Emory Law Journal

Students, Security, and Race
Jason P. Nance Assistant Professor of Law, University of Florida Levin College of Law. I would like to thank the following individuals for the helpful feedback and assistance they provided for this Article: Nancy Dowd, Paul Gugliuzza, Catherine Hancock, Lyrissa Lidsky, Tom C. W. Lin, Richard Peterson, Sharon Rush, and Michael Wolf. I also appreciate the helpful comments provided by faculty members during presentations of earlier drafts of this article at the 2012 Southeast Law Schools Jr./Sr. Faculty Workshop at Tulane Law School, the Southeastern Association of Law Schools 2012 Annual Conference, and 2013 Junior Scholars Conference at the University of Florida Levin College of Law’s Criminal Justice Center. I received outstanding research assistance from Olga Balderas and Dane Ullian. I am grateful for the summer research grant provided by the University of Florida Levin College of Law. I would also like to thank the U.S. Department of Education for providing access to the restricted-use data for the 2009–2010 School Survey on Crime and Safety. Some of the material in this article was presented as an excerpt in the Stanford Law Review Online. See Jason P. Nance, 65 Stan. L. Rev. Online 103 (2013).

Abstract

In the wake of the terrible shootings in Newtown, Connecticut, our nation has turned its attention to school security. For example, several states have passed or are considering passing legislation that will provide new funding to schools for security equipment and law enforcement officers. Strict security measures in schools are certainly not new. In response to prior acts of school violence, many public schools for years have relied on metal detectors, random sweeps, locked gates, surveillance cameras, and law enforcement officers to promote school safety. Before policymakers and school officials invest more money in strict security measures, this Article provides additional factors that should be considered. First, drawing on recent, restricted data from the U.S. Department of Education, this Article presents an original empirical analysis revealing that low-income students and minority students are much more likely to experience intense security conditions in their schools than other students, even when taking into account neighborhood crime, school crime, and school disorder. These findings raise concerns that such inequalities may continue or worsen as policymakers provide additional funding for security measures. Second, this Article argues that strict security measures do not support long-term solutions needed to effectively prevent school violence. Indeed, strict security measures may exacerbate the underlying problems by creating barriers of adversity and mistrust between students and educators.

In addition, this Article offers recommendations to address the disproportionate use of security measures on low-income and minority students and to curb violence more effectively. It urges school officials and policymakers to support programs that build trust and collective responsibility instead of providing grants for strict security measures. Further, it recommends that the Department of Education’s Office of Civil Rights play a more active role in addressing the disproportionate use of strict security measures on minority students.

Introduction

Perhaps never before has our nation been more focused on school security. The horrific massacre of twenty children and six educators at Sandy Hook Elementary School in Newtown, Connecticut, on December 14, 2012, has provoked intense feelings of sadness, anger, and perplexity. 1See, e.g., Tom Raum & Jennifer Agiesta, Poll: Americans Angrier About Sandy Hook than 9/11 Attacks, CNSNews (Jan. 16, 2013), http://cnsnews.com/news/article/poll-americans-angrier-about-sandy-hook-shooting-911-attacks (reporting the anger Americans feel over the shootings); Jeanette Rundquist, Surprise Security Drills Coming to N.J. Schools, NJ.com (Jan. 16, 2013, 9:07 PM), http://www.nj.com/news/index.ssf/2013/01/surprise_school_security_drills.html (“The tragedy in Connecticut was the school community’s 9/11. This has touched the very soul of the country, no less the school community.” (internal quotation marks omitted)). Naturally, parents, educators, policymakers, and communities are searching for immediate solutions to ensure students’ safety. 2See, e.g., Stephanie Banchero & Caroline Porter, In Pursuit of Safety, Schools’ Paths Diverge: Districts Face Tight Budgets, Worries over Armed Security on Campus as They Respond to the Massacre at Sandy Hook, Wall St. J. (Jan. 17, 2013, 7:49 PM) http://online.wsj.com/article/SB10001424127887323783704578248113754373942.html (“Schools are running security drills, planning to install security cameras and bullet-proof glass, and hiring safety consultants.”); Lynh Bui, Montgomery County Parents Ask for More School Security, Teacher Training During Budget Hearing, Wash. Post (Jan. 11, 2013, 3:15 PM) http://www.washingtonpost.com/blogs/maryland-schools-insider/post/montgomery-county-parents-ask-for-more-school-security-teacher-training-during-budget-hearing/2013/01/11/e8d3dcf4-5aab-11e2-9fa9-5fbdc9530eb9_blog.html (explaining the increased concerns parents have for their children’s safety and their demand for increased on-campus security); Motoko Rich, School Officials Look Again at Security Measures Once Dismissed, N.Y. Times (Dec. 18, 2012) http://www.nytimes.com/2012/12/19/education/after-newtown-shootings-schools-consider-armed-security-officers.html?pagewanted=1&_r=1 (describing the reactions of policymakers to the Newtown shootings). In response to this tragedy, many state legislatures and local school boards are considering whether to allocate additional funding to schools for purchasing security equipment and hiring law enforcement officers. 3See, e.g., H.B. 612, 2013 Leg., Reg. Sess. (Ala. 2013) (proposing funding to school districts for “surveillance cameras, metal detectors, and other safety equipment”); S.B. 2230, 2013 Leg., Reg. Sess. (N.Y. 2013) (funding security aid to provide schools with metal detectors, electronically operated partitions, and security cameras); S.B. 2267, 63d Leg. Assemb. (N.D. 2013) (proposing funding to equip schools with alarms, cameras, electronic door locks, emergency response call buttons, intercom systems, key or pass card systems, and metal detectors); H. 2343, Public School Security Equipment Grant Act of 2013, ch. 608 (Va. 2013) (providing funding to schools to for purchasing security equipment); see also Doug Finke, State Board of Education Calls for $874M Increase in School Funding, PJSTAR (Jan. 24, 2013, 9:02 PM), http://www.pjstar.com/news/x1503807290/State-Board-of-Education-calls-for-874M-increase-in-school-funding (“The Illinois State Board of Education called for an $874 million increase in state spending on elementary and secondary education on Thursday, including $20 million that would be available for school security measures.”); Mary Wilson, Pa. Senate Leader Suggests Steep Increase in School Security Funding, Newsworks (Jan 22, 2013), http://www.newsworks.org/index.php/local/item/49955-pa-senate-leader-suggests-steep-increase-in-school-security-funding?linktype=hp_topstorylist (“The leader of Pennsylvania’s Senate wants to increase grant funding twentyfold for school security, including armed guards.”); Liz Hayes, Burrell Board Focuses on School Safety, TribLive (May 7, 2013, 12:51 AM), http://triblive.com/neighborhoods/yourallekiskivalley/yourallekiskivalleymore/3970294-74/burrell-security-wagner#axzz2THQXVCDR (discussing a Pennsylvania School District that is considering using metal detectors more regularly, installing X-ray equipment to screen student backpacks, and making windows “more bulletproof”); Tom Barton, Board Members Unsure Metal Detectors Needed at Beaufort County Schools, Island Packet.com (May 8, 2013), http://www.islandpacket.com/2013/05/08/2495164/board-members-unsure-metal-detectors.html (discussing a South Carolina School Board’s consideration of adding metal detectors in schools).

Before policymakers and school officials make these substantial financial commitments, there is much to consider. Just two days prior to the shootings, on December 12, 2012, another event took place that has since been overshadowed by the Newtown tragedy, but is related to the current response. On that day, a U.S. congressional hearing was held to discuss, for the first time, ending the so-called school-to-prison pipeline. 4See, Ending the School-to-Prison Pipeline Before the Subcomm. on the Constitution, Civil Rights, and Human Rights, 112th Cong. (2012), available at http://www.judiciary.senate.gov/hearings/hearing.cfm?id=b61e5f08eadf22b2ec4ab964fc64ae9f; Susan Ferriss, ‘School to Prison Pipeline’ Hit on Capitol Hill, Ctr. for Pub. Integrity (Dec. 13, 2012, 6:00 AM), http://www.publicintegrity.org/2012/12/13/11921/school-prison-pipeline-hit-capitol-hill. The school-to-prison pipeline refers to the practice of funneling students directly to the juvenile correction system from schools, or suspending or expelling students from schools, thereby creating conditions where those students are more likely to be arrested. 5 See, e.g., Advancement Project et al., Federal Policy, ESEA Reauthorization, and the School-to-Prison Pipeline 2 (2011), available at http://b.3cdn.net/advancement/ceb35d4874b0ffde10_ubm6baeap.pdf; NAACP Legal Def. & Educ. Fund, Inc., School to Prison Pipeline, NAACPLDF, http://www.naacpldf.org/case/school-prison-pipeline (last visited Aug. 12, 2013). This disturbing trend disproportionately affects minority students, especially African-American boys, depriving many of these students of the benefits of an education, future employment, and participation in our democracy. 6Id.

