://www.edreform.com/wp-content/uploads/2013/ 01/CharterLaws2013-Last-8-States.pdf, archived at http://perma.cc/DJ88- PSVA. In March 2015, Alabama became the latest state to enact charter school legislation. Ala. Governor Signs Charter School Bill, EDUC. WEEK, Mar. 25, 2015, at 4. 24 John Byrne, Emanuel Defends Charters in Wake of School Closings, CHI. TRIB., Jan. 29, 2014, at 4; Linda Lutton, Just Months After Closing 50 Schools, Chicago Issues RFP for More Charter Schools, WBEZ91.5 (Aug. 13, 2013), http://www.wbez.org/news/education/just-months-after-closing- 50-schools-chicago-issues-rfp-more-charter-schools-108398, archived at http://perma.cc/BM6K-J99Q. See also Emma Brown, Education Digest: D.C. Charter Schools Have Nation’s 3rd-Highest Market Share, WASH. POST, Dec. 16, 2013, at B3 (discussing how charter schools’ “market share” has been growing in many cities—including in the District of Columbia, where there have been “questions about the future of the traditional public school system, which closed 13 schools for low enrollment”). LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 283 the growth of national for-profit education management organizations (“EMOs”) and nonprofit charter management organizations (“CMOs”).25 Groups of parents, teachers, and/or other community members historically were, and remain, the dominant force in establishing standalone charter schools (i.e., “mom-and-pop” charters).26 At the same time, however, networks of schools operated by for-profit EMOs and nonprofit CMOs,27 with well-developed central- office structures of their own, have increased in size.28 The number of states with EMOs increased from sixteen in 1998– 1999 to thirty-five in 2011–2012 while the number of states with CMOs have increased from ten in 1998–1999 to twenty- nine in 2011–2012.29 This rapid expansion of charter schools, especially those that are part of national networks, has been accompanied by 25 See Caitlin Farrell et al., Charter Management Organizations: An Emerging Approach to Scaling Up What Works, 26 EDUC. POL’Y 499, 499– 500 (2012). 26 Id. at 500; see also Claudio Sanchez, What’s a Charter School If Not a Game Changer?, NPR (Aug. 31, 2012, 4:22 PM), http:// www.npr.org/2012/09/01/160401996/whats-a-charter-school-if-not-a-game- changer, archived at http://perma.cc/CQB9-ND4L (“68 percent [of all charter schools] . . . are so-called ‘mom and pop’ charters, run by parents and local community groups.”). 27 Major EMOs include Imagine Schools and National Heritage Academies, and well-known CMOs include KIPP Foundation and Aspire Public Schools. See CMO and EMO Public Charter Schools: A Growing Phenomenon in the Charter School Sector, NAT’L ALLIANCE FOR PUB. CHARTER SCH. 1, 3 (Nov. 2011), http://www.publiccharters.org/wp-content/ uploads/2014/01/napcs-cmo-emo-dashboard-details_20111103T102812.pdf, archived at http://perma.cc/QY3A-3PXC; see also Gary Miron & Charisse Gulosino, Profiles of For-Profit and Nonprofit Education Management Organizations Fourteenth Edition—2011–2012, NAT’L EDUC. POL’Y CTR. 23–29 (Nov. 2013), http://nepc.colorado.edu/files/emo-profiles-11-12.pdf, archived at http://perma.cc/2UY2-E2L8. 28 See Miron & Gulosino, supra note 27, at 9, 11 (finding that large EMOs and CMOs, which are defined as organizations managing ten or more schools, account for more than seventy-five percent of all EMO- managed schools, and fifty-one percent of all CMO-managed schools, respectively). 29 See id. at 5, 7. LIU - FINAL 284 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 strong support from the business community and “corporate” interests. Contributions from hedge fund managers30 and private philanthropists31 in particular have constituted a sizeable portion of overall funds for charter schools. For example, since 2000, the Walton Family Foundation— governed by the family that founded Walmart—has donated more than $100 million to the Charter School Growth Fund,32 and more than $50 million to the nonprofit national charter school network KIPP.33 This funding has led some critics to condemn charter schools as part of a concerted effort by “corporate” and “rightwing” interests to destroy public schools and replace them with marketized private choice, while making significant profits at the same time.34 Other scholars have dismissed such “conspiracy theories,” 30 Laura Clawson, Education: Follow the Money, DAILY KOS (Oct. 17, 2010, 4:59 PM), http://www.dailykos.com/story/2010/10/17/910960/- Education-follow-the-money#, archived at http://perma.cc/B9BP-5JCX. 31 Joanne Barkan, Got Dough? How Billionaires Rule Our Schools, DISSENT, Winter 2011, at 49. 32 The Charter School Growth Fund invests in CMOs to help them expand. Investing in CMOs, CHARTER SCH. GROWTH FUND, http://chartergrowthfund.org/what-we-do/investing-in-cmos/, archived at http://perma.cc/E6W9-48SW (last visited Jan. 9, 2015). 33 Motoko Rich, A Walmart Fortune, Spreading Charter Schools, N.Y. TIMES, Apr. 26, 2014, at A1. 34 See, e.g., Clawson, supra note 30; Peter DeWitt, Conspiracy Theory: Privatizing Public Education, EDUC. WEEK: FINDING COMMON GROUND (Aug. 23, 2012, 2:07 PM), http://blogs.edweek.org/edweek/finding_comm on_ground/2012/08/conspiracy_theory_privatizing_public_education.html, archived at http://perma.cc/BUE5-DZ55; Valerie Strauss, Ravitch on Obama’s Scary Ed Reform Agenda, WASH. POST ANSWER SHEET (May 19, 2010, 6:30 AM), http://voices.washingtonpost.com/answer-sheet/charter- schools/ravitch-on-obamas-scary-ed-ref.html, archived at http://perma.cc/ JX26-K9ZJ. Some critics of corporate sponsorship of charter schools, including Professor Diane Ravitch, have rejected characterizing their argument as a “conspiracy theory.” See Diane Ravitch, Richard Brodsky: Wall Street and Rightwing Billionaires Are Key Players in Education Policy, DIANE RAVITCH’S BLOG (Oct. 13, 2014), http://dianeravitch.net/ 2014/10/13/richard-brodsky-wall-street-and-rightwing-billionaires-are-key- players-in-education-policy/, archived at http://perma.cc/XB87-G37W. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 285 and have emphasized the diverse group of supporters that originally founded the charter school movement.35 B. Review of Literature on the Inconsistency in Charter School Performance Despite their continued growth, charter schools remain controversial.36 Contributing to this debate is that, in spite of academic interest in this area, there is no strong empirical evidence demonstrating that charter schools are consistently more effective than traditional public schools.37 As such, the fundamental question of whether charter schools better 35 See, e.g., Skanda Amarnath, Economics & Education: Charter Schools and the Corporate Conspiracy, EDLAB (July 11, 2011, 4:21 PM), http://edlab.tc.columbia.edu/index.php?q=node/6087, archived at http:// perma.cc/3JL8-B5Q7; Michael Q. McShane, Putting Charter School Conspiracy Theories to Rest, EDUCATIONNEXT (Feb. 20, 2013), http:// educationnext.org/putting-charter-school-conspiracy-theories-to-rest/, archived at http://perma.cc/MK5R-635D. 36 Joy Resmovits, Charter Schools Continue Dramatic Growth Despite Controversies, HUFFINGTON POST (Dec. 10, 2013), http://www. huffingtonpost.com/2013/12/10/charter-schools_n_4419341.html, archived at http://perma.cc/J8AE-68PL. 37 See, e.g., Robert A. Fox et al., The Line Between Cultural Education and Religious Education: Do Ethnocentric Niche Charter Schools Have a Prayer?, 36 REV. RES. EDUC. 282, 300 (2012); Lance D. Fusarelli, Charter Schools: Implications for Teachers and Administrators, 76 CLEARING HOUSE 20, 22 (2002); Eric A. Hanushek et al., Charter School Quality and Parental Decision Making with School Choice, 91 J. PUB. ECON. 823, 824 (2007) (“[E]ven though charter schools have captured the imagination of many school reformers and the ire of others, little credible evidence about their impact on student achievement is available. This comes about primarily because of the difficulty separating differences in the quality of charter and regular public schools from differences in the students who attend schools in the respective sectors.”). For an overview of key recent studies conducted on charter school effectiveness, see Public Charter School Success: A Summary of the Current Research on Public Charters’ Effectiveness at Improving Student Achievement, NAT’L ALLIANCE FOR PUB. CHARTER SCH. 2–8 (Apr. 23, 2013), http://www.publiccharters.org/wp- content/uploads/2014/01/NAPCS_2013_Research_Summary_20130424T14 5509.pdf, archived at http://perma.cc/25ZB-447J. See also Betts & Tang, supra note 6, at 1, 3–5, 55–58. LIU - FINAL 286 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 educate students remains unresolved.38 Recent studies have also questioned the claim made by many charter school proponents that such schools innovate and experiment with new teaching methods more often and more effectively than traditional public schools.39 Many studies draw broad policy conclusions from analyses of a relatively small population of charter school students or from results in one geographic area—typically one city or state.40 Major differences among state charter school laws,41 however, undermine the value of such studies as a basis for making generalizations about the likely effect 38 Claudio Sanchez, The Charter School vs. Public School Debate Continues, NPR (July 16, 2013, 5:02 PM), http://www.npr.org/2013/07/16/ 201109021/the-charter-school-vs-public-school-debate-continues, archived at http://perma.cc/A9HK-83RF (“[O]ne key question lingers: Do kids in charter schools learn more than kids in traditional public schools?”). 39 See, e.g., Michael Mintrom, Policy Design for Local Innovation: The Effects of Competition in Public Schooling, 1 ST. POL. & POL’Y Q. 343, 358 (2001) (“[M]y findings should give pause to those who see competition in public service supply as a system-enhancing cure-all. Yes, my findings suggest that competition has encouraged some charter schools and some traditional public schools in Michigan to innovate more than schools that are immune from such competition. Nonetheless, . . . there is a large unexplained—and, thus, quite possibly random—component to why some schools in Michigan are more innovative than others.”). 40 See, e.g., Jason M. Barr et al., Charter Schools and Urban Education Improvement: A Comparison of Newark’s District and Charter Schools, 38 THE URB. REV. 291, 291 (2006); Eric P. Bettinger, The Effect of Charter Schools on Charter Students and Public Schools, 24 ECON. EDUC. REV. 133, 133 (2005); Hanushek et al., supra note 37, at 824; Joshua D. Angrist et al., Explaining Charter School Effectiveness (Inst. for the Study of Labor, Working Paper No. 6525, Apr. 2012). 41 For example, Ohio restricts start-up charter schools to “challenged” districts only, but does not cap conversion charter schools. Charter School Laws Across the States 2013: Ohio, CTR. FOR EDUC. REFORM (Jan. 2013), http://www.edreform.com/wp-content/uploads/2013/01/OH-Charter-Law- 2013.pdf, archived at http://perma.cc/T2ZL-H7JH. In contrast, Colorado does not legally limit the number of charter schools. Charter School Laws Across the States 2013: Colorado, CTR. FOR EDUC. REFORM (Jan. 2013), http://www.edreform.com/wp-content/uploads/2013/01/CO-Charter-Law- 2013.pdf, archived at http://perma.cc/27SU-W9GM. See also infra notes 57–76 and accompanying text. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 287 of charter school policies elsewhere. On the other hand, studies that are more national in scope tend to indiscriminately combine results from different states,42 obscuring the possibility that charter school performance levels vary from state to state as a predictable result of policy differences among the states. If that were so, the proper policy question would not be whether charter or traditional public schools should be preferred because one sector outperforms the other, but which identifiable and replicable features of each sector should policymakers concerned about student outcomes endeavor to encourage or discourage. To date, there have been relatively few studies that compare states on their charter school performance.43 The main exceptions are two particularly influential44 and controversial45 multi-state studies from 2009 and 2013, both 42 See, e.g., Melissa A. Clark et al., Do Charter Schools Improve Student Achievement? Evidence from a National Randomized Study, INST. OF EDUC. SCI. OF THE U.S. DEP’T OF EDUC. 1 (2011). 43 Exceptions include: RON ZIMMER ET AL., CHARTER SCHOOLS IN EIGHT STATES: EFFECTS ON ACHIEVEMENT, ATTAINMENT, INTEGRATION, AND COMPETITION (2009); Caroline M. Hoxby, A Straightforward Comparison of Charter Schools and Regular Public Schools in the United States, HARVARD UNIV. & NAT’L BUREAU ECON. RESEARCH (2004), available at http://www. tidioutecharter.com/pdf/drhoxby_study.pdf, archived at http://perma.cc/ 424U-XNC8; Gary Miron et al., Evaluating the Impact of Charter Schools on Student Achievement: A Longitudinal Look at the Great Lakes States, THE GREAT LAKES CTR. FOR EDUC. RESEARCH & PRACTICE (2007), available at http://greatlakescenter.org/docs/Research/Miron_Charter_Achievement/ Miron_Charter%20Achievement.pdf, archived at http://perma.cc/82EF- QPCL. 44 See Valerie Strauss, The Bottom Line on Charter School Studies, WASH. POST ANSWER SHEET (Sep 24, 2013, 10:00 AM), http://www. washingtonpost.com/blogs/answer-sheet/wp/2013/09/24/the-bottom-line-on- charter-school-studies, archived at http://perma.cc/MD86-YWFH (noting that the CREDO study has “taken on such outsized importance”). 45 For a critique of the methodology used in the 2009 CREDO study, see Caroline M. Hoxby, A Serious Statistical Mistake in the CREDO Study of Charter Schools, STANFORD UNIV. & NAT’L BUREAU ECON. RESEARCH (2009), http://credo.stanford.edu/reports/memo_on_the_credo_study.pdf, archived at http://perma.cc/774D-ZDUX. But see Fact vs. Fiction: An Analysis of Dr. Hoxby’s Misrepresentation of CREDO’s Research, CTR. FOR LIU - FINAL 288 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 published by Stanford University’s Center for Research on Education Outcomes (“CREDO”).46 These studies find that charter school performance appears to have improved over the years,47 but that outcomes nonetheless differ significantly by state.48 For example, the 2013 CREDO study concludes that charter schools on average produce better results in math and reading than traditional public schools in Illinois, Indiana, Louisiana, Massachusetts, Michigan, Missouri, New Jersey, New York, Rhode Island, Tennessee, and the District of Columbia.49 On the other hand, the study finds that charter schools in Arizona, Arkansas, Nevada, Ohio, Oregon, Pennsylvania, Texas, and Utah perform worse in math and reading on average than traditional public schools.50 This inconsistency in charter school performance RESEARCH ON EDUC. OUTCOMES (2009), http://credo.stanford.edu/reports/ CREDO_Hoxby_Rebuttal.pdf, archived at http://perma.cc/U33E-24QT. For a critique of the methodology used in the 2013 CREDO study, see Andrew Maul & Abby McClelland, Review of National Charter School Study 2013, NAT’L EDUC. POLICY CTR. 3–7 (July 2013), http://nepc.colorado.edu/files/ttr- _credo2013.pdf, archived at http://perma.cc/EU8X-WNN4. 