While violence and school safety are serious issues that must be addressed, the congressional hearing held just two days before the Newtown shootings highlights another serious problem that our nation faces: the disparate treatment of minority students in public schools. To further illustrate, earlier in 2012, the U.S. Department of Education’s Office of Civil Rights (OCR) released data from a national survey of over 72,000 schools around the country serving approximately eighty-five percent of the nation’s students. 7See Press Release, U.S. Dep’t of Educ., New Data from U.S. Department of Education Highlights Educational Inequalities Around Teacher Experience, Discipline, and High School Rigor (Mar. 6, 2012), available at http://www.ed.gov/news/press-releases/new-data-us-department-education-highlights-educational-inequities-around-teache. The results were troubling, but not surprising. The data revealed that minority students across the country are disciplined more often and more severely than white students, have less access to higher-level courses, and more often have teachers that are less experienced and are lower paid. 8Id. In response to these findings, U.S. Secretary of Education Arne Duncan stated that the “undeniable truth is that the everyday educational experience for many students of color violates the principle of equity at the heart of the American promise.” 9Id. Indeed, the disparate treatment of minorities in schools, especially of black males, is well documented in the literature. 10See e.g., Catherine Y. Kim et al., The School-to-Prison Pipeline: Structuring Legal Reform 53–54 (2010) (describing the overrepresentation of children of color in restrictive special education programs); Angela A. Ciolfi & James E. Ryan, Race and Response-to-Intervention in Special Education, 54 How. L.J. 303, 326–27 (2011) (explaining that African-American students are overrepresented in restrictive educational settings such as alternative schools and other disciplinary programs); Nancy E. Dowd, What Men? The Essentialist Error of the “End of Men, 93 B.U. L. Rev. (forthcoming) (describing the physical marginalization, the pattern of disproportionate discipline, and the overidentification and misidentification of learning disabilities of black males in school); Theresa Glennon, Knocking Against the Rocks: Evaluating Institutional Practices and the African American Boy, 5 J. Health Care L. & Pol’y 10, 11 (2002) (explaining that black males are more likely to be identified as disabled and are more likely to be placed in educational programs that exert greater control and provide fewer services than other children); Theresa Glennon, Looking for Air: Excavating Destructive Educational and Racial Policies to Build Successful School Communities, in Justice for Kids: Keeping Kids Out of the Juvenile Justice System 107, 110–11 (Nancy E. Dowd ed., 2011) (citing studies demonstrating that minority students are disproportionately disciplined); Russell J. Skiba, Suzanne E. Eckes & Kevin Brown, African American Disproportionality in School Discipline: The Divide Between Best Evidence and Legal Remedy, 54 N.Y.L. Sch. L. Rev. 1071, 1086–89 (2009/2010) (discussing the vast empirical evidence of racial disproportionality in school discipline).

While Congress, the U.S. Department of Education, and scholars have rightfully directed their attention to these disconcerting inequalities, another inequality in public education has received far less attention. This inequality also affects the everyday educational experience of many minority students and low-income students and could become more severe if policymakers decide to direct additional money to schools for security measures. I refer to the disproportionate use of strict security measures. 11See infra Part III. Strict security measures include using metal detectors, conducting random sweeps for contraband, hiring law enforcement officers or guards, controlling access to school grounds, and installing security cameras. These measures, particularly when used in combination, can create an intense, prison-like environment that deteriorates the learning climate. 12See infra Part II. As a result, the educational experiences of many minority students and low-income students are often far different—and far worse—than the everyday educational experiences of other students. Before embarking on a new phase of security upgrades, policymakers and school officials should first examine the disproportionate use of strict security measures on minorities and low-income students and consider using alternative means to reduce violence that do not degrade the learning environment.

This Article goes beyond the current literature in three ways. First, drawing on a large, national, restricted-access dataset recently released by the Department of Education, this Article empirically demonstrates that student race and student poverty are strong predictors for whether a school chooses to implement a combination of strict security measures, even after accounting for factors such as school crime, neighborhood crime, school disorder, school location, and school size. 13See infra Part III.

Second, this Article describes the current legal landscape in which these disparities have developed. Specifically, as school safety and discipline have become more pressing concerns for schools over the last three decades, courts have substantially weakened students’ Fourth Amendment rights. 14See infra Part I.A. Courts generally permit school officials to employ many types of suspicionless search practices to assist school officials in their efforts to prevent school crime, particularly when those searches are considered to be minimally intrusive. 15See infra Part I.B. In addition, as part of the recent school accountability movement, federal and state laws require students to be tested each year and impose harsh consequences for failing to meet certain standards. 16See infra Part I.D. Accordingly, schools have an incentive to push low-performing students out of schools by intensifying their surveillance methods. 17See infra Part I.D. Further, federal and state programs have encouraged the use of such measures by providing money to schools to buy equipment and hire law enforcement officers. 18See infra Part I.C. With minimal oversight from courts, the recent accountability movement, and significant funding from federal and state governments to purchase strict security measures, the conditions have been set for yet another inequality to emerge in our public schools.

Third, this Article proposes several recommendations to address these disparities. It urges school officials and policymakers to voluntarily adopt alternative, more effective means to curb school crime and violence instead of resorting to methods that rely on coercion, punishment, and fear. A growing body of research shows that programs promoting a strong sense of community and collective responsibility (a) effectively reduce school crime and student misbehavior without degrading the learning environment and (b) teach students how to resolve conflicts peacefully. Indeed, schools can do much more to prevent violence in our schools and communities by investing in these alternative measures than by purchasing security equipment and hiring law enforcement officers. 19See infra Part IV.B. In addition, this Article encourages federal and state agencies to stop providing money to schools for strict security measures and to use that money to promote these alternative measures. Finally, it recommends that the OCR take an active role in preventing the disproportionate use of strict security measures on minority students. 20In a subsequent scholarly article, I also intend to describe an alternative framework that courts should employ to evaluate students’ Fourth Amendment rights in schools. See infra note 313.

This Article proceeds in four parts. Part I discusses the development of the law that allows and motivates school officials to use a variety of strict security measures on students. Part II describes educational and sociological considerations against adopting strict security measures in schools. Additionally, it discusses the reasons why the disparate use of strict security measures on minorities is particularly harmful. Part III presents the results of the empirical analysis revealing the disparate use of strict security measures along racial and economic lines. Part IV evaluates the concerns presented in the empirical findings and recommends measures to address those concerns. It also urges school officials to adopt alternative measures that will more effectively reduce violence in our schools and communities.

I. The Current State of the Law

In the last three decades, students’ Fourth Amendment rights have steadily declined as the U.S. Supreme Court has provided school officials with more constitutional leeway to maintain order and discipline. 21See Jason P. Nance, Random, Suspicionless Searches of Students’ Belongings: A Legal, Empirical, and Normative Analysis, 84 U. Colo. L. Rev. 367 (2013); see also James E. Ryan, The Supreme Court and Public Schools, 86 Va. L. Rev. 1335, 1415 (2000) (explaining that “the Court’s decisions regarding student searches rest on the value-laden view that maintaining discipline is necessary to preserve the educational process of schools”). Scholars have long debated how far students’ constitutional rights should extend in schools. But, as Catherine Kim recently observed, “both sides of the debate share a common starting point: such restrictions must be justified, if at all, by pedagogical goals.” Catherine Y. Kim, Policing School Discipline, 77 Brook. L. Rev. 861, 866 (2012); see also, e.g., Anne Proffitt Dupre, Should Students Have Constitutional Rights? Keeping Order in the Public Schools, 65 Geo. Wash. L. Rev. 49, 53 (1996) (applauding the Supreme Court’s recent jurisprudence that restricts students’ Fourth Amendment rights because doing so enhances the ability of school officials to provide students with a serious education); Betsy Levin, Educating Youth for Citizenship: The Conflict Between Authority and Individual Rights in the Public School, 95 Yale L.J. 1647, 1648–49 (1986) (describing the need to find an equilibrium between respecting students’ constitutional rights so that students learn to value those rights, and restricting students’ rights to help school officials maintain an orderly environment conducive to learning); Ryan, supra, at 1338–43, 1424–26 (maintaining that when a school engages in its core, universal function of imparting academic and vocational skills, it should be treated more deferentially and given some constitutional leeway to achieve its goals). School officials are not required to obtain a warrant, show probable cause, or even have an individualized suspicion that a student participated in wrongdoing before conducting a search. 22See Vernonia School District 47J v. Acton, 515 U.S. 646, 653–54 (1995); New Jersey v. T.L.O., 469 U.S. 325, 340–41 (1985). The absence of any meaningful protections from random, suspicionless searches has created an environment in which school officials can now freely employ a variety of strict security measures, even when those measures have the cumulative effect of creating an intense environment that is detrimental to learning. 23See infra Part I.B. In addition, federal and state laws have motivated schools to purchase strict security measures and intensify their surveillance methods. 24See infra Part I.C. These conditions set the stage for another racial inequality to develop in our public schools.

A. The Foundational Cases

The U.S. Supreme Court evaluated students’ Fourth Amendment rights in schools for the first time in New Jersey v. T.L.O. and initiated what would become a steady decline of those rights. 25See Nance, supra note 21, at 376–94. In T.L.O., a school official searched high school freshman T.L.O.’s purse for cigarettes after a teacher claimed that she spotted T.L.O. smoking in the bathroom and T.L.O. denied the accusation. 26T.L.O., 469 U.S. at 328. During the search, the school official discovered marijuana and other materials suggesting that T.L.O. was dealing marijuana. 27Id. The school official turned the evidence over to the police, who brought delinquency charges against T.L.O. in juvenile court. 28Id. at 328–29. T.L.O. moved to suppress the evidence, arguing that the school official’s search violated the Fourth Amendment. 29Id. at 329. The Supreme Court disagreed. 30See id. at 332–33.

The Court evaluated the constitutionality of the search by balancing T.L.O.’s expectation of privacy against the school’s need to maintain an orderly environment conducive to learning. 31Id. at 339–40. The Court first explained that students have legitimate expectations of privacy in the personal items they bring to school. 32See id. at 337–38 (“A search of a child’s person or of a closed purse or other bag carried on her person, no less than a similar search carried out on an adult, is undoubtedly a severe violation of subjective expectations of privacy.” (footnote omitted)). At the same time, the Court recognized that school officials also have an important need to maintain an appropriate learning environment. 33Id. at 339 (citing Nat’l Inst. of Educ., U.S. Dep’t of Health, Educ. & Welfare, Violent Schools––Safe Schools: The Safe School Study Report to Congress 1 (1978)). To strike a balance, the Court held that school officials were not required to obtain a warrant before searching a student, and a school official’s level of suspicion need not rise to the level of “probable cause.” 34See id. at 340–41. Rather, the constitutionality of a search in school would depend on its reasonableness under the circumstances. 35Id. at 337. The determination of “reasonableness” is a two-fold inquiry: (1) “‘whether the . . . action was justified at its inception,’” and (2) “whether the search as actually conducted ‘was reasonably related in scope to the circumstances which justified the interference in the first place.’” 36Id. at 341 (quoting Terry v. Ohio, 392 U.S. 1, 20 (1968)). Using this framework, the Court concluded that the search was constitutional. 37Id. at 343.

Although the search in T.L.O. was based on individualized suspicion, the Court emphasized that it was not deciding whether individualized suspicion was essential to justify a search conducted by school authorities. 38See id. at 342 n.8 (“We do not decide whether individualized suspicion is an essential element of the reasonableness standard we adopt for searches by school authorities.”). Ten years later, however, in Vernonia School District 47J v. Acton, 39Vernonia School District 47J v. Acton, 515 U.S. 646 (1995). the Court determined that individualized suspicion was unnecessary to conduct minimally intrusive searches of students when certain conditions are present. 40Id. at 653.

In Vernonia, the Court evaluated a school district’s suspicionless drug-testing program on student athletes. 41Id. at 650. Students wishing to participate in interscholastic sports would be tested at the beginning of the season, and each week of the season a student, under the supervision of two adults, would blindly select ten percent of the student-athlete population for drug testing. 42Id. In fall 1991, James Acton, a seventh-grade student who was well behaved and who did not have a drug problem, signed up to play football. 43See Robert M. Bloom, The Story of Pottawatomie County v. Lindsay Earls: Drug Testing in the Public Schools, in Education Law Stories 337, 346 (Michael A. Olivas & Ronna Greff Schneider eds., 2008) (“At the time, James said, ‘I was like one of the smartest kids in class. I never got a referral (to the principal’s office) and I thought that was probably enough for them to see I wasn’t taking drugs.’”). The school officials would not allow James to participate in school athletics because he and his parents refused to sign a drug-testing consent form. 44Vernonia, 515 U.S. at 651. The Actons claimed that Vernonia’s drug-testing program violated the Fourth Amendment, but the Court disagreed. 45Id. at 651, 664–65.

After dispensing with the individualized suspicion requirement, 46Id. at 653 (“The school search we approved in T.L.O., while not based on probable cause, was based on individualized suspicion of wrongdoing. As we explicitly acknowledged, however, ‘the Fourth Amendment imposes no irreducible requirement of such suspicion.’” (quoting New Jersey v. T.L.O., 469 U.S. 325, 342 n.8 (1985)) (internal quotation marks omitted)). the Court established a new framework for evaluating suspicionless searches. Under its new framework, the Court balanced three factors: (1) “the scope of the legitimate expectation of privacy at issue”; (2) “the character of the intrusion that is complained of”; and (3) “the nature and immediacy of the governmental concern at issue . . . and the efficacy of this means for meeting it.” 47Id. at 654, 658, 660. The Court applied these factors in the following manner. First, while acknowledging that students retained some expectation of privacy while at school, the Court explained that the scope of students’ privacy rights “are different in public schools than elsewhere.” 48Id. at 656. According to the Court, “the ‘reasonableness’ inquiry [could not] disregard the schools’ custodial and tutelary responsibilit[ies],” requiring students’ expectation of privacy to be diminished. 49See id. Next, the Court considered drug tests to be minimally intrusive because drug-testing conditions resembled conditions that students commonly encounter in public restrooms. 50Id. at 658. Finally, the Court determined that the school district’s interest in deterring drug use was important and immediate in light of the rampant use of drugs among the student athletes in Vernonia School District, outweighing any right of privacy the student-athletes possessed. 51Id. at 661–63.

In her dissent, Justice O’Connor criticized the Court’s decision to further restrict students’ Fourth Amendment rights by eliminating the individualized suspicion requirement. 52Id. at 681 (O’Connor, J., dissenting). Justice O’Connor believed that suspicionless searches send a harmful pedagogical message to students: that school officials do not trust students, and that students have to prove their innocence to school officials. 53Id. at 682. She argued,

[I]ntrusive, blanket searches of schoolchildren, most of whom are innocent, for evidence of serious wrongdoing are not part of any traditional school function of which I am aware. Indeed, many schools, like many parents, prefer to trust their children unless given reason to do otherwise. As James Acton’s father said on the witness stand, “[suspicionless testing] sends a message to children that are trying to be responsible citizens . . . that they have to prove that they’re innocent . . . , and I think that kind of sets a bad tone for citizenship.” 54Id. (all but the first alteration in original) (quoting Tr. 9 (Apr. 29, 1992)).

Seven years later, in Board of Education v. Earls, 55Bd. of Educ. v. Earls, 536 U.S. 822 (2002). the Court arguably abridged students’ Fourth Amendment rights even further. There, the Court upheld a program requiring students enrolled in extracurricular activities to submit to random drug testing. 56Id. at 825. But unlike in Vernonia, where the Court justified those suspicionless searches in part because the drug problem in the district was “alarming,” the school district in Earls had made no such showing. 57Id. at 849 (Ginsburg, J., dissenting).

In Earls, Lindsay Earls, a student enrolled in “the show choir, the marching band, the Academic Team, and the National Honor Society,” challenged Pottawatomie School District’s drug-testing policy as unconstitutional. 58Id. at 826–27 (majority opinion). Earls argued that Pottawatomie had failed to identify a special need for implementing its random drug-testing program, because it had not demonstrated that her school had a drug problem. 59Id. at 827. In a five-to-four decision, the Supreme Court upheld Pottawatomie’s drug-testing policy. 60Id. at 824–25.

The Court balanced the same three factors that it did in Vernonia, largely reaching the same conclusions. 61Id. at 830–38; see also Nance, supra note 21, at 384–87. However, while noting that Pottawatomie “presented specific evidence of drug use,” the Court held that it was unnecessary to identify a drug abuse problem before imposing a suspicionless drug-testing policy. 62Earls, 536 U.S. at 834–35. The Court justified the program because “the nationwide drug epidemic makes the war against drugs a pressing concern in every school.” 63Id. at 834. This broad holding has provided a clear path for schools to conduct a sweeping array of suspicionless search practices without first having to demonstrate a drug or weapons problem. 64See infra Part III. But see Nance, supra note 21, at 391–94 (arguing that the Fourth Amendment requires school officials to have particularized evidence of a substance use or weapons problem before performing suspicionless searches that are “highly intrusive,” such as a search through a student’s personal belongings).More recently, the Supreme Court addressed the constitutionality of a student search in Safford Unified School District Number 1 v. Redding, 557 U.S. 364 (2009). In Redding, the Court did not evaluate the constitutionality of a random, suspicionless search as it did in Vernonia and Earls; instead, the Court evaluated the legality of a strip search performed on a thirteen-year-old female student who was accused of bringing unauthorized prescription and over-the-counter drugs to school. Id. at 368. Relying on the same two factors the Court presented in T.L.O., the Court concluded that the search violated the Constitution because it was excessively intrusive in light of the age of the student and the nature of the school violation. Id. at 378 (citing New Jersey v. T.L.O., 469 U.S. 325, 342 (1985)). Nevertheless, the Court refused to hold the school officials accountable because the school officials were protected under the qualified immunity doctrine. Id. at 378–79. As some scholars have noted, this case demonstrates how difficult it is for students to recover for unlawful searches. E.g., Barry C. Feld, T.L.O. and Redding’s Unanswered (Misanswered) Fourth Amendment Questions: Few Rights and Fewer Remedies, 80 Miss. L.J. 847, 947–54 (2011) (explaining the impediments for students to recover for unlawful searches).

B. Lower Courts Permit Schools to Employ a Host of Random, Suspicionless Search Practices

The foundational cases illustrate that as school safety and discipline have become more pressing concerns, students’ Fourth Amendment rights have steadily declined. As a result of this movement in the law, lower courts routinely justify the use of a variety of random, suspicionless search practices in schools. For example, courts uniformly permit schools to use metal detectors (whether hand-held or walk-through). 65Cases that permit the use of metal detectors include Hough v. Shakopee Pub. Sch., 608 F. Supp. 2d 1087, 1104 (D. Minn. 2009); In re F.B., 726 A.2d 361, 366 (Pa. 1999); In re Latasha W., 70 Cal. Rptr. 2d 886, 886–87 (Ct. App. 1998); and State v. J.A., 679 So. 2d 316, 319–20 (Fla. Dist. Ct. App. 1996). Courts uphold random, suspicionless searches of students’ lockers. 66Cases that uphold random locker searches include State v. Jones, 666 N.W.2d 142, 150 (Iowa 2003); In re Isiah B., 500 N.W.2d 637, 641 (Wis. 1993); and In re Patrick Y., 746 A.2d 405, 414–15 (Md. 2000). For an extended discussion on the disagreement among courts regarding whether students possess an expectation of privacy in their lockers, see Kim et al., supra note 10, at 115–17; Feld, supra note 64, at 933–37; and Nance, supra note 21, at 411–12 and accompanying notes. Further, there is no rational basis to conclude that students should lose their expectation of privacy in their personal belongings simply because they place them in their lockers. See Nance, supra note 21, at 411–12 and accompanying notes. Further, if the issue were presented to courts, courts most likely would justify the use of surveillance cameras in public spaces at schools on the theory that recording individuals in public places is minimally intrusive. 67See, e.g., United States v. Taketa, 923 F.2d 665, 677 (9th Cir. 1991) (“Videotaping of suspects in public places, such as banks, does not violate the [F]ourth [A]mendment . . . .”). In certain locations, however, video surveillance—particularly surreptitious video surveillance—inappropriately intrudes on liberties that are fundamental to human dignity. See Brannum v. Overton Cnty. Sch. Bd., 516 F.3d 489, 499 (6th Cir. 2008) (holding that surreptitious video surveillance of a student locker room violates the Fourth Amendment because the surveillance intrudes on a student’s inherent right to bodily privacy).