46 Multiple Choice: Charter School Performance in 16 States, CTR. FOR RESEARCH ON EDUC. OUTCOMES (2009), http://credo.stanford.edu/reports/ multiple_choice_CREDO.pdf, archived at http://perma.cc/P3TY-DCUD [hereinafter 2009 CREDO STUDY]; Edward Cremata et al., National Charter School Study 2013, CTR. FOR RESEARCH ON EDUC. OUTCOMES (2013), http://credo.stanford.edu/documents/NCSS%202013%20Final%20 Draft.pdf, archived at http://perma.cc/J9Z6-4KTZ [hereinafter 2013 CREDO STUDY]. 47 2013 CREDO STUDY, supra note 46, at 3. 48 Id. at 52–53. See also Marcus A. Winters, Op-Ed., Yes, NYC Charter Schools Are Working, N.Y. DAILY NEWS (Mar. 12, 2014, 5:33 PM), http:// www.nydailynews.com/opinion/yes-nyc-charter-schools-working-article-1.1 719604 (explaining that it is a “blatant mischaracterization” and a “misleading” tactic for charter school opponents to only cite CREDO’s nationwide conclusion that charter schools and traditional public schools are equally effective on average, because CREDO finds that “charter schools’ effectiveness varies dramatically from place to place”). 49 See id. 50 See id. The study also finds some states (e.g., California, Colorado, Georgia, Minnesota, North Carolina) with inconsistent results across subjects—i.e., charter schools perform better than traditional public LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 289 across states provides a further motivation for identifying the differences between states that account for the differential performance of their charter schools. A fruitful place to begin that inquiry is states’ charter school laws, which are the focus of this Note. C. Review of Literature on the Variation in Charter School Laws Charter schools and the laws establishing them are often treated as though they are all the same.51 A study conducted by the Institute of Education Sciences of the U.S. Department of Education exemplifies the assumption that charter school laws are largely uniform. This study analyzes whether charter schools improve achievement by focusing on thirty-six charter middle schools in fifteen states.52 Looking only at charter schools with randomized lottery admissions, the study finds that student performance varies significantly from school to school.53 Even though the study recognizes that differences in environment (e.g., urban vs. suburban vs. rural) may have contributed to the variation in performance, it does not examine the effect that differences in state charter laws and their purposes may have had on student achievement.54 Instead, it appears to assume that state schools in math, but worse in reading, or better in reading, but worse in math. See id. 51 Sarah Yatsko, Buried Treasure: Inside Charter Schools, CTR. ON REINVENTING PUB. EDUC.: THE LENS (Apr. 9, 2014), http:// www.crpe.org/thelens/buried-treasure-inside-charter-schools (“I’ve had conversation after conversation, professional and personal, that shows how people think—still—that all charter schools are the same.”). 52 Melissa A. Clark et al., Do Charter Schools Improve Student Achievement? Evidence from a National Randomized Study, INST. OF EDUC. SCI. OF THE U.S. DEP’T OF EDUC. 1 (2011). 53 Id. (finding that, while in terms of overall average, charter schools improved student outcomes at the same level as traditional public schools, charter school “impacts varied significantly across schools and students, with positive impacts for more disadvantaged schools and students and negative impacts for the more advantaged”). 54 See id. at 2, 4. LIU - FINAL 290 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 charter laws are so similar to one another that their differences need not be taken into account. The study sample simply aggregates data across many states.55 Although every state charter school law embodies the “basic bargain” of providing autonomy to charter schools in exchange for accountability,56 many other studies recognize that charter laws still differ significantly from one another. Sui generis state and local politics drive varying legislative compromises.57 Professor Katrina Bulkley of Montclair State University concludes that charter schools can attract the support of diverse political coalitions for different reasons. Charter schools could easily be called the “all things to all people” reform . . . . Free-market conservatives see them as a way to enhance competition in education and a step in the direction of vouchers. Teachers’ union leaders . . . see them as a way to increase the power of teachers. Cultural conservatives hope that they will increase parental control over the values taught in schools their children attend, while those interested in restructuring schools see them as a way to further their goals. Moderate Democrats hope that charter schools will provide parental choice, competition and 55 Id. at 1. 56 Marc Dean Millot, Autonomy, Accountability, and the Values of Public Education: A Comparative Assessment of Charter School Statutes Leading to Model Legislation, CTR. ON REINVENTING PUB. EDUC. vii–ix (1996). See also Arnold F. Shober, Paul Manna & John F. Witte, Flexibility Meets Accountability: State Charter School Laws and Their Influence on the Formation of Charter Schools in the United States, 43 POL’Y STUD. J. 563, 563 (2006). 57 See, e.g., Michael Mintrom, Policy Design for Local Innovation: The Effects of Competition in Public Schooling, 1 ST. POL. & POL’Y Q. 343, 345 (2001) (restricting the study to a single state “in part because the details of charter school laws differ significantly among the states”); Amy Stuart Wells et al., Defining Democracy in the Neoliberal Age: Charter School Reform and Educational Consumption, 39 AM. EDUC. RES. J. 337, 345 (2002); Thomas Mauhs-Pugh, Charter Schools 1995: A Survey and Analysis of the Laws and Practices of the States, 3 EDUC. POL’Y ANALYSIS ARCHIVES (1995), http://epaa.asu.edu/ojs/article/viewFile/656/778, archived at http://perma.cc/6WNX-BLAC. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 291 accountability while avoiding actual privatization through school vouchers.58 As a result, a charter school law may not easily be labeled as “Republican” or “Democrat.”59 Professor Michael Kirst of Stanford University has built on this research in an analysis of groups that support and groups that oppose charter laws, concluding that “[i]t is hard to generalize about charter politics because of the extreme variations among 50 states and thousands of school districts.” Because “[t]he United States is a nation of states, each with its distinctive politics,” differences in charter school policy among states are “vast.”60 For example, Kirst finds that, while Michigan’s charter school legislation was designed to “create competition with traditional public schools,” Georgia’s legislation was meant to “deregulate and decentralize education” to counter “a period of increasing centralization to the state.”61 D. Review of Literature Evaluating Charter School Laws Given the variation in state charter school laws and the inconsistency in charter school outcomes, to provide recommendations for improving performance, education reform advocacy organizations and scholars have developed different methods for evaluating the quality of a state’s charter school law.62 58 Katrina Bulkley, Understanding the Charter School Concept in Legislation: The Cases of Arizona, Michigan, and Georgia, 18 INT’L J. QUALITATIVE STUD. EDUC. 527, 527 (2005). 59 Kathryn A. McDermott, What Causes Variation in States’ Accountability Policies?, 78 PEABODY J. EDUC. 153, 172 (2003). 60 Michael W. Kirst, Politics of Charter Schools: Competing National Advocacy Coalitions Meet Local Politics, 82 PEABODY J. EDUC. 184, 189, 199–200 (2007) (“There is not a cohesive state or local charter political pattern given the variations in charter schools and their contexts.”). 61 Id. at 187. 62 In addition to evaluation methods that focus solely on the quality of charter school legislation, there are many assessments that “grade” each state’s overall educational system and include charter school policy in their analysis. See, e.g., State Policy Report Card 2014: The National LIU - FINAL 292 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 One influential method of “grading” state charter laws is the Center for Education Reform’s (“CER”) Annual Charter Law Rankings.63 CER highly values laws that allow for greater independence and flexibility for charter schools (i.e., encourages a free-market system).64 It ranks state policy based on four components: (1) the existence of independent and/or multiple authorizers; (2) the number of schools allowed and state caps; (3) operational and fiscal autonomy; and (4) equitable funding.65 Since 2009, the National Alliance for Public Charter Schools (“NAPCS”) has promulgated a different annual charter law ranking system.66 NAPCS sought input from charter stakeholders to develop its own version of a “model” law that includes twenty “essential components,” ranging from providing for multiple authorizers to requiring a strong authorizer accountability Report, STUDENTSFIRST 5–6, 53–64 (2014), http://reportcard.studentsfirst. org/assets/2014NationalReport.pdf, archived at http://perma.cc/8MWJ- D4YR; Quality Counts 2014: District Disruption & Revival, EDUC. WEEK, Jan. 9, 2014, available at http://www.edweek.org/ew/toc/2014/01/09/ index.html, archived at http://perma.cc/W6R8-YJWM. These kinds of evaluations are often promulgated by education reform organizations advocating for specific policy goals, such as school choice, test-based accountability, and centralization of governance. See Sherman Dorn & Ken Libby, Review of State Policy Report Card, NAT’L EDUC. POL’Y CTR. 1, 6 (Feb. 2013), http://nepc.colorado.edu/files/nepc-ttr-stdfirst-grades-rhee.pdf, archived at http://perma.cc/2F2R-9CRE. 63 See 2012 CER RANKINGS, supra note 16. 64 Wendy C. Chi & Kevin G. Welner, Charter Ranking Roulette: An Analysis of Reports that Grade States’ Charter School Laws, 114 AM. J. EDUC. 273, 273 (2008). 65 2012 CER RANKINGS, supra note 16, 6–7; see also Understanding Charter School Laws and How They Are Ranked, CTR. FOR EDUC. REFORM (Feb. 5, 2013), http://www.edreform.com/2013/02/understanding-charter- school-laws-and-how-they-are-ranked/, archived at http://perma.cc/742Z- FPJJ. 66 Todd Ziebarth & Louann Bierlein Palmer, Assessing the Increasing Strength of Charter Laws Between 2010 and 2013, NAT’L ALLIANCE FOR PUB. CHARTER SCH. 2 (July 30, 2013), http://www.publiccharters.org/wp- content/uploads/2014/01/Charter-School-Law-Strength_20130730T113930. pdf, archived at http://perma.cc/8LXJ-8SP8. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 293 system.67 NAPCS then ranks state charter laws based on the extent to which they coincide with or deviate from the NAPCS model law.68 While CER rankings are generally proportional to how free charter schools are from state regulation,69 NAPCS gives greater preference to how effectively state laws enforce quality control and accountability.70 These “grades” of charter school laws generally promote what each organization itself views as important to developing “strong” charter schools. Professors Wendy Chi and Kevin Welner at the University of Colorado at Boulder criticize such charter law evaluation systems as ignoring the wide variety of purposes that motivated the charter school movement in the first place.71 They developed a competing system that ranks a state based on how closely its components adhere to the following seven goals: (1) instructional innovation; (2) regulatory waivers; (3) maintenance of the public nature of charter schools; (4) increased access to opportunities for at- risk students; (5) performance-based accountability; (6) increased student achievement; and (7) promotion of best practices through evaluation of initial small-scale efforts.72 Chi and Welner derived these goals from a review of early charter school advocacy and the rationales stated in state charter school laws.73 Unlike the proponents of these other evaluation methods, Western Michigan University Professor Gary Miron has tried to determine the characteristics of “strong” charter 67 Measuring Up to the Model: A Ranking of State Charter School Laws, NAT’L ALLIANCE FOR PUB. CHARTER SCH. 6–7 (Jan. 2014), available at http://www.publiccharters.org/wp-content/uploads/2014/01/StateRankings 2014.pdf, archived at http://perma.cc/4VEY-VN2Z. 68 Id. 69 See supra notes 64–65 and accompanying text. 70 Measuring Up to the Model: A Ranking of State Charter School Laws, supra note 67, at 6–7. 71 See Chi & Welner, supra note 64, at 275–76. 72 See id. at 282. 73 See id. LIU - FINAL 294 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 school laws more empirically by analyzing six states.74 He argues that laws should lead to positive outcomes (e.g., improved student performance) and minimize negative outcomes (e.g., re-segregation of schools by race).75 Miron praises state charter laws with the following components and history: (1) rigorous approval process; (2) rigorous oversight; (3) provision of technical assistance; (4) limited role for EMOs; (5) adequate financial support; (6) rapid rate of charter school sector growth; and (7) bipartisan support.76 This Note proposes a different approach for determining what effective charter school legislation looks like: first categorize state charter school laws based on their legislative purposes, then determine whether these purposes correlate with achievement outcomes, and finally, derive the key components of laws in each purpose category that may affect outcomes. Evaluations of the quality of state charter school laws by CER, NAPCS, and Miron have generally determined what components (e.g., no teacher certification requirement) should be included in laws based on what they themselves have determined should be charter schools’ goals. They do not analyze what purposes may actually motivate the differences in laws and meaningfully categorize states accordingly. Chi and Welner do look at purpose, but in grading charter school laws, they make the assumption that all of the laws must reflect all seven of the goals that they believe have motivated various aspects of the charter school movement. They generally provide policy recommendations without necessarily basing them on achievement outcomes. No study has yet provided a framework that categorizes state laws based on purpose to explain the inconsistency in charter school performance, and subsequently derive policy recommendations for improving student achievement. 74 Gary Miron, Strong Charter School Laws Are Those that Result in Positive Outcomes, AM. EDUC. RESEARCH ASS’N 2 (Apr. 2005), http://homepages.wmich.edu/~miron/publics/aera_2005_paper_charter_sch ool_laws.pdf, archived at http://perma.cc/VS32-A3NL [hereinafter Miron, Strong Charter School Laws]. 75 Id. at 1. 76 Id. at 4–8. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 295 III. STUDY METHODOLOGY This study attempts to (1) develop a framework for categorizing state charter school laws based on their underlying purposes, (2) assess charter school student achievement outcomes associated with each category of law, and (3) determine the typical provisions of each category of law to provide charter school policy recommendations for improving student outcomes.77 A. Development of the Framework Based on Legislative Purpose This Note evaluates the hypothesis that there are two principal purposes that differentiate, and underlie the passage of, state charter school laws. The first is to close achievement gaps by improving the outcomes of at-risk student populations in particular (i.e., “gap-closing states”). The second is to provide parents and local communities with greater choice and control in education (i.e., “libertarian- oriented states”). In addition, there may be states that promote both purposes (i.e., “mixed states”). Given the diverse political coalitions78 that have come together to establish charter schools, it makes sense that these coalitions may have sought to use the same vehicle— charter schools—to accomplish potentially conflicting goals. Moreover, especially in light of such inconsistent charter school performance,79 it is possible that improving student outcomes may simply have not been the sole or main overarching purpose for all charter schools. By offering 77 This Note focuses on student test scores, and not what effect charter schools may have in other areas, such as college enrollment, incidence of risky behavior, and health. See Will Dobbie & Roland G. Fryer, Jr., The Medium-Term Impacts of High-Achieving Charter Schools on Non-Test Score Outcomes (Nat’l Bureau of Econ. Research, Working Paper No. 19581, 2013). 78 See, e.g., Chris Pipho, Bipartisan Charter Schools, 75 THE PHI DELTA KAPPAN 102, 103 (1993). See also supra notes 57–61 and accompanying text. 79 See supra Part II.B. LIU - FINAL 296 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 greater autonomy and accountability, charter schools can meet the divergent needs of both those seeking to depart from traditional teaching methods to address the difficult challenge of closing significant achievement gaps, and those seeking to exert greater parental and local community control. To test this hypothesis, this Note first analyzes the forty- three80 state81 charter school laws in place to identify their underlying purposes. The statement of purposes specified in each statute provides a primary source of information about the law’s purposes. Although not all purposes are necessarily codified in a statute,82 this declaration of legislative intent is nevertheless strong evidence of the motivation behind the passage of charter school legislation.83 However, some states do not include a statement of purpose in their legislation for establishing charter schools.84 Additionally, this Note examines whether state charter laws contain provisions that implicitly suggest the legislative purpose. Specifically, these include (1) provisions requiring or permitting preferences to be given to “at-risk” student 80 Alabama recently became the latest state to have a charter school law, see supra note 23, but it is excluded from analysis due to publication time constraints. 81 For ease of language, throughout this Note, references to “states” include the District of Columbia. 82 See MILLOT, supra note 56, at 7 (noting that while imperfect, statements of legislative intent are still valuable because courts often use them to help interpret other provisions of the statute). 83 The declaration of legislative intent is furthermore important because it is the actual text that a legislature agreed and voted on. See, e.g., Kenneth R. Dortzbach, Legislative History: The Philosophies of Justices Scalia and Breyer and the Use of Legislative History by the Wisconsin State Courts, 80 MARQ. L. REV. 161, 186 (1996) (citing Green v. Bock Laundry Mach. Co., 490 U.S. 504, 528 (1989) (Scalia, J., concurring)). 84 See, e.g., Courtney L. Malloy & Priscilla Wohlstetter, Working Conditions in Charter Schools: What’s the Appeal for Teachers?, 35 EDUC. & URB. SOC’Y. 219, 221 (2003) (noting that Alaska, Mississippi, Ohio, Texas, and the District of Columbia do not specify purposes of charter schools in their respective laws). LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 297 populations85 in the charter school application or enrollment process and (2) provisions relating to parental involvement.86 At-risk preferences imply an intent to serve those at-risk populations in particular. States with these preferences are more likely to fall under the gap-closing states category. By contrast, parental involvement provisions imply an intent to provide parents with greater control over education. States with many of these provisions are more likely to fall under the libertarian-oriented states category. Lastly, any academic literature relating the circumstances of a state charter school law’s enactment provides a secondary source for analyzing the purposes of the law and determining how it should be categorized. B. Assessment of the Correlation Between Legislative Purpose and Outcomes To determine whether a framework categorizing state charter laws based on legislative purpose can explain the variation in charter school performance seen across states, this Note assesses whether or not outcomes are correlated with purpose. For this analysis, this Note uses the performance data from the 2013 CREDO National Charter School study, which is the most comprehensive nationwide study on charter school impact to date.87 The study examines 85 Examples include academically low-achieving students, drop-outs, low-income students, English Language Learners, and students with disabilities. See, e.g., ME. REV. STAT. ANN. tit. 20-A § 2401(1) (2011) (“‘At- risk pupil’ means a pupil who has an economic or academic disadvantage that requires special services and assistance to enable the student to succeed in educational programs. ‘At-risk pupil’ includes, but is not limited to, pupils who are members of economically disadvantaged families, pupils who are identified as having special educational needs, pupils who are limited in English proficiency, pupils who are at risk of dropping out of high school and pupils who do not meet minimum standards of academic proficiency.”). 86 See, e.g., D.C. CODE § 38-1802.05(a)(2) (2011) (requiring that the Board of Trustees of a public charter school include at least two members who are parents of students attending the school). 87 2013 CREDO STUDY, supra note 46, at 2. See supra note 45 for past critiques of this study. LIU - FINAL 298 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 how the 2008–2011 academic growth88 of charter school students in reading and math compares to that of similar students that attended traditional public schools.89 Academic growth is determined by the longitudinal change in a student’s test scores from one school year to the next and is represented in terms of days of learning added each year (relative to a traditional public school student).90 The CREDO study covers twenty-six91 of the forty-three states with charter laws, including the District of Columbia.92 Although a study that covers all forty-three states would provide a more robust sample size, the twenty-six states 88 To facilitate comparison of charter school performance across states, the 2013 CREDO study uses academic growth, instead of achievement test scores, for measuring charter impact. This preference is due to the significant variability in where states start off in terms of their average charter student test scores. For example, Missouri has the lowest average test scores in both math and reading, while North Carolina has the highest scores. See id. at 5–6, 21. 89 Id. at 12. 90 The actual outputs of statistical methods used for each analysis are test scores expressed in terms of standard deviation from the statewide average academic performance of all tested students (i.e., a score of 0 corresponds to the 50th percentile of performance in the state). However, the CREDO study includes a transformation of results into days of learning to make the data more meaningful to non-technical readers. For example, a standard deviation of 0.01 converts to seven days of learning. A school week equals five days of learning, and a school month equals twenty. Id. at 12–13, 21. 91 The 2013 CREDO study separately analyzes charter school impacts for New York City and the rest of the state of New York because the demographics, performance, and size of New York City make it a “unique sub-population.” 2013 CREDO STUDY, supra note 46, at 3 n.4, 7. The rest of the state of New York (where charter students had seventy-nine more days of learning in math and thirty-six more days in reading than traditional public school students) and New York City (where charter students had ninety-four more days in math and comparable performance in reading) still see similarities in results, however—charter schools generally have a relatively positive impact. Id. at 53. See also Winters, supra note 48. This Note focuses on the rest of the state of New York for analysis, but using New York City data also does not significantly change the analysis. See infra note 165. 92 2013 CREDO STUDY, supra note 46, at 4, 7. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 299 covered here nevertheless educate over ninety-five percent of all charter school students.93 Overall, the data from the CREDO study cover seventy-nine percent of all (charter and traditional) public school students.94 Thus, while there is room for additional study beyond the states covered here, this Note nonetheless provides a suggestive analysis of the charter school performance.95 C. Analysis of Key Charter School Law Features by Legislative Purpose Finally, to understand what policies legislatures should adopt in order to improve charter school outcomes, this Note determines the key features of charter school laws that can be found in each legislative purpose category (i.e., gap- closing, libertarian-oriented, and mixed). Table One lists the charter school law components this Note analyzes. They were selected because they have been considered relevant to charter school outcomes and have generally attracted the most interest from policymakers, education reform advocacy organizations, and scholars. These components appear to fall into four main groups: (1) charter school sector flexibility (overall strength of the charter school market); (2) charter school autonomy (degree of freedom that each school has over operations, curriculum, etc.); (3) accountability (extent to which both authorizers and schools are held accountable for outcomes); and (4) funding (school’s financial sustainability). 93 Id. at 2. 94 Id. at 4. 95 There are a few other multi-state charter school achievement studies, but as they cover fewer than ten states, they do not provide a sufficient number of data points for a robust analysis and generally do not cover states beyond those that the 2013 CREDO study already covers. See supra note 43 and accompanying text. Moreover, this Note does not combine one-state or one-city studies to create a dataset for analysis because the differences in these studies’ methodologies and areas of focus prevent an apples-to-apples comparison of charter school performance across states. LIU - FINAL 300 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE ONE: KEY CHARTER SCHOOL LAW FEATURES FOR ANALYSIS Category Charter School Law Features Charter School Sector Flexibility  Multiple Authorizers: Availability of a range of authorizing options, from local school boards to municipalities, and an appeals process if an authorizer denies a charter school application  Caps: Existence of limits on the number of charter schools or students  Types of Schools: Options for start-up, conversion, or virtual schools  Education Service Providers: Involvement of for-profit education management organizations and nonprofit charter management organizations in operating charter schools Charter School Autonomy  Independence from State and Local Regulations: Automatic, partial, or case-by-case exemptions of state and local regulations that apply to traditional public schools  Collective Bargaining Agreement Requirements: Automatic, partial, or case-by-case exemptions from being subject to district collective bargaining agreements  Teacher Certification Requirements: Automatic, partial, or case-by-case exemptions from being subject to teacher certification requirements  Statewide Retirement System Requirements: Requirement for all or some teachers to participate in the statewide retirement system, or prohibition on participation LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 301 TABLE ONE: KEY CHARTER SCHOOL LAW FEATURES FOR ANALYSIS (CONT.) Category Charter School Law Features Account- ability  Authorizer Accountability: Existence of oversight body that conducts regular reviews and has authority to sanction authorizers on performance  Overall Charter School System Accountability: Periodic evaluations of charter school programs and outcomes  Charter School Accountability: Oversight of charter schools to ensure accountability for financial and student achievement outcomes, notification to schools of problems, opportunity for schools to remedy problems, and authority to take corrective actions short of revocation Funding  Operational Funding: Funding for operations, access to categorical federal and state grants, and funding for student transportation that are comparable to what are provided to traditional public schools  Capital Funding and Facilities Access: Equitable funding and access to school property LIU - FINAL 302 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 First, with respect to components affecting charter school sector flexibility, the availability of authorizers, existence of caps, types of schools, and the role of educational service providers are highly contentious issues. Policymakers, education reform advocacy organizations, and scholars have all spent considerable time debating and evaluating them. Authorizers96—from which charter school applicants must obtain approval before they can establish schools—play a crucial role as a gatekeeper.97 They also often are considered important for providing ongoing monitoring of charter school performance and for deciding whether to revoke or renew a charter.98 Caps99 on the number of charter schools or 96 Examples of past state charter law evaluations by education reform advocacy organizations that had the availability of authorizers as a criterion include: 2012 CER RANKINGS, supra note 16, at 6; Measuring Up to the Model: A Ranking of State Charter School Laws, NAT’L ALLIANCE FOR PUB. CHARTER SCH. 6 (Jan. 2012), http://www.publiccharters.org/wp- content/uploads/2014/01/NAPCS_2012_StateLawRankings_Final_2012011 7T162953.pdf [hereinafter 2012 NAPCS RANKINGS]. There has also been significant scholarly interest on this topic. See, e.g., Mead & Rotherham, supra note 9, at 3–5. 97 See, e.g., 2009 CREDO STUDY, supra note 46, at 38–40 (“Empowering entities outside the traditional public K-12 education milieu to grant charters provides diversity in perspective about school choice and can bring a wider range of knowledge and experience to bear on charter growth. At the same time, authorizers exercise a considerable degree of latitude in their practice, which may present schools with different opportunity costs when given a choice of authorizers, and the response mechanism of schools is unclear.”). 98 Louann Bierlein Palmer & Rebecca Gau, Charter School Authorizing: Are States Making the Grade?, THOMAS B. FORDHAM INST. 1 (2003), available at http://repository.asu.edu/attachments/55818/content/ Palmer2003.pdf, archived at http://perma.cc/2AN9-428Q (“The operator must run a successful school that delivers the results it promised. The authorizer must see that this happens, providing various forms of oversight and assistance, renewing the charter if all goes well—and pulling the plug if it does not. The role of the authorizer, therefore, is pivotal to the charter movement’s overall success.”). 