But perhaps more importantly, there is no law prohibiting school officials from using a combination of these strict security measures, even when their cumulative use creates an intense environment that is not conducive to a healthy learning climate. For example, many schools monitor students by using a combination of metal detectors, random sweeps, surveillance cameras, locked gates, and law enforcement officers. 68See infra Part III. While courts may deem each of these measures as “minimally intrusive” or not a cognizable search at all under the Fourth Amendment, arguably the cumulative effect of using all of these measures together amounts to a substantial invasion of students’ privacy, harming students’ educational progress. 69See infra Part IV. The fact that the cumulative effect of using several strict measures together arguably amounts to a substantial invasion of students’ privacy provides justification for courts to apply a modified framework under the Fourth Amendment. I will explore this concept in greater detail in a subsequent scholarly article.

C. Federal and State Programs Have Encouraged the Use of Strict Security Measures

Not only do courts permit the use of strict security measures, but federal and state programs have encouraged the use of such measures by providing money to schools to buy equipment and hire law enforcement officers. For example, under the U.S. Department of Justice’s Community Oriented Policing Services (COPS) initiatives, schools could request up to $500,000 to support half the cost of a security program. 70See Office of Cmty. Oriented Policing Servs., 2011 Secure Our Schools Program, U.S. Dep’t of Justice, http://www.cops.usdoj.gov/pdf/2011AwardDocs/CSPP-SOS-CHP/SOSMethodology.pdf (last visited Aug. 12, 2013). Schools used the awarded funds for metal detectors, locks, lighting, security assessments, security training of personnel, coordination with local law enforcement, and other deterrent measures. 71Id. On September 8, 2011, COPS announced more than $13 million in grants to local law enforcement agencies and municipalities to enhance school safety in thirty-two states. 72See Press Release, U.S. Dep’t of Justice, Office of Cmty. Oriented Policing Servs., U.S. Department of Justice COPS Office Announces over $13 Million in School Safety Grants (Sept. 8, 2011), available at http://www.cops.usdoj.gov/Default.asp?Item=2599. Since 1995, COPS has invested approximately $913 million in school security measures across the country. 73See Office of Cmty. Oriented Policing Servs., Fact Sheet: Secure Our Schools, U.S. Dep’t of Justice (Sept. 2011), http://www.cops.usdoj.gov/pdf/2011AwardDocs/CSPP-SOS-CHP/2011-SOS-Post-FactSheet.pdf. In another example, under the Safe and Drug-Free Schools and Communities Act, 7420 U.S.C. §§ 7101–7165 (2006). Congress authorized disbursements to state and local education agencies for “[a]cquiring and installing metal detectors, electronic locks, surveillance cameras, or other related equipment and technologies.” 75Id. § 7115(b)(2)(E)(ii). Notably, the grant application procedures make clear that the Department of Education views school security equipment as only one component to an overall strategy designed to create a safe and healthy learning environment. See U.S. Dep’t of Educ. et al., Safe Schools/Healthy Students: Information and Application Procedures for Fiscal Year 2009, at 22 (2009), available at http://www2.ed.gov/programs/dvpsafeschools/2009-184l.pdf. In the 2009 application, no more than ten percent of the grant could be used to fund security equipment. Id. The remainder of the grant must be used for programs designed to address student safety and health holistically by providing behavioral, social, and emotional support; mental health services; early childhood social and emotional learning programs; and alcohol, tobacco, and other drug prevention activities for students, their families, and the community. Id. at 22–24. This funding initiative currently is suspended. U.S. Dep’t of Educ., Safe Schools-Healthy Students Initiative: Application Information, ED.gov, http://www2.ed.gov/programs/dvpsafeschools/applicant.html (last modified Nov. 4, 2011).

In addition, several states currently make grants available to schools for strict security measures, 76See Ala. Code § 41-15B-2.2 (2012) (“School Safety Enhancement Programs eligible for grants shall be designed to prevent or reduce violence in the schools . . . . The programs shall relate to one or more of the following: . . . (v) Safety plans involving the use of metal detectors, other security devices, uniforms, school safety resource officers, or other personnel employed to provide a safe school environment.”); Ga. Code Ann. § 20-2-1185 (West 2012) (“A public school may request funding assistance from the state for the installation of safety equipment including, but not limited to, video surveillance cameras, metal detectors, and other similar security devices.”); 24 Pa. Stat. Ann. § 13-1302-A (West 2012) (“[T]he office is authorized to make targeted grants to school entities to fund programs which address school violence, including . . . metal detectors, protective lighting, surveillance equipment . . . and training in the use of security-related technology.”); S.C. Code Ann. § 59-66-30 (2012) (“Using funds appropriated by the General Assembly, each public middle, junior high, and high school in the State must be equipped with one hand-held metal detector.”). and the Newtown tragedy has motivated other state legislatures to pass or consider passing legislation that will funnel additional money to schools for school security. 77See supra note 3. No doubt such initiatives will continue to serve as a strong impetus for many schools to adopt strict security measures, especially as many parents and other community members demand that policymakers and school officials respond to the Newtown shootings.

D. The School Accountability Laws May Motivate Schools to Adopt Strict Security Measures

Further, school accountability laws may motivate some school officials to adopt strict security measures as part of an overall effort to push low-performing students out of their schools. Federal and state laws require students to be tested each year and carry harsh consequences for schools that fail to meet certain standards. For example, under the No Child Left Behind Act of 2001 (NCLBA), 78Pub. L. No. 107-110, 115 Stat. 1425 (2002) (codified as amended in scattered sections of 20 U.S.C.). all schools receiving federal funds must test students at various points during grades three through twelve in reading, language arts, math, and science and demonstrate improvement across all subgroups of students. 79Torin Monahan & Rodolfo D. Torres, Introduction to Schools Under Surveillance: Cultures of Control in Public Education 1, 5 (Torin Monahan & Rodolfo D. Torres, eds. 2010); U.S. Dep’t of Educ., Testing: Frequently Asked Questions, ED.gov, http://www2.ed.gov/nclb/accountability/ayp/testing-faq.html (last visited on Aug. 12, 2013). Schools that do not meet these standards risk receiving a negative label, being placed on probation, and eventually being taken over by the state. 80Monahan & Torres, supra note 79, at 5. Many scholars argue that schools have an incentive to push low-performing students out by suspending, expelling, or referring low-performing students to the juvenile justice system under their zero-tolerance policies. 81See, e.g., Advancement Project, Test, Punish, and Push Out: How “Zero Tolerance” and High-Stakes Testing Funnel Youth into the School-to-Prison Pipeline 28–33 (2010), available at http://b.3cdn.net/advancement/d05cb2181a4545db07_r2im6caqe.pdf; Linda Darling-Hammond, Race, Inequality, and Educational Accountability: The Irony of ‘No Child Left Behind, 10 Race, Ethnicity & Educ. 245, 252–55 (2007); Deborah Gordon Klehr, Addressing the Unintended Consequences of No Child Left Behind and Zero Tolerance: Better Strategies for Safe Schools and Successful Students, 16 Geo. J. on Poverty L. & Pol’y 585, 602–03 (2009); Nance, supra note 21, at 397; James E. Ryan, The Perverse Incentives of the No Child Left Behind Act, 79 N.Y.U. L. Rev. 932, 969–70 (2004). Under the Federal Gun-Free Schools Act, states receiving federal education funds must have a state law that requires schools to expel students for at least one year for bringing a firearm to school. See 20 U.S.C. § 7151(b)(1). Although the NCLBA reenacted a softened version of this law by permitting superintendents to modify the expulsion requirement on a case-by-case basis, see id.; see also Federal Law on Guns in Schools, L. Center to Prevent Gun Violence (May 21, 2012), http://smartgunlaws.org/federal-law-on-guns-in-schools/, many states and schools adopted laws and policies modeled after the original Gun-Free Schools Act of 1994 by creating strict rules that impose predetermined consequences for certain acts, irrespective of the surrounding circumstances or intent of the students. See Klehr, supra, at 589. These so-called zero tolerance policies have pushed more students out of schools and have increased referrals to the juvenile justice system. Id. at 590. As scholar Jim Ryan noted, “the temptation to exclude low-performing students, enhanced by the NCLBA, can hardly be denied: One less student performing below the proficiency level increases the overall percentage of students who have hit that benchmark.” 82Ryan, supra note 81, at 969. Indeed, this observation is bolstered by the fact that the NCLBA does not have the same accountability standards for graduation rates as it does for testing. 83Id. at 970. For schools that adopt this exclusionary ethos, strict security measures play an integral role. When school officials using intense surveillance methods discover low-performing students carrying contraband, they may expel these students, transfer them to alternative educational settings, or refer them to law enforcement officials to avoid having to count their low test scores against the school’s overall pass rate.

II. Educational and Sociological Considerations Against Strict Security Measures

In the aftermath of Newtown, policymakers, educators, and, of course, parents want nothing more than to keep children safe. Not only must school officials protect students from dangerous intruders, they also must address safety and discipline issues from their own students. 84See Timeline: School Violence in the U.S., CNN.com (Dec. 14, 2012, 12:11 PM), http://www.cnn.com/2012/12/14/us/timeline-school-violence/index.html (detailing the number of violent acts in schools committed by students and outsiders). Although few would argue that using strict security measures in schools is ideal, many will respond that the safety of our children in schools is paramount, which overrides any pedagogical concerns their use creates.

The safety of children in schools is extremely important. No one can plausibly argue otherwise. But a difficult truth that all must accept is that those determined to commit violent acts inside schools will succeed. 85Cf. Amanda Terkel, Columbine High School Had Armed Guard During Massacre in 1999, Huffington Post (Dec. 21, 2012, 11:07 PM), http://www.huffingtonpost.com/2012/12/21/columbine-armed-guards_n_2347096.html (describing the Columbine High School shooting tragedy that occurred in the presence of an armed security officer). Indeed, the Columbine High School shootings occurred in the presence of armed guards and metal detectors. 86See id. (noting presence of armed guards); Marcus Wright, Experts Say Intrusive Security at Public Schools Reproduces Social Inequality, Mich. Citizen (Nov. 15, 2012), http://michigancitizen.com/dps-eaa-tighten-security/ (noting presence of metal detectors). It is simply impossible to ensure the safety of all our children at school at all times. 87See Arne Duncan, Resources for Schools to Prepare for and Recover from Crisis, Homeroom (Dec. 17, 2012), http://www.ed.gov/blog/2012/12/resources-for-schools-to-prepare-for-and-recover-from-crisis/ (explaining that not all tragedies in schools can be prevented).

Ironically, although highly publicized acts of school violence skew the public’s perception of school safety, 88See, e.g., Monahan & Torres, supra note 79, at 2–3 (explaining that although schools continue to be the safest places for children, the threat of another school shooting haunts the social imagery); Matthew J. Mayer & Peter E. Leone, School Violence and Disruption Revisited: Equity and Safety in the School House, 40 Focus on Exceptional Child. Sept. 2007, at 1, 6 (“[M]edia coverage of school violence has shaped the public’s beliefs, and in many cases has led to a distorted perception of violence in schools, as well as adolescent violence more generally.” (citations omitted)); Randy Borum et al., What Can Be Done About School Shootings? A Review of the Evidence, 39 Educ. Researcher 27, 27 (2010) (“[S]chool shootings receive such intense publicity, and are such inherently disturbing events, that they generate an inflated perception of danger.” (citation omitted)). empirical evidence demonstrates that schools remain among the safest places for children. 89See Duncan, supra note 87 (“Schools are among the safest places for children and adolescents in our country, and, in fact, crime in schools has been trending downward for more than a decade.”); see also Barbara Fedders, Jason Langberg & Jennifer Story, School Safety in North Carolina: Realities, Recommendations & Resources 4 (May 2013), available at http://www.legalaidnc.org/public/learn/media_releases/2013_MediaReleases/school-safety-in-north-carolina.pdf (“School violence that results in death is extremely rare. Young people are much more likely to be harmed in the home or on the streets than they are in schools.” (footnote omitted)); Randall R. Beger, The “Worst of Both Worlds”: School Security and the Disappearing Fourth Amendment Rights of Students, 28 Crim. Just. Rev. 336, 338 (2003) (“Contrary to popular belief, schools remain among the safest places for children.”); Borum et al., supra note 88, at 27 (explaining that the number of homicides that occur on school grounds represents less than one percent of the annual homicides of children ages five to eighteen, and that “any given school can expect to experience a student homicide about once every 6,000 years”); Pedro A. Noguera, Schools, Prisons, and Social Implications of Punishment: Rethinking Disciplinary Practices, 42 Theory into Prac. 341, 343 (2003) (“Despite surveys that suggest a growing number of teachers and students fear violence in school, schools in the United States are generally safe places.” (citation omitted)). For example, during the 2008–2009 school year, there were a total of 1,579 homicides of youth ages five to eighteen, but only seventeen of those homicides occurred at school. 90See Simone Robers et al., Indicators of School Crime and Safety: 2011, at iv (2012), available at http://nces.ed.gov/pubs2012/2012002rev.pdf. Likewise, in 2008, four in one thousand students ages twelve to eighteen were victims of serious violent crimes such as rape or assault at school, but eight in one thousand students of the same age group were victims of those crimes away from school. 91See Simone Robers et al., Indicators of School Crime and Safety: 2010, at iv (2010). But see id. (“In 2008, students ages 12 to 18 were victims of about 1.2 million nonfatal crimes (theft plus violent crime) at school, compared to about 1 million nonfatal crimes away from school.” (footnotes omitted)). Of course, some schools face serious safety and discipline issues that must be addressed. But before school officials and policymakers decide to make significant financial investments in strict security measures, they should examine the educational and sociological considerations against using strict security measures and consider alternative methods that reduce violence more effectively and do not harm the learning environment.

A. Strict Security Measures Are Not Aligned with Students’ Best Interests

When student Minerva Dickson saw her high school for the first time in Brooklyn, New York, it reminded her of a prison. 92Daryl Khan, Perps or Pupils? Safety Policy Creates Friction in New York City Schools, Center for Pub. Integrity (Sept. 21, 2012, 6:00 AM), http://www.publicintegrity.org/2012/09/21/10950/perps-or-pupils-safety-policy-creates-friction-new-york-city-schools. Every day before entering school, she would wait in a long line as each student went through various security checks. 93Id. First, she would swipe an identification card through a machine. 94Id. Once cleared, she would head to the metal detectors that were monitored by several police officers. 95Id. While the police officers watched, Minerva would remove her jewelry, hairpins, and shoes, then place her personal bags on a conveyer belt to be scanned. 96Id. Next, an officer would run a security wand as she stood with her arms out and legs spread. 97Id. Minerva would then collect her things, put her shoes on, and hurry to her first class. 98Id. When asked how she felt about the security process, she responded, “They treat[] us like criminals. It ma[kes] me hate school. When you cage up students like that it doesn’t make us safe, it makes things worse.” 99Id.

In another example, in September 2012, Bushwick Community High School, a transfer high school for troubled teenagers, installed metal detectors for the first time. 100Michael Powell, In a School Built on Trust, Metal Detectors Inject Fear, N.Y. Times (Sept. 17, 2012), http://www.nytimes.com/2012/09/18/nyregion/in-a-brooklyn-school-metal-detectors-inject-fear.html. Because the junior high school with which Bushwick shares its building developed a reputation for violence, city education officials determined that all students in the building should walk through metal detectors every day. 101Id. This security measure had a profound effect on many Bushwick students. One student said that the first time she saw those machines she felt sick and wanted to run. 102Id. When the security guard told her that she would feel different about the detectors after another student injured her with a weapon, she replied, “Really? When I came here last year, I remember feeling one thing: This is family, and that makes me feel safe.” 103Id. (internal quotation marks omitted). Another student remarked that the detectors “add[ed] to [his] stress and isolation.” 104Id. The longtime dean and teacher maintained that “[t]he glue in this building is love, and the metal trespasses on that.” 105Id. (internal quotation marks omitted). Michael Powell, the author of the article, aptly observed:

Any school that works has an ethos, a culture and a language known to these particular students and these particular teachers. Here, on Palmetto Street in a very nonhipster corner of Bushwick, that ethos is scouring honesty and trust. You cannot pull one over; you will be challenged, but you will not be discarded. 106Id.

The new security measures, however, threatened that ethos.

Educators have long understood that trust is a vital component to establishing a healthy climate conducive to learning. 107See David Domenici & James Forman Jr., What It Takes to Transform a School Inside a Juvenile Justice Facility: The Story of the Maya Angelou Academy, in Justice for Kids: Keeping Kids Out of the Juvenile Justice System, supra note 10, at 283, 289 (“High-achieving schools are places where a culture of trust dominates.”); Roger D. Goddard, Megan Tschannen-Moran & Wayne K. Hoy, A Multilevel Examination of the Distribution and Effects of Teacher Trust in Students and Parents in Urban Elementary Schools, 102 Elementary Sch. J. 3, 3 (2001) (explaining that trusting in others is an important element of the learning process); Megan Tschannen-Moran & Wayne K. Hoy, A Multidisciplinary Analysis of the Nature, Meaning, and Measurement of Trust, 70 Rev. Educ. Res. 547, 547 (2000) (explaining that trusting in others is an important element of the learning process). Indeed, students learn best when they feel respected, have positive attitudes toward their classmates and their teachers, and feel good about themselves. 108See Donna Lieberman, Over-Policing in Schools on Students’ Education and Privacy Rights, NYCLU, http://www.nyclu.org/content/over-policing-schools-students-education-and-privacy-rights (last visited Aug. 12, 2013). Strict security measures, however, undermine the climate of trust needed to effectively educate children. Their use sends a negative message that students are harmful, dangerous, and prone to commit illegal, violent acts. 109 See Martin R. Gardner, Student Privacy in the Wake of T.L.O.: An Appeal for an Individualized Suspicion Requirement for Valid Searches and Seizures in the Schools, 22 Ga. L. Rev. 897, 943 (1988). That message “sour[s] students’ attitudes toward school and school authorities and undermine[s] a positive, respectful academic environment.” 110Paul Hirschfield, School Surveillance in America: Disparate and Unequal, in Schools Under Surveillance: Cultures of Control in Public Education, supra note 79, at 38, 46.