99 Examples of past state charter law evaluations by education reform advocacy organizations that had caps on charter school growth as a criterion include: 2012 CER RANKINGS, supra note 16, at 6; 2012 NAPCS RANKINGS, supra note 96, at 6. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 303 students permitted may have a number of different effects on outcomes, including “generat[ing] pressure to allocate charters to maximize the chance of high performance” and “raising a barrier to entry that sends talented or successful operators to less restrictive states.”100 And while there appears to be a consensus about the value of permitting start-ups, conversions, and nonprofit CMOs, the impact of virtual schools101 and for-profit EMOs102 continues to be disputed.103 The next two sets of charter law features for analysis relate to the fundamental concept of charter schools. Providing charter schools greater autonomy in exchange for greater accountability has long been considered the key means by which charter schools could produce better outcomes than traditional public schools.104 Unsurprisingly, the components affecting the degree to which states should exempt charter schools from various regulations105 and hold 100 2009 CREDO STUDY, supra note 46, at 38–40. 101 See, e.g., Mead & Rotherham, supra note 9, at 9 (“Virtual schools can be innovative and engage students who might otherwise not be in school at all, but they also create new policy challenges, especially in the areas of funding, teacher quality, and accountability. And financial and enrollment scandals . . . have exacerbated policymakers and the public’s pre-existing suspicions about such unconventional schools.”). 102 See, e.g., 2013 CREDO STUDY, supra note 46, at 50–51 (noting that there are many ongoing debates about charter management organizations); Mead & Rotherham, supra note 9, at 7–9 (“The experience of [states with EMOs] suggests that there are both pros—increased capital, scalability, and quality control across multiple sites—and cons— less innovation, reduced community control, potential for conflicts of interest and other scandals—to having a largely EMO-run charter sector.”). 103 Examples of past state charter law evaluations by education reform advocacy organizations that had types of schools and education service providers as criteria include: 2012 CER RANKINGS, supra note 16, at 6; 2012 NAPCS RANKINGS, supra note 96, at 6. 104 See, e.g., Chi & Welner, supra note 64, at 282. 105 Education reform advocacy organizations have often “graded” state charter school laws on how much they free charter schools from state and local regulations and requirements on collective bargaining, teacher certification, and retirement. 2012 CER RANKINGS, supra note 16, at 6; LIU - FINAL 304 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 both authorizers and charter schools accountable106 have generated substantial interest and discussion. As such, as Table One shows, this Note analyzes many of these components. However, the duration of an initial charter term is omitted.107 There is such significant variation in state charter school laws that a meaningful comparison is not feasible.108 Last but not least, like past research that have evaluated the quality of charter school laws, this Note examines the adequacy of funding to support charter school operations, student transportation, and facilities needs.109 Whether or not charter schools need to, or should, receive as much funding as traditional public schools do has been the subject of much debate.110 2012 NAPCS RANKINGS, supra note 96, at 6. But some scholars have questioned the value of autonomy. See Ron Zimmer & Richard Buddin, Getting Inside the Black Box: Examining How the Operation of Charter Schools Affects Performance, 82 PEABODY J. EDUC. 231, 271 (2007) (“Our analysis suggests that although charter school principals do have greater control over decision making than their counterparts in TPSs, we found no strong evidence that this autonomy leads to higher test scores.”). 106 Accountability provisions have been an important part of the 2012 NAPCS RANKINGS, supra note 96, at 6. 107 MILLOT, supra note 56, at x, xiii–xiv (“Limiting the duration of a charter school’s contract promotes accountability but constrains autonomy.”).7 108 States may specify the precise duration of an initial charter term, prescribe a range of years, set a minimum number of years, set a maximum number of years, not specify any duration, or do any combination of these for different types of charter schools. See infra Appendix Table B. 109 Examples of past state charter law evaluations by education reform advocacy organizations that had funding as a criterion: 2012 CER RANKINGS, supra note 16, at 7; 2012 NAPCS RANKINGS, supra note 96, at 6. 110 Miron, Strong Charter School Laws, supra note 74, at 6–7 (“[S]tates that can insure that more resources get to the charter schools in a timely fashion are more likely to have successful and strong charter schools. Some early charter school promoters made sweeping assumptions about charter schools being able to do a better job with less money. The expectations that charter schools would be innovative and serve as research and development units for public education were naive, given LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 305 For the charter school features listed in Table One, a database on every state’s key provisions was developed by examining the language of the laws themselves, utilizing the charter law ranking reports from the Center for Education Reform111 and the National Alliance for Public Charter Schools,112 as well as referencing the profiles of state charter school policies from the Education Commission of the States.113 Because this Note utilizes the 2013 CREDO study data on state charter school impacts through 2011, it also examines laws as they existed in 2011.114 that charter schools typically receive the same or less money as traditional public schools.”). 111 2012 CER RANKINGS, supra note 16. 112 2012 NAPCS RANKINGS, supra note 96. 113 State Profiles—Charter Schools Database, EDUC. COMM’N OF THE STATES (Oct. 2010), http://ecs.force.com/mbdata/mbchsmap, archived at http://perma.cc/Z6UC-TXWU [hereinafter ECS STATE PROFILES]. The Education Commission of the States is a “nationwide, nonpartisan interstate compact devoted to education.” Id. 114 This approach may not precisely capture which provisions in state charter school laws may have affected the charter school outcomes that the 2013 CREDO study finds. While states rarely amend the legislative statements of purpose in their charter school laws, states are more likely to amend other provisions, such as on authorizer accountability. See generally Robert A. Fox & Nina K. Buchanan, A State Charter School Law in Transition, 1 J. SCH. CHOICE 145, 171 (2006) (“Legislation is a dynamic, rather than a static, thing.”). See also Kerry A. King, Charter Schools in Arizona: Does Being a For-Profit Institution Make a Difference?, 41 J. ECON. ISSUES 729, 732 tbl. 1 (2007) (listing when state charter laws were most recently amended, regardless of how major or minor an amendment was). This Note nonetheless provides a suggestive analysis on charter school legislation because meaningful, substantive changes to charter school laws do not occur constantly. See Shober et al., supra note 56, at 568 tbl. 1. A relatively small number of states may amend their charter school legislation in a given year. For example, less than a quarter of the states with charter laws partially or entirely lifted caps on their charter school growth in 2011. See 2012 NAPCS RANKINGS, supra note 96, at 3. Moreover, “many state legislatures have left their charter school laws untouched for a decade or more to avoid reopening politically contentious debates.” JULIET SQUIRE ET AL., THE ROAD TO REDEMPTION: TEN POLICY RECOMMENDATIONS FOR OHIO’S CHARTER SCHOOL SECTOR 6 (2014). LIU - FINAL 306 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 This Note provides a suggestive—rather than an absolutely decisive—analysis on charter school legislation purposes and their correlation with charter school outcomes. There may be other factors that could be determinative, but that this study has not identified and ruled out. There should be further research and comprehensive empirical analysis in this area. For example, this Note assumes that the legislative intent and the various components of charter school laws (e.g., providing equitable funding, increasing school autonomy) are faithfully carried out. In other words, the assumption is that there is no inherent execution problem that would help explain inconsistent or poor charter school performance. This assumption is supported by a 2011 nationwide study conducted by University of Pennsylvania Professors Katherine Barghaus and Erling Boe.115 They conclude that “charter schools have been implemented much as intended by legislation” and that lack of improvement in student achievement “cannot be attributed to a failure to implement the charter school concept.”116 However, an assumption that implementation is faithful may be risky since implementation oftentimes does deviate from intent. In California, a referendum passed in 2000 required school districts to provide charters with “equitable access to school facilities.”117 Yet, “districts have often subverted the law’s intent.”118 Similarly, “[b]oth Congress and the District of Columbia City Council have passed legislation to give charter schools access to millions of square feet in unused public school space held by the District of Columbia Public Schools, but the city’s Board of Education has resisted releasing the space to charter schools.”119 115 Katherine M. Barghaus & Erling E. Boe, From Policy to Practice: Implementation of the Legislative Objectives of Charter Schools, 118 AM. J. EDUC. 57, 57 (2011). 116 Id. 117 Mead & Rotherham, supra note 9, at 13. 118 Id. 119 Id. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 307 In sum, further research should be conducted on other possibly determinative factors such as poor implementation, but this Note nonetheless provides a suggestive analysis on charter school legislation and outcomes. V. AN ORIGINAL FRAMEWORK: GAP-CLOSING VERSUS LIBERTARIAN-ORIENTED STATES This Note concludes that the data support the hypothesis that state charter school laws may be distinguished on the basis of their legislative purpose: (1) “gap-closing” states emphasize closing achievement gaps by improving the outcomes of low-achieving or at-risk student populations in particular; (2) “libertarian-oriented” states emphasize providing parents and local communities with greater control in education; and (3) “mixed” states promote both purposes. A. Determination of How to Best Categorize Legislative Purposes In order to understand whether state charter school laws may be categorized as gap-closing, libertarian-oriented, or mixed, this Note first examines the range and types of purposes specified in legislation. Nearly all states specify more than one purpose. As Table Two shows, purposes can range from providing greater choice to improving student achievement for at-risk populations. This Note classifies most of these purposes as generally relating to the broader goals of increasing local control and flexibility, ensuring accountability, and improving student achievement. LIU - FINAL 308 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE TWO: FREQUENCY OF MENTION OF LEGISLATIVE PURPOSE IN CHARTER LAWS (n=36)120 Legislative Purpose # States (%) (1) Increase Local Control and Flexibility (a) Facilitate innovation in teaching, governance, etc. 34 (94%) (b) Provide greater choice and increase competition 31 (86%) (c) Create professional development opportunities for teachers 30 (83%) (d) Deregulate and provide greater autonomy 20 (56%) (e) Increase parental and community involvement 13 (36%) (2) Ensure Accountability and Measure Outcomes 32 (89%) (3) Improve Achievement (a) Enhance student learning in general 35 (97%) (b) Improve the existing public education system overall 22 (61%) (c) Serve low-achieving or at-risk student populations 17 (47%) (4) Other121 11 (31%) This Note finds that, of the nine purposes that state charter school laws commonly specify, five do not provide an effective means to distinguish between charter laws because of how frequently they are specified. The vast majority (i.e., greater than eighty percent) of states seek to: (1)(a) facilitate innovation in teaching and governance; (1)(b) provide greater 120 Besides the seven states without charter school laws (i.e., Kentucky, Montana, Nebraska, North Dakota, South Dakota, Vermont, and West Virginia) and Alabama’s recently enacted charter school law, see supra note 80, Table Two excludes the seven states with charter school laws that do not specify their purposes (i.e., Alaska, Connecticut, Hawaii, Michigan, Missouri, Ohio, and Wisconsin). 121 See, e.g., FLA. STAT. § 1002.33(2) (2011) (Charter schools may fulfill the purpose of “[m]itigat[ing] the educational impact created by the development of new residential dwelling units.”). LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 309 choice and increase competition; (1)(c) create professional development opportunities for teachers; (2) ensure accountability and measure outcomes; and (3)(a) enhance student learning in general.122 The four other purposes, each of which less than two- thirds of state charter school laws specify, are much more informative for distinguishing the laws, and they confirm this study’s hypothesis. The purpose of serving low-achieving or at-risk student populations ((3)(c))123 strongly weighs in favor of concluding that a key overarching goal of state charter school laws is “gap-closing.” In addition, a law that specifies the purpose of improving the existing public education system overall ((3)(b)),124 which would benefit all public school students, including those lacking the resources to attend a better school, slightly favors categorizing it as gap-closing. Specifying this purpose also suggests a desire to improve student achievement from within the public education system, rather than to shift gradually to a private education system that is free from governmental control. In comparison, the purposes of deregulating and providing greater autonomy ((1)(d))125 and increasing parental and 122 The near universality of these purposes reflects what the charter school movement has long been best known for. See, e.g., Eric A. Hanushek et al., Charter School Quality and Parental Decision Making with School Choice, 91 J. PUB. ECON. 823, 823–24 (2007) (“Charter schools have been championed as the politically feasible form of school choice that offers most of the advantages of school voucher without sacrificing the benefits of government oversight. The freedom from many of the constraints under which regular public schools operate allows for a diversity of educational approaches and increased competition within the public sector.”). 