According to scholar Paul Hirschfield, strict security measures produce barriers between students and their schools and are “a frequent cause of disunity or discord within the school community.” 111Id. Scholar Martin Gardner explained it this way: “In a very real sense, each and every student stands accused, has become a ‘suspect,’ in generalized school searches, especially given the special relationship of trust which supposedly exists between student and teacher.” 112Gardner, supra note 109, at 943. Searches taking place in schools are much different than searches in other environments such as airports. 113Id. As Gardner reasons,

Surely a student even indirectly accused by his teacher as a possible thief or drug user suffers a greater indignity and loss of self-esteem by being subjected to a generalized search than does an airline passenger passing through a metal detector or a driver [through] a checkpoint. Far from “morally neutral,” school searches are instead particularly rife with moral overtones. 114Id. (footnotes omitted).

Further, airline passengers voluntarily decide to subject themselves to being searched. Students, on the other hand, are subject to mandatory school attendance laws and must submit to these searches against their will.

After interviewing students subjected to such security measures, scholar Jen Weiss reported that strict security measures caused students to “feel consistently watched, to distrust, hide from, and avoid authority figures.” 115Jen Weiss, Scan This: Examining Student Resistance to School Surveillance, in Schools Under Surveillance: Cultures of Control in Public Education, supra note 79, at 213, 227. She found that instead of promoting a greater sense of safety and civic-minded school community, students felt a sense of danger and disillusion. 116Id. at 213–14. She concluded that students in schools using strict security measures “experience firsthand what it is to be monitored, contained, and harassed, all in the name of safety and protection.” 117Id. at 213. Similarly, Donna Liebermann, Executive Director of the New York Civil Liberties Union, testified that strict security measures produce environments that are not conducive to educational and social growth. 118See Lieberman, supra note 108. Instead, strict security measures “foster environments where children perceive that they are being treated as criminals; where they are diminished by such perceptions; and where they, consequently, cultivate negative attitudes toward their schools.” 119Id. One teacher put it this way: “The medium is the message. And the message that [strict security measures] give[] out is that we are afraid of our students.” 120Carol Ascher, Gaining Control of Violence in the Schools: A View from the Field, ERIC Digest, Sept. 1994, at 1, 4 (1994) (internal quotation marks omitted), available at http://www.eric.ed.gov/PDFS/ED377256.pdf (“Rather than offering reassurance, metal detectors and other mechanical devices, as well as security forces, are seen as providing a false sense of safety, if not a harsh symbol of the failure to create safe schools.”).

Notably, while strict security measures negatively affect the learning environment, several studies call into doubt whether these measures reduce school violence at all. 121See Advancement Project, Education on Lockdown: The Schoolhouse to Jailhouse Track 8 (2005), available at http://www.advancementproject.org/sites/default/files/publications/FINALEOLrep.pdf (explaining that while strict security measures “produce a perception of safety, there is little or no evidence that they create safer learning environments or change disruptive behaviors”); John Blosnich & Robert Bossarte, Low-Level Violence in Schools: Is There an Association Between School Safety Measures and Peer Victimization?, 81 J. Sch. Health 107, 107 (2011) (concluding that school security measures did not reduce violent behaviors related to bullying); Abigail Hankin, Marci Hertz & Thomas Simon, Impacts of Metal Detector Use in Schools: Insights from 15 Years of Research, 81 J. Sch. Health 100, 105 (2011) (concluding that there is insufficient evidence to demonstrate whether metal detectors reduce school violence); Matthew J. Mayer & Peter E. Leone, A Structural Analysis of School Violence and Disruption: Implications for Creating Safer Schools, 22 Educ. & Treatment Child. 333, 350, 352 (1999) (finding that student disorder and victimization were higher in schools using strict security measures than in schools that did not use such measures); Richard E. Redding & Sarah M. Shalf, The Legal Context of School Violence: The Effectiveness of Federal, State, and Local Law Enforcement Efforts to Reduce Gun Violence in Schools, 23 Law & Pol’y 297, 319 (2001) (“It is hard to find anything better than anecdotal evidence” to demonstrate that strict security measures such as metal detectors and guards reduce violence in schools). But see Ctrs. for Disease Control & Prevention, Violence-Related Attitudes and Behaviors of High School Students—New York City, 1992, 42 Morbidity & Mortality Wkly. Rep. 773 (1993), available at http://www.cdc.gov/mmwr/preview/mmwrhtml/00022011.htm (showing that students who attended schools using metal detectors were less likely to carry a weapon inside a school (7.8% versus 13.6%), but the presence of metal detectors did not reduce overall school violence); Renee Wilson-Brewer & Howard Spivak, Violence Prevention in Schools and Other Community Settings: The Pediatrician as Initiator, Educator, Collaborator, and Advocate, 94 Pediatrics 623, 626–27 (1994) (stating that one school system in New York City reported that after the school security staff began using hand-held metal detectors to conduct unannounced lobby searches of students at the beginning of the school day, weapon-related incidents decreased in thirteen of fifteen schools); Rachana Bhatt & Tomeka Davis, The Impact of Random Metal Detector Searches on School Violence, Contraband Possession, and Perceptions of Safety 22–23 (Dec. 2012) (unpublished manuscript), available at http://www2.gsu.edu/~ecorrb/index_files/RandomSearch.pdf (finding that, when comparing two geographically adjacent school districts in Florida, the school district that used random metal detector searches reduced the probability of a student bringing a weapon to school). Research on the effectiveness of school security measures is extremely limited, especially causal research demonstrating the effects of strict security measures. See Mayer & Leone, supra note 88, at 12. In fact, many argue that implementing strict security measures increases student behavioral problems and crime by alienating students. 122See Ascher, supra note 120, at 5 (maintaining that strict security measures “increase, rather than alleviate, tension in schools”); Beger, supra note 89, at 340 (citing several studies demonstrating that “aggressive security measures produce alienation and mistrust among students”); Michael Easterbrook, Taking Aim at Violence, 32 Psychol. Today, July–Aug. 2009 at 52, 56 (providing evidence that strict security measures alienate students); Clifford H. Edwards, Student Violence and the Moral Dimensions of Education, 38 Psychol. Schs. 249, 250 (2001) (arguing that “intrusive strategies are likely to undermine the trust needed to build cooperative school communities capable of really preventing violence”); Mayer & Leone, supra note 121, at 350, 352 (finding that student disorder and victimization were higher in schools using strict security measures than in schools that did not use such measures); Pedro A. Noguera, Preventing and Producing Violence: A Critical Analysis of Responses to School Violence, 65 Harv. Educ. Rev. 189, 190–91 (1995) (arguing that a “get tough” approach does not create a safe environment because coercive measures create mistrust and resistance among the student body); accord Aaron Kupchik, Homeroom Security: School Discipline in an Age of Fear 7, 15–18 (2010) (explaining that student misbehavior is likely to increase rather than decrease when students perceive they are treated unfairly and with disrespect). For example, after conducting an empirical study involving almost 7,000 students, scholars Matthew Mayer and Peter Leone concluded that schools’ reliance on metal detectors, locked doors, locker checks, and security guards may lead to more disorder, crime, and violence in schools. 123See Mayer & Leone, supra note 121, at 349. They observed:

Overall, the model may suggest that less attention should be paid to running schools in an overly restrictive manner and rather, schools should concentrate more on communicating individual responsibility to students. Viewed in the context of a reciprocal relationship, the data may suggest that disorder and restrictive management of the school premises may go hand in hand and may feed off of each other. 124Id. at 351.

A description of one high school in the Chicago Public Schools that relies on strict security measures is illustrative. Scholars Matthew Steinberg, Elaine Allensworth, and David Johnson, with the cooperation of the Consortium on Chicago School Research at the University of Chicago Urban Education Institute, conducted an in-depth study of school safety in the Chicago Public Schools and observed the following:

Inside Lake Erie, the physical environment is dominated by crowd-control mechanisms: metal detectors, which are present throughout CPS high schools, greet students upon entering; folding tables corral students at the main entrance and at informal security “checkpoints” throughout hallways; folding metal gates are pulled across entrances to stairwells and padlocked. There is a constant police presence outside and inside the school. 125Matthew P. Steinberg, Elaine Allensworth & David W. Johnson, Student and Teacher Safety in Chicago Public Schools: The Roles of Community Context and School Social Organization 15 (2011) (emphasis removed), available at http://ccsr.uchicago.edu/sites/default/files/publications/SAFETY%20IN%20CPS.pdf.

Yet, despite all of these strict security measures, substantial violence and disorder abounded in that school. Steinberg, Allensworth, and Johnson continue:

Nearly all teachers at Lake Erie report problems with robbery in the building, gang activity, fights, disorder, and disrespect, and three-quarters of teachers report that students threaten them with violence. Interactions between students and teachers are frequently hostile and mutually disrespectful; students’ and teachers’ frustration with one another are easily visible. An algebra teacher at Lake Erie complains that constant disruption “impedes the teaching process”; repeated conflicts make it difficult, he continues, for teachers “to reach students who want to learn as deeply as you know [they] could.” Another teacher observes, “I see behavior problems I have never seen before . . . I get cursed out almost daily.” . . . Violence inside and outside the school creates a climate of mistrust, antagonism, and fear. 126Id. (emphasis removed).

Rather than relying on strict security measures and punitive discipline approaches to reduce school violence, the study concludes that schools “must make stronger efforts to foster trusting, collaborative relationships with students and their parents.” 127Id. at 2.

Further, one must not overlook that strict security measures cannot completely prevent even serious acts of violence from occurring at school. For example, in 2004, a student in a Washington, D.C., high school was shot by another student inside the school. 128See Sara Neufeld & Sumathi Reddy, Violent Week Renews Metal Detector Debate, Balt. Sun, Oct. 14, 2006, at 1A, available at http://articles.baltimoresun.com/2006-10-14/news/0610140131_1_metal-detectors-school-students-park-elementary-school. The school had metal detectors, perimeter fencing, and guards. 129Id. The infamous Columbine massacre occurred in a school that used metal detectors and had armed guards. 130See Wright, supra note 86. According to Ronald Stephens, an executive director of the National School Safety Center, “[R]ule-followers will follow the rules. Rule-breakers will break the rules. . . . Sometimes the metal scanning is more of a comfort.” 131Neufeld & Reddy, supra note 128 (internal quotation marks omitted). Scholar Pedro Noguera maintains that most students he spoke with during his visits to urban schools understood that anyone who wanted to bring a weapon into a school could get it into the building without being discovered by a metal detector. 132Noguera, supra note 122, at 193; see also Ascher, supra note 120, at 5 (“[T]hose few students intent on bringing in weapons are inevitably a step ahead of the security devices, which means that enforcement activities alone cannot create a safe school.”); Neufeld & Reddy, supra note 128 (explaining that students interviewed in focus groups maintained that it was “easy to get around” metal detectors). Scholar Crystal Garcia reports that only thirty-two percent of the school safety administrators she surveyed believed that metal detectors effectively prevented violent crime in their schools. 133 See Crystal A. Garcia, School Safety Technology in America: Current Use and Perceived Effectiveness, 14 Crim. Just. Pol’y Rev. 30, 40 (2003).

But more importantly, even if strict security measures deter some students from bringing dangerous weapons to schools, these measures do not support long-term solutions needed to effectively prevent school and community violence. Lasting solutions include helping children develop socially responsible behavior, build social and emotional stability, understand collective responsibility, and resolve conflict peacefully. 134See infra Part IV.B. Unfortunately, strict security measures exacerbate underlying problems by forging barriers of adversity and mistrust between students and educators. 135See supra notes 121–22. Furthermore, investing millions of dollars in security equipment and law enforcement officers diverts scarce resources away from other educational and mental health services that students desperately need. 136See, e.g., Hankin, Hertz, & Simon, supra note 121, at 101 (“The Cleveland public school system estimated a cost of $3.7 million to incorporate walk-through metal detectors and X-ray scanners in each of 111 public schools.”); id. at 105 (“Metal detector programs are expensive, and funds spent on metal detectors would not be available for other programs and strategies that have been shown to be effective at reducing youth risk for violence and promoting pro-social behaviors.”); Dakarai I. Aarons, Mayor Herenton Vows to Add More Officers, Detectors to City’s Public Middle and High Schools, Com. Appeal (Feb. 13, 2008, 10:48 AM), http://www.commercialappeal.com/news/2008/feb/13/board-member-says-memphis-city-schools-should-be-s/ (explaining that a “feasibility study conducted by school administrators said it would cost the 115,000-student school district $4.5 million in equipment and staffing to conduct metal-detector screenings daily in middle and high schools”).

B. Strict Security Measures Applied Disproportionately to Minority Students Are Particularly Harmful

That schools serving primarily minority students are more inclined to rely on strict security measures is particularly harmful for several reasons. First, social scientists and criminologists maintain that the presence of strict security in schools with high minority populations perpetuates racial inequalities. 137See Aaron Kupchik & Geoff K. Ward, Reproducing Social Inequality Through School Security: Effects of Race and Class on School Security Measures 3–10 (unpublished manuscript), available at http://www.edweek.org/media/kupchikward-02security.pdf. Loïc Wacquant argues that inner-city schools located in impoverished communities have “deteriorated to the point where they operate in the manner of institutions of confinement whose primary mission is not to educate but to ensure ‘custody and control.’” 138Loïc Wacquant, Deadly Symbiosis: When Ghetto and Prison Meet and Mesh, 3 Punishment & Soc’y 95, 108 (2001). He maintains:

[T]he carceral atmosphere of schools and the constant presence of armed guards in uniform in the lobbies, corridors, cafeteria, and playground of their establishment habituates the children of the hyperghetto to the demeanor, tactics, and interactive style of the correctional officers many of them are bound to encounter shortly after their school days are over. 139Id.

Lamentably, schools that focus primarily on custody and control of students deprive their students of quality educational experiences that affluent white students typically enjoy. 140See Kupchik & Ward, supra note 137, at 7 (“Marginalized youth are presumed to be young criminals and treated as such through exposure to criminal justice oriented practices (e.g., police surveillance and metal detectors), while youth with social, political and cultural capital are presumed to be well-behaved, treated as such, and empowered to be productive citizens.”). Such inferior educational experiences may inhibit minority students’ future educational and employment opportunities, further exacerbating social inequalities. 141See Hirschfield, supra note 110, at 40 (arguing that the disproportionate use of strict security measures prepares urban minority students for certain positions in the post-industrial order, such as those of prisoners, soldiers, or service sector workers).

Second, strict security measures in schools skew minority students’ perceptions of government institutions and condition them to accept intense surveillance by government authorities. 142Kupchik & Ward, supra note 137, at 6–7. Henry Leonardatos, an experienced school administrator, observes that subjecting students to pat-downs, metal detectors, and other strict security measures “ends up teaching students to not think for themselves.” 143Khan, supra note 92. He continues:

They play the role that is expected of them—they will play the role of the criminal and victimizer because the cops will say, ‘don’t do this and don’t [do] that.’ When you do that to a kid you’re telling the kid that this is how the world is supposed to be. You end up putting the idea in the kid’s head that this is what he’s supposed to be doing. 144Id. (internal quotation marks omitted).

According to Leonardatos, students in poor, minority communities are routinely stopped and frisked in their neighborhoods, then find they are treated the same way in their schools. 145Id.; accord Victor M. Rios, Punished: Policing the Lives of Black and Latino Boys, at xiv, 133–38 (2011) (finding that low-income youth and young men of color disproportionately experience surveillance, and that there is a “system in which schools, police, probation officers, families, community centers, the media, businesses, and other institutions systematically treat young people’s everyday behaviors as criminal activity”). The result is that these students perceive their school as simply “another appendage to the police state.” 146Khan, supra note 92.

Third, the trust needed to establish a healthy climate conducive to learning is undermined when minority students are cognizant that strict security measures are applied disproportionately. 147See Glennon, supra note 10, at 112 (arguing that schools cannot promote a positive school climate for learning when students are aware of the disparate application of discipline along racial lines). For example, when Minerva Dickson, a student who was subject to intense conditions at school, 148See Khan¸ supra note 92. discovered during a conference that other students were not subject to these conditions, it “blew her mind.” 149Id. She said, “I thought all schools were like mine . . . . I couldn’t believe a student could just walk into their school without dealing with all that.” 150Id. (internal quotation marks omitted). Paul Hirschfield’s interview of a former inner-city student also illustrates the sense of unfairness that students perceive at schools that rely on strict security measures. This student explained:That school was run more like a prison than a high school. It don’t have to be nothing illegal about it. But you’re getting arrested. No regard for if a college going to accept you with this record. No regard for none of that, because you’re not expected to leave this school and go to college. You’re not expected to do anything.Paul J. Hirschfield, Preparing for Prison?: The Criminalization of School Discipline in the USA, 12 Theoretical Criminology 79, 79 (2008). Indeed, the disproportionate use of strict security measures on minorities may exacerbate the distrust that many minority students already have toward educators. 151See Constance Flanagan et al., School and Community Climates and Civic Commitments: Patterns for Ethnic Minority and Majority Students, 99 J. Educ. Psychol. 421, 423 (2007) (stating that studies have shown that minority groups have reported “a lower sense of school belonging than . . . their European American peers”); Rosa Hernández Sheets, Urban Classroom Conflict: Student–Teacher Perception: Ethnic Integrity, Solidarity, and Resistance, 28 Urb. Rev. 165, 175–76 (1996) (reporting in a study on classroom conflict that minority students believed that teachers did not respect them or care about them, that teachers abused their authority, that they were purposely pushed to the edge and were expected and encouraged to be hostile, and that teachers expected and wanted them to act out to justify disciplinary actions); Susan Rakosi Rosenbloom & Niobe Way, Experiences of Discrimination Among African American, Asian American, and Latino Adolescents in an Urban High School, 35 Youth & Soc’y 420, 434 (2004) (reporting that “[w]hen African American and Latino students were asked about their experiences with discrimination, they described hostile relationships with adults in positions of authority such as . . . teachers in school”); cf. Noguera, supra note 122, at 201 (describing the sentiment in many black communities that black children are being treated unfairly in schools).