123 See, e.g., N.Y. EDUC. LAW § 2850(2)(b) (McKinney 2011) (listing one of the law’s objectives as “[i]ncreas[ing] learning opportunities for all students, with special emphasis on expanded learning experiences for students who are at risk of academic failure”). 124 See, e.g., LA. REV. STAT. ANN. § 17:3972(B)(1)(a) (2011) (listing one of the law’s objectives as “[i]mprov[ing] pupil learning and, in general, the public school system”). 125 See, e.g., N.H. STAT. § 194-B:1-a(V) (2011) (listing one of the law’s objectives as “[e]xempt[ing] charter schools from state statutes and rules, other than where specified, to provide innovative learning and teaching in a unique environment”). LIU - FINAL 310 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 community involvement ((1)(e))126 both support finding that another key overarching goal of state charter school laws is “libertarian-oriented.” They advocate for placing greater power in the hands of local community members including parents, students, and teachers. Using this framework of gap-closing, libertarian-oriented, and mixed states, to assign each law to its appropriate purpose category, this Note also analyzes other relevant provisions—those concerning at-risk preferences or parental involvement. Table Three shows that, of the laws that provide for any preferences in charter school applications or enrollment procedures (e.g., for a sibling of student currently enrolled in the charter school,127 for local area residents128), a majority either permit or mandate preferences for at-risk student populations. Mandating such a preference strongly supports deeming a state charter school law to be gap- closing. TABLE THREE: PROVISIONS IN CHARTER LAWS REQUIRING / PERMITTING AT-RISK PREFERENCES (n=43)129 Preferences in Charter School Applications or Enrollment # States (%) No Preferences Indicated 4 (9%)130 Overall – Required / Permitted At-Risk Preference 30 (77%)131 Required Preference 14 (47%) Permitted Preference 16 (53%) 126 See, e.g., UTAH CODE § 53A-1a-503(6) (2011) (listing one of the law’s objectives as “provid[ing] opportunities for greater parental involvement in management decisions at the school level”). 127 For example, Arizona requires a preference for siblings of a child currently enrolled in the school. ECS STATE PROFILES, supra note 113. 128 For example, Georgia requires a preference for local area residents. ECS STATE PROFILES, supra note 113. 129 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; 2012 CER RANKINGS, supra note 16; and ECS STATE PROFILES, supra note 113. 130 Out of the 43 states with charter school laws. 131 Out of the 39 states with charter school laws that mention any required or permissible preferences (e.g., for siblings). LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 311 Table Four summarizes the many different provisions relating to parental involvement in charter school legislation and favoring categorizing a law as libertarian-oriented. Results show that these provisions vary widely in the degree to which parents may influence the direction of educational programs. Provisions that institutionalize parental influence in a school’s governance or operational structure (e.g., requiring parents to be members of charter schools’ governing boards132) are given significantly greater weight than provisions contemplating relatively minimal parental engagement (e.g., merely requiring that teachers’ resumes be made available to parents133). The raw number of provisions supporting parental involvement also is taken into consideration. On average, a state charter law includes three such provisions. All states but one (Maryland134) have at least one provision relating to parental involvement. The most common parental involvement provisions in state charter laws are the requirements for parent, teacher, or other community support in order to establish a charter school135 and for a charter application to include a plan for parental involvement.136 132 DEL. CODE ANN. tit. 14 § 512(1) (2011) (“At the time at which the school commences its instructional program and at all times thereafter, the board of directors must include a teacher from at least 1 of the charter schools operated by the board and at least 1 parent of a student enrolled in a charter school operated by the board”). 133 TEX. EDUC. CODE ANN. § 12.130 (West 2011) (“Each open- enrollment charter school shall provide to the parent or guardian of each student enrolled in the school written notice of the qualifications of each teacher employed by the school.”). 134 MD. EDUC. CODE ANN. §§ 9-101 to -112 (West 2011). 135 See, e.g., KAN. STAT. ANN. § 72-1923(b) (2011) (“The application shall include the following: . . . (2) a description of the interest and support for partnerships between the public innovative district, parents and the community.”). 136 See, e.g., MISS. CODE. ANN. § 37-28-15(4) (West 2011) (“In addition to all other requirements, the request for proposals must require charter applications to provide or describe thoroughly all of the following mandatory elements of the proposed school plan: . . . (x) Opportunities and expectations for parent involvement.”) LIU - FINAL 312 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE FOUR: PROVISIONS IN CHARTER LAWS SUGGESTING PARENTAL INVOLVEMENT IN OPERATIONS (From High to Low Degree of Influence) (n=43) Provision Relating to Parental Involvement # States (%) (1) Institutionalized Requirements – Mandatory Parental Influence (a) Governing Board / Board of Directors Membership 11 (26%) (b) Authorizing Body Membership 4 (9%) (c) Revocation / Termination of Charter Participation 6 (14%) (d) Parent / Community Support to Establish / Convert School 33 (77%) (2) Formalized Requirements – Advisory Parental Influence (a) Advisory Council / Parent-Teacher Council Membership 10 (23%) (b) Public Meetings Participation 13 (30%) (c) Parental Grievance Procedures 6 (14%) (3) Soft Requirements – Encouraged Parental Influence (a) Parental Involvement Plan in Charter Application 26 (60%) (b) Designated Representative to Facilitate Parental Involvement 2 (5%) (c) Training for Staff / Parents on Parental Involvement 2 (5%) (d) Regular Tracking of Parental Satisfaction and Involvement 9 (21%) (4) No Requirement for Parental Influence (a) Parental Access to Teacher / Staff Qualifications 4 (9%) (b) Regular Performance Report Submission to Parents 18 (42%) (5) No Mention of Parental Influence 1 (2%) LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 313 B. Assignment of Charter School Laws to Legislative Purpose Categories In assigning state charter school laws to purpose categories, this Note finds that these laws overall do not overwhelmingly favor one type of purpose over another. Table Five indicates that a plurality (forty-two percent) of states with charter school laws are libertarian-oriented, more than a quarter are gap-closing, and a third are mixed. This breakdown of laws also does not appear to fall into easily discernible patterns, e.g., based on geography137 or political affiliation.138 137 For example, gap-closing states include Washington in the West, Massachusetts in the Northeast, Tennessee in the South, and Illinois in the Midwest. 138 For example, libertarian-oriented states include Minnesota, which has consistently voted for the Democratic presidential candidate long before it enacted its charter school law in 1991. Minnesota, 270 TO WIN, http://www.270towin.com/states/Minnesota, archived at http://perma.cc/ 7BWX-U5K7 (last visited Dec. 18, 2014); 2012 CER RANKINGS, supra note 16, at 47. They also include Kansas, which has consistently voted for the Republican presidential candidate long before it enacted its charter school law in 1994. Kansas, 270 TO WIN, http://www.270towin.com/states/Kansas, archived at http://perma.cc/SC69-ESGS (last visited Dec. 18, 2014); 2012 CER RANKINGS, supra note 16, at 38. LIU - FINAL 314 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE FIVE: STATE CHARTER SCHOOL LAWS BY PURPOSE CATEGORY (n=43) Gap-Closing Libertarian- Oriented Mixed 11 States (26%) 18 States (42%) 14 States (33%) Connecticut Illinois Louisiana Massachusetts Missouri New York North Carolina Rhode Island Tennessee Washington Wisconsin Alaska Arizona Georgia Idaho Indiana Iowa Kansas Maryland Michigan Minnesota New Hampshire New Jersey Oklahoma Oregon Pennsylvania Texas Utah Wyoming Arkansas California Colorado Delaware District of Columbia Florida Hawaii Maine Mississippi Nevada New Mexico Ohio South Carolina Virginia For many states, categorizing state charter school legislation as gap-closing or libertarian-oriented involves a relatively straightforward analysis of the legislation’s specified purposes, provisions on at-risk student preferences and parental involvement, as well as any relevant academic literature on a law’s passage. Louisiana’s charter school law exemplifies a gap-closing state that focuses primarily on addressing the needs of at-risk students. The law cites neither increasing parental involvement, nor increasing autonomy, as its purposes.139 Instead, not only does the Louisiana charter school law emphasize improving the existing public school system overall, but it also expressly 139 LA. REV. STAT. ANN. § 17:3972 (2011). LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 315 makes “the best interests of at-risk pupils” the “overriding consideration” in implementing the law.140 A provision in the law requiring specific percentages of at-risk students to be enrolled further supports the conclusion that Louisiana is a gap-closing state.141 Similarly, Massachusetts’ charter school law is gap-closing because of its emphasis on closing achievement gaps,142 including the requirement that “not less than 2 of the new commonwealth charters” target low- achieving students every year.143 The categorization of the Illinois charter school law, in contrast, is not as clear-cut. The law cites both increasing parental and community involvement and addressing the needs of at-risk students as its purposes.144 Nonetheless, the Illinois charter school law’s provisions relating to parental involvement are relatively limited,145 while it requires a preference for charter schools “designed to enroll and serve a substantial portion of at-risk children.”146 This mandatory provision favors categorizing Illinois’ charter school law as gap-closing on balance.147 140 Id. 141 ECS STATE PROFILES, supra note 113. 142 James Vaznis, Lawmakers Approve Education Bill: Will Help State’s Bid for U.S. Funds, Patrick Says, BOS. GLOBE, Jan. 15, 2010, at B1; Brian MacQuarrie, Patrick Trumpets Education Legislation: Governor Signs Bill Designed to Close Achievement Gaps, BOS. GLOBE, Jan. 19, 2010, at B1. 143 MASS. GEN. LAWS ch. 71 § 89(i)(2) (2011). 144 ILL. COMP. STAT. tit. 105 § 5/27A-2 (2011). 145 Besides the (common) requirement that an existing, traditional public school seeking to convert to a charter school must demonstrate parental support, the Illinois charter school law merely requires that a charter school application include a description of the nature and extent of parental and other community involvement. Id. § 5/27A-8. There are no provisions mandating, for example, that parents serve on the school’s governing board and play a role in determining revocation, termination, or renewal of a charter. See id. §§ 5/27A-1 to -14. 146 Id. § 5/27A-8(a). See also Ron Zimmer et al., Examining Charter Student Achievement Effects Across Seven States, 31 ECON. EDUC. REV. 213, 214–15 (2011); ECS STATE PROFILES, supra note 113. 147 Traditional public schools’ poor student performance appears to have particularly motivated the growth of charter schools in Illinois, or at LIU - FINAL 316 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 Among libertarian-oriented state charter school laws, Arizona’s law is a leading example of a legislature’s clear intent to deregulate in order to provide parents and local community members with greater control over education. In addition to the general purpose of “improv[ing] pupil achievement,” Arizona’s charter school legislation simply states that it established charter schools to “provide additional academic choices for parents and pupils.”148 The law does not mentions low-achieving or at-risk student populations.149 This reliance on competition to achieve better outcomes reflects the policy environment at the time.150 With Republicans controlling the legislative and executive branches, Arizona passed its charter school legislation in a conservative political environment that had a strong “anti- bureaucratic spirit.”151 Many opponents of school choice accepted charter school legislation as a compromise to avoid the threat of school voucher legislation.152 Another example of a libertarian-oriented charter school law is Georgia’s. While Arizona emphasized competition as a way to provide parents with greater control over education, Georgia focused on decentralizing education to provide greater autonomy.153 least in Chicago. See Robin J. Lake & Lydia Rainey, Chasing the Blues Away: Charter Schools Scale Up in Chicago, PROGRESSIVE POL’Y INST., 8, 12 (2005) (noting that then-U.S. Education Secretary William J. Bennett’s statement that the Chicago Public Schools were the “worst urban school system in America” in 1987 may have provided strong motivation to support charter schools). 148 ARIZ. REV. STAT. ANN. § 15-181(A) (2011). 149 Id. §§ 15-181 to -189. 150 Michael W. Kirst, Politics of Charter Schools: Competing National Advocacy Coalitions Meet Local Politics, 82 PEABODY J. EDUC. 184, 187 (2007). 151 Bryan C. Hassel & Michelle Godard Terrell, The Rugged Frontier: A Decade of Charter Schooling in Arizona, PROGRESSIVE POL’Y INST. 9 (2004). 152 Id. 153 MILLOT, supra note 56, at 13 (noting that Georgia’s state charter law “focuses more on self-determination by stakeholders in the local school”). See also Kirst, supra note 150, at 187 (“In Georgia, . . . policymakers were uninterested in competition and sought LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 317 Georgia’s law does not mention targeting at-risk students in its purposes, but does include five provisions relating to parental involvement, including requiring that parents serve on a charter school’s governing council.154 As such, Georgia is a libertarian-oriented state favoring control by local community stakeholders rather than by the government. While states’ charter school laws like those of Louisiana and Arizona clearly fall under the gap-closing and libertarian-oriented categories, respectively, other states’ laws appear to have more conflicting purposes and provisions. Consequently, they are categorized as mixed states. A primary example of a state that promotes both gap- closing and libertarian-oriented purposes in its charter school legislation is Arkansas. On the one hand, its law prominently features a “special emphasis” for at-risk students in its statement of purposes and mandates preferences for students qualifying for free or reduced-price lunches and for districts in academic distress.155 On the other hand, Arkansas’ law also cites increasing parental choices as a key purpose,156 in addition to including other provisions relating to parental involvement, such as requiring a charter school application to include a plan for “substantial” parental involvement.157 Because of the emphases on addressing the needs of at-risk students and increasing local as opposed to government control, Arkansas’ charter school law is categorized as mixed. Having developed an original framework for interpreting state charter school laws, this Note next analyzes whether it can help explain inconsistency in charter school performance across states. methods to deregulate and decentralize education after a period of increasing centralization to the state.”). 154 GA. CODE ANN. § 20-2-2062(5.1) (2011) (defining “governing council” for a charter school to include parents as members, in addition to teachers, administrators, and others). 155 ARK. CODE ANN. §§ 6-23-102 & -304 (2011). See also ECS STATE PROFILES, supra note 113. 156 ARK. CODE ANN. § 6-23-102(5). 157 Id. § 6-23-202. LIU - FINAL 318 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 V. CORRELATION BETWEEN LEGISLATIVE PURPOSE AND OUTCOMES To determine whether differences in charter school laws’ purposes are correlated with differences in charter school performances seen across states, this Note analyzes the 2013 CREDO study158 data on twenty-six states159 that have charter school laws. Results indicate that charter schools in states that emphasize improving the outcomes of at-risk populations tend to do significantly better than charter schools in states that emphasize having greater parental and local community control in education. Nearly all gap-closing states see charter schools consistently outperforming traditional public schools, while libertarian-oriented and mixed states present a less consistent picture for charter school performance.160 A. Overall Comparison of Charter and Traditional Public School Performance Table Six indicates that gap-closing state charter school laws tend to be much more successful than libertarian- oriented and mixed state laws in producing effective charter schools. Charter schools in nearly all (eighty-eight percent) of the gap-closing states outperform traditional public schools in terms of academic growth in both math and reading. The remaining outperform traditional public schools in only one subject (reading). By comparison, charter schools in fully half of the libertarian-oriented states perform less well than traditional public schools in both reading and math. Mixed states have greater variation in outcomes. Fifty percent have different results for math and reading and 158 2013 CREDO STUDY, supra note 46. The 2013 CREDO study results indicate significant variation in states’ charter school performance. See infra Appendix Table A for the study’s data on state charter school impacts. For additional detail on past studies finding inconsistent charter school performance, see supra notes 36–50 and accompanying text. 159 These twenty-six states educate ninety-five percent of the total charter school population. See supra notes 91–94 and accompanying text. 160 See infra Tables Six to Eight. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 319 nearly forty percent have traditional public schools outperforming charter schools in both math and reading. These results support the conclusion that the underlying purpose driving a state’s charter school law is correlated with the resulting charter school performance. TABLE SIX: PROPORTION OF STATES WHERE CHARTER SCHOOLS OUTPERFORM TRADITIONAL PUBLIC SCHOOLS IN 2008–2011 BY PURPOSE161 (n=26) Type of Outcome Gap- Closing (n=8) Libertarian- Oriented (n=10) Mixed (n=8) Charter Schools Outperform TPS162 in Both Math and Reading 88% 30% 13% TPS Outperform Charter Schools in Both Math and Reading 0% 50% 38% Different Results for Math and Reading163 13% 20% 50% B. Average Magnitude of Charter School Impact As Table Seven depicts, the average impact on academic growth of charter schools relative to traditional public schools is significantly greater in gap-closing states than in libertarian-oriented and mixed states for both math and reading. In reading, gap-closing states on average produce forty-three more days of learning each year compared to traditional public schools, while libertarian-oriented states produce only five more days of learning. In math, the difference is even starker—gap-closing states produce fifty- 161 See infra Appendix Table A for the underlying 2013 CREDO study data on state charter school impacts. 162 Traditional public schools. 163 Charter schools outperform traditional public schools only in one subject (i.e., math or reading). LIU - FINAL 320 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 three more days of learning while libertarian-oriented states produce eleven fewer days. Interestingly, mixed states appear to perform worse than both gap-closing and libertarian-oriented states. It is unclear why mixed states do not perform closer to the level of gap- closing states when they both have an emphasis on closing achievement gaps. Further study is warranted, but one possible explanation could be that the competing motivations and vectors of enforcement in these states may create policy noise, and consequently, the weakest kind of oversight of charter school performance (e.g., a combination of a relative lack of state monitoring and a relative lack of parental accountability). In addition, as Table Eight shows, the range in results within each purpose category is high, due to the relatively small sample size. Nonetheless, these estimates of charter schools’ average impact on academic growth each year relative to traditional public schools reinforce the conclusion that gap-closing state charter school laws yield better charter schools than libertarian-oriented and mixed state laws do. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 321 TABLE SEVEN: AVERAGE CHARTER SCHOOL IMPACT ON 2008– 2011 ACADEMIC GROWTH RELATIVE TO TRADITIONAL PUBLIC SCHOOLS BY PURPOSE AND SUBJECT164 (n=26) Subject Gap-Closing (n=8)165 Libertarian- Oriented (n=10) Mixed (n=8) Reading 43 days 5 days -6 days Math 53 days -11 days -18 days TABLE EIGHT: VARIATION IN AVERAGE CHARTER SCHOOL IMPACT ON 2008–2011 ACADEMIC GROWTH RELATIVE TO TRADITIONAL PUBLIC SCHOOLS BY PURPOSE AND SUBJECT166 (n=26) Measure of Variation Gap-Closing (n=8) Libertarian- Oriented (n=10) Mixed (n=8) Reading Minimum 14 days -29 days -108 days Maximum 86 days 43 days 72 days Standard Deviation 27 days 27 days 47 days Math Minimum -7 days -50 days -137 days Maximum 108 days 58 days 101 days Standard Deviation 35 days 36 days 61 days 164 In terms of additional days of learning each year. See infra Appendix Table A for the underlying 2013 CREDO study data on state charter school impacts. 165 If this analysis were to use the 2013 CREDO study data on the charter school impacts of New York City (instead of the rest of the state of New York), the results for the gap-closing states category would not significantly change. The average impacts would be thirty-nine additional days of learning for reading, and fifty days for math. For an explanation of why the 2013 CREDO study separately analyzed New York City, see supra note 91. 166 In terms of additional days of learning each year. LIU - FINAL 322 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 This analysis of the 2013 CREDO study on charter school performance relative to traditional public schools indicates that this Note’s proposed framework provides an effective means for explaining the inconsistency in charter school performance. Because charter laws of gap-closing states are associated with better outcomes, this Note next analyzes these laws’ typical features to determine what legislation should include to maximize charter school success. VI. KEY FEATURES OF CHARTER SCHOOL LAWS BY LEGISLATIVE PURPOSE In assessing the key differences between state charter school laws according to legislative purpose, this Note focuses on four types of features: (1) charter school sector flexibility; (2) charter school autonomy; (3) accountability; and (4) funding. Compared to libertarian-oriented states, gap-closing states generally provide less sector flexibility, similar school autonomy, greater accountability, and more funding. Mixed states present a less consistent picture; depending on the policy component, they may be more like gap-closing states, be more like libertarian-oriented states, or fall in the middle. A. Charter School Sector Flexibility Gap-closing states tend to restrict the development of the overall charter school sector more than both libertarian- oriented and mixed states do. They are less flexible in provisions ranging from the variety of authorizing paths offered and the total number of charter schools and students permitted, to the establishment of virtual charter schools and the involvement of for-profit EMOs.167 The key area where gap-closing states are generally similar to the other states concerns the types of charter school authorizers available.168 167 See infra Tables Ten to Fifteen. 168 See infra Table Nine. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 323 Given the important gate-keeping function that authorizers serve, this Note first examines the variety of authorizers available to charter school applicants. Table Nine shows that charter school applicants across gap-closing, libertarian-oriented, and mixed states largely have similar authorizer options. There is a diverse range of authorizers, including municipalities and nonprofits. The Illinois charter school legislation, for example, provides a means for school district voters, by majority vote, to mandate that an authorizer approves a charter.169 Nevertheless, across all states with charter school laws, local school boards are the most common authorizers available—more than ninety percent of states have these boards as authorizers. There is no clear indication that gap-closing states have a particular type of authorizer that other states lack and that would improve charter school outcomes. TABLE NINE: TYPES OF AUTHORIZERS BY PURPOSE170 (n=43) Authorizer Type Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Local School Boards 100% 94% 93% State Boards of Education 64% 78% 50% Independent State Board 45% 33% 36% Colleges / Universities 45% 44% 36% Nonprofits 9% 6% 14% Municipalities / Mayors 9% 6% 0% Other Authorizers 27% 6% 21% 169 ECS STATE PROFILES, supra note 113. 170 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; 2012 CER RANKINGS, supra note 16; and ECS STATE PROFILES, supra note 113. LIU - FINAL 324 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 Unlike the relative similarity across states in the types of authorizers available to applicants, Table Ten indicates significant differences between gap-closing state charter school laws and other charter school laws in terms of the number of authorizer options available. More than seventy percent of gap-closing states limit the number of options in some way, while only twenty-eight percent of libertarian- oriented states do the same. This result suggests that imposing some limit on authorizing options reduces the risk that low-quality charter school applicants are nonetheless approved just because, if applicants fail to get approved by one authorizer, they can seek out another, less demanding authorizer.171 Legislatures may potentially encourage higher quality authorizing through reasonably limiting the number of authorizers available. Table Ten also shows that the availability of an appeals process after an authorizer rejects a charter application is fairly similar across all states. Thus, this provision does not explain the differences in performance between gap-closing and other states. 171 This result supports the theory tested in the 2009 CREDO STUDY, supra note 46, at 40 (concluding that applicants are strategic in their choice of authorizer and look for the option that is “easiest” on charters). LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 325 TABLE TEN: NUMBER OF AUTHORIZERS AVAILABLE BY PURPOSE172 (n=43) Authorizer Availability Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Multiple Authorizing Options Available 27% 61% 57% Limited Authorizing Options Available 36% 11% 21% Only One Authorizer Available 36% 28% 21% Appeal Available 55% 61% 57% Table Eleven depicts the stark difference between limits in gap-closing states and those in libertarian-oriented states. Nearly three-quarters of gap-closing states impose caps on charter school growth, while more than sixty percent of libertarian-oriented states do not impose any limits. Mixed states are roughly equally divided between imposing at least some limits, and not imposing any at all. The popularity of caps in gap-closing states appears to support the theory that such limits “generate[ ] pressure to allocate charters to maximize the chance of high performance,” assuming that authorizers can successfully predict the likely future performance of charter school applicants.173 In addition, these limits do not appear to drive effective charter school operators to less restrictive states.174 Table Twelve furthermore shows that the existence of these caps is often not too stringent a limit on growth in practice. Nearly all 172 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; 2012 CER RANKINGS, supra note 16; and ECS STATE PROFILES, supra note 113. 173 2009 CREDO STUDY, supra note 46, at 39. 174 Id. LIU - FINAL 326 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 states that impose limits on charter school sector growth leave at least some room for further growth.175 An “extreme” example is California, which has long had a cap that is far above the current number of charter schools in operation.176 These results suggest that an effective policy for improving charter school outcomes is to impose a cap, but also to ensure that there is some room for expansion. TABLE ELEVEN: EXISTENCE OF LIMITS ON CHARTER SCHOOL SECTOR GROWTH BY PURPOSE177 (n=43) Caps on Number of Charter Schools or Students Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Yes 73% 28% 43% Some 0% 11% 14% None 27% 61% 43% TABLE TWELVE: EXISTENCE OF ADEQUATE ROOM FOR GROWTH DESPITE CAP BY PURPOSE178 (n=23) Room for Charter School Growth Gap-Closing (n=8) Libertarian- Oriented (n=7) Mixed179 (n=8) Adequate Room 25% 29% 25% Some Room 75% 71% 63% No Room 0% 0% 13% 175 2012 NAPCS RANKINGS, supra note 96. 176 In mid-2009, the cap provided room for an additional 400 charter schools to open, or five times the number of charter schools that opened each year. 2009 CREDO STUDY, supra note 46, at 38. 177 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; 2012 CER RANKINGS, supra note 16; and ECS STATE PROFILES, supra note 113. 178 Underlying data was compiled using 2012 NAPCS RANKINGS, supra note 96. 179 Percentages do not total to 100% due to rounding. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 327 Table Thirteen shows that nearly all states permit newly created charter schools (start-ups) and traditional public schools converted into charters (conversions). However, gap- closing states are much less likely to allow for virtual schools and instead restrict charter school formation to traditional brick-and-mortar schools only. Only fifty-five percent of gap- closing states permit virtual schools, compared to eighty- three percent for libertarian-oriented states, and seventy- nine percent for mixed states. This difference may imply that virtual schools tend to perform less well than other kinds of charter schools. TABLE THIRTEEN: TYPES OF CHARTER SCHOOLS PERMITTED BY PURPOSE180 (n=43) Charter School Type Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Start-Up 100% 100% 93% Conversion 100% 94% 86% Virtual 55% 83% 79% Likewise, gap-closing states are more restrictive with for- profit EMOs than other states are, as Table Fourteen shows. No libertarian-oriented state prohibits EMOs, but nearly forty percent of gap-closing states do. Mixed states appear to be slightly less restrictive than gap-closing states, but also appear to be more restrictive than libertarian-oriented states. Furthermore, per Table Fifteen, EMOs are less prevalent in gap-closing states in practice as well. In 2011– 2012, compared to sixty-seven percent for libertarian- oriented states, eighty-two percent of gap-closing states had for-profit EMOs fully managing less than ten percent of their charter schools. The libertarian-oriented state of Michigan, where for-profit EMOs run seventy-nine percent of charter schools, is especially friendly to EMO involvement.181 Not 180 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; 2012 CER RANKINGS, supra note 16; and ECS STATE PROFILES, supra note 113. 181 Miron & Gulosino, supra note 27, at 18. LIU - FINAL 328 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 only does Michigan provide relatively limited funding for charter schools, making EMOs valuable for securing necessary resources, but Michigan also prohibits EMO- employed teachers from paying into its expensive statewide retirement system, making it cheaper for EMOs to operate charter schools.182 The dominance of EMOs in the state has attracted significant controversy in light of the ongoing debate over whether for-profit EMOs negatively impact student outcomes.183 The availability of for-profit EMOs to operate charter schools is a key difference between gap-closing and libertarian-oriented states. On the other hand, as Table Fourteen shows, all states are generally open to nonprofit CMOs. This lack of difference in the availability of CMOs across all types of states implies that it does not help explain the inconsistency in outcomes. Overall, gap-closing states tend to offer less flexibility for charter school sector growth than do libertarian-oriented and mixed states. 182 Mead & Rotherham, supra note 9, at 9–10. 183 Many studies have sought to determine the impact of for-profit EMOs on student outcomes. See, e.g., Kerry A. King, Charter Schools in Arizona: Does Being a For-Profit Institution Make a Difference?, 41 J. ECON. ISSUES 729, 730 (2007); Ron French, Pursuit of Money, Learning Mix, BRIDGE MAGAZINE (Feb. 28, 2013), http://bridgemi.com/2013/02/ pursuit-of-money-learning-mix/, archived at http://perma.cc/PQ5M-9D2W. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 329 TABLE FOURTEEN: TYPES OF CHARTER SCHOOL OPERATORS PERMITTED BY PURPOSE184 (n=43) Degree of Restriction on Involvement Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed185 (n=14) For-Profit Education Management Organization (“EMO”) No Limits 64% 100% 71% Some Limits 0% 0% 7% Prohibited 36% 0% 21% Nonprofit Charter Management Organization (“CMO”) No Limits 91% 100% 93% Some Limits 9% 0% 7% Prohibited 0% 0% 0% TABLE FIFTEEN: LEVEL OF FOR-PROFIT EMO MANAGEMENT OF CHARTER SCHOOLS IN 2011–2012 BY PURPOSE186 (n=43) Proportion of EMO-Managed Charter Schools in Each State Gap-Closing (n=11) Libertarian- Oriented187 (n=18) Mixed (n=14) <10% 82% 67% 79% 10–20% 9% 22% 7% >20% 9% 12% 14% B. Charter School Autonomy This Note finds that, for district collective bargaining agreements, libertarian-oriented states provide greater freedom from regulation, while for retirement system and teacher certification requirements, gap-closing states are 184 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; and 2012 CER RANKINGS, supra note 16. 185 Percentages do not total to 100% due to rounding. 186 See Miron & Gulosino, supra note 27, at 20. 187 Percentages do not total to 100% due to rounding. LIU - FINAL 330 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 more likely to provide some freedom from regulation.188 There does not appear to be major differences across states based on legislative purpose in the availability of blanket exemptions from state and local regulations.189 Table Sixteen shows that all states with charter school laws generally provide paths for charter schools to obtain exemptions from at least some of the state and local regulations that typically apply to traditional public schools. Roughly half of the states within each purpose category provide automatic exemptions from all laws,190 while the remainder offer automatic exemptions from some of the regulations and/or opportunities for charter schools to apply for a waiver on a case-by-case basis. TABLE SIXTEEN: INDEPENDENCE FROM STATE AND LOCAL REGULATIONS BY PURPOSE191 (n=43) Degree of Freedom from Regulation Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Automatic Exemption from All Laws192 55% 44% 57% Partial or Case-by- Case Exemption from Laws 45% 56% 43% No Exemption from Laws 0% 0% 0% 188 See infra Tables Seventeen to Nineteen. 189 See infra Table Sixteen. 190 Exemptions from all laws, except those covering health, safety, civil rights, student accountability, employee criminal history checks, open meetings, freedom of information, and generally accepted accounting principles. 191 Underlying data was compiled using state charter school statutes and 2012 NAPCS RANKINGS, supra note 96. 192 There are still some types of laws to which charter schools are subject. See supra note 190. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 331 Table Seventeen illustrates that a difference between gap-closing and libertarian-oriented states relates to the degree to which charter school teachers are subject to a local school district’s collective bargaining agreements (“CBAs”). Traditional public school teachers generally are subject to these agreements. As expected, a significant majority (sixty- seven percent) of libertarian-oriented states automatically exempt charter school teachers from CBAs, while gap-closing states are less likely to do the same. Less than half of gap- closing states provide similarly expansive exemptions. These states are more likely to provide exemptions for only certain groups of teachers or permit charter schools to apply for a waiver on a case-by-case basis. For example, the gap-closing Massachusetts charter school legislation permits their commonwealth (i.e., start-up) charter school teachers to negotiate as a separate bargaining unit or work independently, but requires Horace Mann (i.e., conversion) charter teachers to remain covered by the district CBA.193 Mixed states are more evenly divided between providing automatic exemptions for everyone, and partial or case-by- case exemptions. TABLE SEVENTEEN: FREEDOM FROM COLLECTIVE BARGAINING AGREEMENTS BY PURPOSE194 (n=43) Degree of Freedom from CBAs Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) No Exemption 0% 11% 14% Partial / Case-by- Case Exemption 64% 22% 43% Automatic Exemption 36% 67% 43% 193 2012 CER RANKINGS, supra note 16. 194 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; 2012 CER RANKINGS, supra note 16; and ECS STATE PROFILES, supra note 113. LIU - FINAL 332 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 In contrast to how libertarian-oriented states tend to provide charter schools with greater freedom from CBA requirements, gap-closing states appear slightly more likely to at least partially exempt charter schools from teacher certification requirements, as Table Eighteen shows. Mixed states are relatively similar to gap-closing states here. More than seventy percent of both gap-closing and mixed states either exempt some charter school teachers from these requirements, or provide alternatives for meeting licensing requirements. The gap-closing state of Connecticut, for example, requires a minimum of fifty-percent of a charter school’s teachers to be certified, but also issues charter school educator permits as an alternative.195 In addition, a majority of libertarian-oriented states provide some freedom from teacher certification requirements, but a third provide no exemption at all. The libertarian-oriented state of Texas only requires English Language Learner and special education teachers to be certified, while the libertarian- oriented state of Kansas does not provide any exemptions.196 195 CONN. GEN. STAT. § 10-66dd (2011). These permits are issued provided that teachers: (1) pass the state reading, writing, and math competency test for teacher certification candidates or meet state board criteria for a testing waiver; (2) pass the same state test as a teacher or administrator certification candidate seeking to work in the same subject or administrative area; and (3) demonstrate effectiveness as a teacher or school administrator, as appropriate. Id. 196 Measuring Up: Alignment to the Model Law—Texas, NAT’L ALLIANCE FOR PUB. CHARTER SCH. (Jan. 2014), http://www.public charters.org/get-the-facts/law-database/states/tx/. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 333 TABLE EIGHTEEN: FREEDOM FROM TEACHER CERTIFICATION REQUIREMENTS BY PURPOSE197 (n=43) Degree of Freedom from Requirements Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed198 (n=14) No Exemption 18% 33% 21% Partial / Case-by- Case Exemption 73% 56% 71% Automatic Exemption 9% 11% 7% As Table Nineteen shows, gap-closing states also are more likely than libertarian-oriented states to exempt charter school teachers from participation in the statewide retirement system. Given the underlying purpose of libertarian-oriented states, it may be surprising that a majority (sixty-one percent) of libertarian-oriented states provide no exemption from participation in the statewide retirement system, while less than a third of gap-closing states do the same. Mixed states tend to have more diverse policies, with one state even prohibiting charter school teachers from participating in the statewide retirement system at all. 197 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; 2012 CER RANKINGS, supra note 16; and ECS STATE PROFILES, supra note 113. 198 Percentages do not total to 100% due to rounding. LIU - FINAL 334 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE NINETEEN: PARTICIPATION IN STATEWIDE RETIREMENT SYSTEM BY PURPOSE199 (n=43) Type of Participation in Retirement System200 Gap- Closing201 (n=11) Libertarian- Oriented (n=18) Mixed (n=14) No Exemption 27% 61% 50% Partial Exemption 45% 6% 14% Automatic Exemption / Optional Participation 27% 33% 29% Prohibition on Participation 0% 0% 7% In sum, state charter school laws are relatively similar in generally exempting charter schools from at least some state and local regulations that affect charter school autonomy. It is interesting, though, that while libertarian-oriented states are more likely than gap-closing to exempt charter schools from CBA requirements, they are somewhat less likely to exempt charter schools from teacher certification and statewide retirement system requirements. C. Accountability In analyzing differences in accountability provisions, this Note looks at accountability requirements for both authorizers and charter schools. It finds that gap-closing states tend to impose greater requirements in this area than 199 Underlying data was compiled using: state charter school statutes; 2012 NAPCS RANKINGS, supra note 96; 2012 CER RANKINGS, supra note 16; and ECS STATE PROFILES, supra note 113. 200 There are a few exceptions to these generally applicable categories, including Florida, Wisconsin, and the District of Columbia. For example, D.C. permits employees transferring from a local district school to a charter school to stay in the D.C. retirement system if desired, but prohibits all other employees from participating in the retirement system. 2012 NAPCS RANKINGS, supra note 96, at 23. 201 Percentages do not total to 100% due to rounding. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 335 libertarian-oriented states. Notable differences include that gap-closing states are more likely to subject authorizers to oversight to ensure high-quality authorizing202 as well as to empower authorizers to hold charter schools accountable for student outcomes.203 For example, nearly three-quarters of gap-closing states provide authorizers with the authority to take corrective action short of revocation of charters if necessary, while less than half of libertarian-oriented states do the same.204 Libertarian-oriented states, however, are more likely than gap-closing states to require publishing annual school performance reports.205 An area where gap- closing and libertarian states have relatively similar policies concern requiring financial accountability.206 Finally, mixed states once again present a less consistent picture in terms of how stringent their accountability provisions are. Table Twenty compares states on authorizer accountability. Gap-closing charter school laws are more likely than libertarian-oriented laws to include at least some provisions to hold authorizers accountable for their performance. Mixed states seem to fall somewhere in the middle. Ninety percent of gap-closing states require that an authorizer oversight body conduct at least some regular review of authorizers, compared to seventy-nine percent for mixed states and sixty-one percent for libertarian-oriented states. Similar proportions within each purpose category require the establishment of an oversight body that has the authority to sanction authorizers for performing poorly. Thus, promoting high-quality authorizing behavior is a key distinguishing feature of gap-closing states. 202 See infra Table Twenty. 203 See infra Table Twenty-Two. 204 Id. 205 Id. 206 Id. LIU - FINAL 336 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE TWENTY: DEGREE OF AUTHORIZER ACCOUNTABILITY BY PURPOSE207 (n=43) Type of Accountability Gap- Closing208 (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Regular Review Process by Authorizer Oversight Body Yes 45% 11% 36% Some 45% 50% 43% None 9% 39% 21% Authorizer Oversight Body with Authority to Sanction Authorizers Yes 45% 22% 36% Some 45% 39% 43% None 9% 39% 21% Table Twenty-One highlights another key difference between gap-closing states and libertarian-oriented states stemming from the degree to which the overall charter school system performance is monitored. Gap-closing states are significantly more likely to require at least some periodic evaluation of the overall charter program and its outcomes. While more than eighty percent of gap-closing states require such evaluations, a majority of libertarian-oriented states do not. Also, the proportions of mixed states that provide these evaluations are fairly similar to gap-closing states here. 207 Underlying data was compiled using state charter school statutes and 2012 NAPCS RANKINGS, supra note 96. 208 Percentages do not total to 100% due to rounding. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 337 TABLE TWENTY-ONE: OVERALL CHARTER SCHOOL SYSTEM ACCOUNTABILITY BY PURPOSE209 (n=43) Type of Accountability Gap- Closing (n=11) Libertarian- Oriented (n=18) Mixed210 (n=14) Periodic Evaluation of Overall Charter Program / Outcomes 64% 39% 71% Some Evaluation of Overall Charter Program / Outcomes 18% 0% 14% No Periodic Evaluation of Overall Charter Program / Outcomes 18% 61% 14% Table Twenty-Two shows that while nearly all states, regardless of purpose category, require holding charter schools accountable financially, there are significant differences with respect to other accountability provisions. Gap-closing states are more likely to require annual collection and analysis of student outcome data, authorizer authority for oversight activities, notification to schools of problems and provision of opportunity to remedy them, and finally, authorizer authority to take corrective actions (e.g., sanctions) short of revocation. Libertarian-oriented states, on the other hand, are slightly more likely than gap-closing states to require at least some publication of annual school performance reports. 209 Underlying data was compiled using state charter school statutes and 2012 NAPCS RANKINGS, supra note 96. 210 Percentages do not total to 100% due to rounding. LIU - FINAL 338 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE TWENTY-TWO: DEGREE OF CHARTER SCHOOL ACCOUNTABILITY BY PURPOSE211 (n=43) Type of Accountability Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Financial Accountability212 Yes 100% 83% 93% Some 0% 6% 7% None 0% 11% 0% Annual Collection and Analysis of Student Outcome Data Yes 45% 22% 36% Some 27% 50% 50% None 27% 28% 14% Publication of Annual School Performance Reports Yes 36% 17% 36% Some 36% 61% 50% None 27% 22% 14% Authorizer Authority for Oversight Activities Yes 82% 61% 86% Some 0% 6% 14% None 18% 22% 0% Notification to Schools of Problems and Opportunity to Remedy Problems Yes 73% 33% 79% Some 9% 17% 7% None 18% 50% 14% Authorizer Authority to Take Corrective Actions Besides Revocation Yes 73% 33% 79% Some 9% 22% 7% None 18% 44% 14% 211 Underlying data was compiled using state charter school statutes and 2012 NAPCS RANKINGS, supra note 96. 212 Includes complying with Generally Accepted Accounting Principles and submitting an independent annual audit of charter schools to the authorizer. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 339 Overall, with the exception of financial accountability and publication of annual performance reports, gap-closing state charter school laws tend to impose greater accountability requirements on charter schools than libertarian-oriented laws. These results appear to align well with the underlying purpose of libertarian-oriented state charter laws, which generally rely on greater, informed parental choice and inter- school competition, rather than increased government regulation, as the key mechanism for holding schools accountable. D. Funding This Note finds that most state charter school laws do not provide charter schools with the operational funding, capital funding, and facilities access that are comparable to the resources provided to traditional public schools. This conclusion generally applies whether the law is gap-closing, libertarian-oriented, or mixed, although there are some notable differences. Table Twenty-Three shows that gap-closing states are more likely than libertarian-oriented states to provide charter schools operational funding and access to categorical federal and state funding. However, both of these categories of states are relatively similar to one another in terms of the equity of the student transportation funding they provide. Table Twenty-Four illustrates that there are no major differences between gap-closing and libertarian-oriented states in terms of equity in capital funding and facilities access. Most states’ charter school laws, regardless of their underlying purposes, do not ensure such equity (e.g., no grants, no loans, lack of right of first refusal to unused public school property). LIU - FINAL 340 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE TWENTY-THREE: EQUITY IN OPERATIONAL FUNDING BY PURPOSE213 (n=43) Type of Funding Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Equitable Operational Funding Yes 9% 0% 7% Some 82% 44% 64% None 9% 56% 29% Equal Access to Categorical Federal and State Funding Yes 27% 22% 14% Some 64% 33% 64% None 9% 44% 21% Equitable Transportation Funding214 Yes 36% 22% 57% Some 18% 17% 7% None 45% 56% 36% TABLE TWENTY-FOUR: EQUITY IN CAPITAL FUNDING AND FACILITIES BY PURPOSE215 (n=43) Type of Funding Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Per-Pupil Facilities Allowance that Reflects Actual Costs Yes 9% 0% 7% Some 18% 39% 36% None 73% 61% 57% 213 Underlying data was compiled using state charter school statutes and 2012 NAPCS RANKINGS, supra note 96. 214 This criterion is not applicable for Michigan because the state does not provide any transportation funding or other support to traditional public schools, either. Measuring Up: Alignment to the Model Law—Texas, NAT’L ALLIANCE FOR PUB. CHARTER SCH. (Jan. 2014), http://www.public charters.org/get-the-facts/law-database/states/mi/. 215 Underlying data was compiled using state charter school statutes and 2012 NAPCS RANKINGS, supra note 96. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 341 TABLE TWENTY-FOUR: EQUITY IN CAPITAL FUNDING AND FACILITIES BY PURPOSE216 (CONT.) (n=43) Type of Funding Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) State Grant Program for Charter School Facilities Yes 9% 11% 29% Some 0% 0% 0% None 91% 89% 71% State Loan Program for Charter School Facilities Yes 9% 11% 21% Some 0% 0% 0% None 91% 89% 79% Equal Access to Tax-Exempt Bonding Authorities Yes 36% 61% 64% Some 36% 22% 7% None 27% 17% 29% Credit Enhancement for Charter School Facilities Yes 0% 11% 21% Some 0% 6% 0% None 100% 83% 79% Equal Access to Facilities Programs Available to Traditional Public Schools Yes 27% 6% 36% Some 0% 11% 7% None 73% 83% 57% Right of First Refusal to Access Unused Public School Property Yes 9% 11% 29% Some 9% 6% 29% None 82% 83% 43% 216 Underlying data was compiled using state charter school statutes and 2012 NAPCS RANKINGS, supra note 96. LIU - FINAL 342 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 E. Overall Policy Implications Categorizing state charter school laws by their underlying legislative purposes helps explain the inconsistency in charter school performance across states. Because state laws assigning priority to closing achievement gaps are correlated with better outcomes, this Note proposes that legislatures adopt the key distinguishing features of gap-closing states’ charter school legislation. Overall, gap-closing states are generally more likely to regulate the development of the charter school sector, impose greater accountability requirements on authorizers and charter schools, and provide greater equity in operational funding to charter schools. Therefore, while state charter school legislation should still ensure a minimum baseline of charter school flexibility and autonomy, it should also include certain reasonable restrictions to ensure high-quality outcomes. Table Twenty-Five summarizes key provisions to include in charter school legislation. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 343 TABLE TWENTY-FIVE: PROFILE OF RECOMMENDED CHARTER SCHOOL LAW FEATURES Category Charter School Law Features Charter School Sector Flexibility  Multiple Authorizers: Ensure that there is an adequate number of authorizing options available, but there may be a need to place some limits to ensure that only high-quality charter schools are approved  Caps: Place some limits on the growth of the charter school market, but still provide reasonable room for future expansion  Types of Schools: Permit both start-ups and conversions, but limit the establishment of virtual schools until they prove quality of outcomes  Education Service Providers: Restrict the involvement of for-profit education management organizations, and potentially encourage the use of nonprofit charter management organizations Charter School Autonomy  Independence from State and Local Regulations: Provide automatic exemptions from most state and local regulations, or at the minimum, provide for partial or case-by-case exemptions  Collective Bargaining Agreements, Teacher Certification, and Statewide Retirement System Requirements: Provide at least some freedom from these requirements217 217 The differences between gap-closing and libertarian-oriented states appear to be less significant here. See supra Part VI.B. LIU - FINAL 344 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE TWENTY-FIVE: PROFILE OF RECOMMENDED CHARTER SCHOOL LAW FEATURES (CONT.) Category Charter School Law Features Accounta- bility  Authorizer Accountability: Ensure authorizers are held accountable by subjecting them to a regular review process and an oversight body that has authority to sanction them  Overall Charter School System Accountability: Provide for periodic evaluations of charter school programs and outcomes  Charter School Accountability: Require accountability for financial and student performance outcomes, especially by creating charter school oversight bodies, notifying schools of problems, providing schools opportunities to remedy problems, and giving authorizers authority to take corrective actions short of revocation Funding  Operational Funding: Provide charter schools with funding and access to categorical federal and state grants that are relatively comparable to what are provided to traditional public schools  Capital Funding and Facilities Access: States are generally similar to one another in that they do not provide equitable funding and access and thus it is unclear whether providing such equity is needed to improve student achievement outcomes LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 345 These findings suggest that charter schools work best when they are relatively less marketized and choice-driven and more guided by public policy. Moreover, the fact that these types of charter schools, such as those run by the nonprofit organization KIPP, are also the schools that are most strongly supported by the business community218 is an indication that charter schools are not merely a “conspiracy” by hedge fund managers and other “corporate” interests to destroy public schools and replace them with marketized private choice.219 To the contrary, the relatively more marketized and choice-driven for-profit charter schools— which are more common in libertarian-oriented states220— are not a focus of corporate philanthropy, suggesting that the business community may be trying to improve the “public” school system instead. Finally, while this Note does not recommend adoption of typical features of libertarian-oriented state charter school laws, it recognizes the potential importance and value of liberty. By providing parents greater control over their children’s schooling, parents can better preserve and promote their religious commitments and other personal interests. It may be valid and reasonable for states to pursue this approach if they desire. However, at least in the debate over whether charter schools can and generally do outperform traditional public schools, the performance of these libertarian-oriented charter school states—which is less consistent than in gap-closing states221—should be excluded. This study’s findings suggest that charter schools can be a valuable tool for improving public education under gap-closing states’ charter school policies. 218 See supra notes 30–33 and accompanying text. 219 See supra notes 34–35 and accompanying text. 220 See supra Tables Fourteen to Fifteen and accompanying text. 221 See supra Part V. LIU - FINAL 346 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 VII. CONCLUSION The debate over the value of charter schools has continued to be fierce,222 but the conversation thus far has not fully recognized the important role that underlying differences in the purposes of each state’s charter school legislation play in determining the resulting charter school performance outcomes. This Note concludes that state charter school laws can be effectively categorized according to whether they emphasize closing achievement gaps, greater parental and local community control, or both. Moreover, while further research should be conducted to build on this study’s suggestive analysis, this original framework of gap-closing, libertarian-oriented, and mixed states helps explain the inconsistency found in charter school performance across states. It also helps determine which charter school law features may be most effective. Because gap-closing states, which promote improving the outcomes of at-risk student populations in particular, are correlated with better charter school performance outcomes, this Note argues that legislatures should adopt key distinguishing features of gap-closing states in order to improve outcomes. Contrary to the influential recommendations of the Center for Education Reform, which has long stressed the importance of free-market mechanisms and opposed governmentally imposed regulations,223 greater restrictions on the formation and operation of charter schools may be crucial to improving charter school performance. In conclusion, with the proper analytical framework to better understand how differences in state charter school laws’ purposes and features may drive variation in outcomes, states can more effectively improve charter school policy and performance. 222 See supra notes 36–39 and accompanying text. 223 See supra note 64 and accompanying text. LIU - FINAL No. 1:273] SOLVING THE PUZZLE OF CHARTER SCHOOLS 347 APPENDIX TABLE A: 2013 CREDO STUDY ON CHARTER SCHOOL IMPACTS IN 2008–2011224 State Reading (Days of Learning) Math (Days of Learning) Arizona -22 -29 Arkansas -22 -22 California 22 -7 Colorado 7 -7 District of Columbia 72 101 Florida -7 0 Georgia 14 -14 Illinois 14 22 Indiana 36 14 Louisiana 50 65 Massachusetts 36 65 Michigan 43 43 Minnesota 14 -7 Missouri 14 22 Nevada -108 -137 New Jersey 43 58 New Mexico 0 -29 New York 36 79 New York City 0 94 North Carolina 22 -7 Ohio -14 -43 Oregon -22 -50 Pennsylvania -29 -50 Rhode Island 86 108 Tennessee 86 73 Texas -22 -29 Utah -7 43 224 2013 CREDO STUDY, supra note 46, at 52–53. LIU - FINAL 348 COLUMBIA BUSINESS LAW REVIEW [Vol. 2015 TABLE B: DURATION OF INITIAL CHARTER TERM BY PURPOSE225 (n=43) Length of Initial Charter Term226 Gap-Closing (n=11) Libertarian- Oriented (n=18) Mixed (n=14) Specified Years227 45% 33% 64% <5 Years 0% 33% 22% 5-10 Years 80% 50% 67% ≥10 Years 20% 33% 22% Maximum Years228 55% 50% 29% <5 Years 0% 0% 0% 5-10 Years 67% 56% 75% ≥10 Years 33% 44% 25% Minimum Years229 18% 22% 7% <5 Years 50% 75% 100% 5-10 Years 50% 25% 0% ≥10 Years 0% 0% 0% None Specified 0% 6% 7% 225 Underlying data was compiled using state charter school statutes and 2012 NAPCS RANKINGS, supra note 96. 226 Percentages do not necessarily total to 100%, because some state charter school laws may include more than one requirement for the duration of the initial charter term. For example, a state may specify both the minimum and the maximum years of a charter term and/or specify the duration of the charter term only for certain kinds of charter schools. 227 The state charter school legislation specified, for at least some types of charter schools, the precise number of years that the initial charter term must be. 228 The state charter school legislation specified, for at least some types of charter schools, the maximum number of years that the initial charter term may be. 229 The state charter school legislation specified, for at least some types of charter schools, the minimum number of years that the initial charter term may be.