Finally, disproportionate exposure to strict security measures perpetuates inequalities by affecting minority students’ opportunities for advancement. Strict security measures, especially when used in conjunction with zero-tolerance policies, increase the likelihood of suspension, expulsion, and arrest—all of which can have profound consequences on students’ future educational and employment opportunities. 152Khan¸ supra note 92. Many are concerned that post-Newtown legislation that could provide more funds for school security measures will exacerbate the school-to-prison pipeline. See, e.g., Jamelle Bouie, How Obama Might Make the School-to-Prison Pipeline Worse, Am. Prospect (Jan. 18, 2013), http://prospect.org/article/how-obama-might-make-school-prison-pipeline-worse; Julianne Hing, After Newtown, School Communities Brace for Another Influx of Police, Colorlines (Jan. 10, 2013), http://colorlines.com/archives/2013/01/militarized_post-newtown-schools.html; Simon McCormack, Gun Control Legislation Must Not Include More Cops in Schools: ACLU Letter to Biden, Huffington Post (Jan. 15, 2013), http://www.huffingtonpost.com/2013/01/14/gun-control-legislation_n_2474188.html.

III. Empirical Evidence Reveals Disparities in the Use of Strict Security Measures

The use of strict security measures in schools certainly is not new. After other highly publicized shootings in schools, there were heavy demands to tighten school security and substantial government funding provided to schools for strict security measures. 153See, e.g., David Firestone, After Shootings, Nation’s Schools Add to Security, N.Y. Times, Aug. 13, 1999, at A1, available at http://www.nytimes.com/1999/08/13/us/after-shootings-nation-s-schools-add-to-security.html?pagewanted=all&src=pm; John D. Sutter, Columbine Massacre Changed School Security, CNN.com (Apr. 20, 2009, 9:30 AM), http://articles.cnn.com/2009-04-20/living/columbine.school.safety_1_metal-detectors-and-security-national-school-safety-school-security?_s=PM:LIVING. Yet, what types of schools implemented those measures and to what degree is not clear. Further, while a few empirical studies measure the general use of strict security measures in schools, 154See, e.g., Julie Kiernan Coon, Security Technology in U.S. Public Schools (2007); Kupchik & Ward, supra note 137. there are no national, empirical studies that evaluate the types of schools that apply combinations of strict security measures creating an intense environment that may not be conducive to a healthy learning climate. This study seeks to fill that gap.

I empirically tested the hypothesis that low-income students and minority students are subjected to intense surveillance methods more often than other students, even after taking into account factors such as school crime, neighborhood crime, school disorder, school location, and school size. To test this hypothesis, I analyzed restricted data from the U.S. Department of Education’s 2009–2010 School Survey on Crime and Safety (SSOCS). The SSOCS is a rich, national dataset that contains information submitted by school principals regarding school crime, security practices, disorder, demographics, and other related information. 155See Nat’l Ctr. for Educ. Stats., School Survey on Crime and Safety Principal Questionnaire: 2009–10 School Year 5 (2010) [hereinafter 2009–2010 SSOCS Questionnaire], available at http://nces.ed.gov/surveys/ssocs/pdf/SSOCS_2010_Questionnaire.pdf (last visited Aug. 12, 2013).

My study builds on Aaron Kupchik and Geoff Ward’s work based on SSOCS data from the 2005–2006 school year. 156See Kupchik & Ward, supra note 137, at 14. In that study, Kupchik and Ward found that, after controlling for several other factors that might influence the use of school security measures, student race was a strong predictor for the presence of metal detectors. 157Id. at 2. Kupchik and Ward also found that student poverty was a predictor for having law enforcement officers in middle and elementary schools, locked gates in elementary schools, and metal detectors in middle schools. 158Id. In addition, they found that surveillance cameras were ubiquitous across all social strata at all schools levels, and that race was not a predictor for using law enforcement officers or locked gates. 159Id. In fact, the percentage of racial/ethnic minority students was actually a negative predictor for use of drug-sniffing dogs in high schools. Id. at 42.

This study builds on Kupchik and Ward’s work by using many of the same independent variables that they used to control for other factors that might influence the use of strict security measures. However, this study differs from Kupchik and Ward’s study in two critical respects. First, this study used the SSOCS data from the 2009–2010 school year, the most recent SSOCS data available, whereas Kupchik and Ward’s study examined data from the 2005–2006 school year. 160Id. at 14. Second, and of particular importance, this study sought to uncover the characteristics of schools that employ a combination of strict security measures that could create an intense environment that may not be conducive to a healthy learning environment. Specifically, it sought to identify whether race and poverty were strong predictors for schools using a combination of metal detectors, law enforcement officers, surveillance cameras, locked gates, and random sweeps for contraband. In contrast, Kupchik and Ward’s study examined schools using only one of these practices. 161See id. at 41–42.

First, I describe the 2009–2010 SSOCS dataset I used for this empirical analysis. Next, I describe the variables I employed. Finally, I report the results of the study.

A. Data and Sample

Data for this study came from the School Survey on Crime and Safety for the 2009–2010 school year, published by the U.S. Department of Education’s National Center for Education Statistics (NCES). 162See 2009–2010 SSOCS Questionnaire, supra note 155. The dataset is the restricted-access version, meaning it contains detailed, sensitive data, such as information pertaining to the number of violent incidents that occurred at a school during the school year. 163See Statistical Standard Program: Getting Started, Nat’l Ctr. for Educ. Stats., http://nces.ed.gov/statprog/instruct_gettingstarted.asp (last visited Aug. 12, 2013). The restricted-use data “have a higher level of detail in the data compared to public-use data files.” Id. Although the restricted-use datasets are not available to the general public, datasets that contain less sensitive data for prior school years are currently available. Id. Those datasets can be downloaded at http://nces.ed.gov/surveys/ssocs/data_products.asp. The restricted dataset became available in June 2011 to researchers who met specified conditions. 164NCES provides restricted-use datasets to certain researchers in qualified organizations. Id. To qualify, an organization must provide a justification for access to the restricted-use data, submit the required legal documents, agree to keep the data safe from unauthorized disclosures at all times, and to participate fully in unannounced, unscheduled inspections of the researcher’s office to ensure compliance with the terms of the License and the Security Plan form.Id.; see also Statistical Standards Program: Applying for a Restricted-Use Data License, Nat’l Ctr. for Educ. Stats., http://nces.ed.gov/statprog/instruct_apply.asp?type=rl (last visited Aug. 12, 2013) (providing guidelines for applying to receive restricted-use datasets). It is the most recent data available on school safety from NCES.

NCES used the 2007–2008 school year Common Core of Data Public Elementary/Secondary School Universe File (CCD), 165The Common Core of Data “is an NCES annual census system that collects fiscal and nonfiscal data on all public schools, public school districts, and state education agencies in the United States.” Nat’l Ctr. for Educ. Stats., 2009–2010 School Survey on Crime and Safety (SSOCS): Restricted-Use Data File User Manual 8 (2011) [hereinafter 2009–2010 Restricted-Use Manual] (on file with author); see also Helen M. Marks & Jason P. Nance, Contexts of Accountability Under Systemic Reform: Implications for Principal Influence on Instruction and Supervision, 43 Educ. Admin. Q. 3, 10–11 (2007) (describing the Common Core of Data). The CCD includes regular schools, charter schools, and schools that have magnet programs in the United States. It excludes schools in the U.S. territories and outlying areas, such as American Samoa, Guam, the Commonwealth of the Northern Mariana Islands, the U.S. Virgin Islands, and Puerto Rico, as well as overseas Department of Defense schools, newly closed schools, home schools, Bureau of Indian Education schools, nonregular schools, ungraded schools, and schools with a high grade of kindergarten or lower. 2009–2010 Restricted-Use Manual, supra at 8. the most complete list of public schools available, as a sampling frame 166A “sampling frame” is a list of units that could be selected for study. See Richard L. Scheaffer et al., Elementary Survey Sampling 43 (5th ed. 1996). to select schools to participate in the study. 167See 2009–2010 Restricted-Use Manual, supra note 165, at 8. After the sample frame was subdivided to ensure that subgroups of interest would be adequately represented, 168See id. at 9. The sample was stratified by instructional level (e.g., elementary school, middle school, high school), locale (e.g., rural, suburb, urban), enrollment size, and region (e.g., Northeast, Midwest, South, and West). In addition, the sample frame was stratified by percent of combined student population as black/African-American, Hispanic/Latino, Asian, native Hawaiian/other Pacific Islander, or American Indian/Alaska native. Id. 3,480 schools were randomly selected to participate in the study. 169Id. at 10. A total of 2,650 public schools (approximately seventy-six percent of the selected sample) submitted usable questionnaires. 170Id. at 1, 9–13. A response rate of seventy-six percent is very good and reduces bias in the data. Earl Babbie, The Practice of Social Research 256 (9th ed. 2001). NCES notes that some schools were more likely than others to respond to the survey. For example, schools more likely to respond included rural schools, schools with fewer students, combined schools, or those with a low percentage of combined black/African-American, Hispanic/Latino, Asian, native Hawaiian/other Pacific Islander, and American Indian/Alaska native students. 2009–2010 Restricted-Use Manual, supra note 165, at 9–10. While no category had a response rate lower than sixty-nine percent, see id. at 13, using a sample weight to analyze the data helped ameliorate the effects of the discrepancies in the response rates. See infra note 215. NCES collected the data from February 24 to June 11, 2010. 1712009–2010 Restricted-Use Manual, supra note 165, at 1.

B. Dependent Variables

The 2009–2010 SSOCS restricted-use dataset provides a unique opportunity to view, as a cross section and on a national scale, the types of strict security measures that schools employ. In the 2009–2010 SSOCS, school principals were asked to respond to several questions relating to school security. 1722009–2010 SSOCS Questionnaire, supra note 155, at 5, 8. For example, principals were asked if, during the 2009–2010 school year, it was a practice in the principal’s school to

  • require students to pass through metal detectors each day;

  • perform one or more random metal detector checks on students;

  • perform one or more random sweeps for contraband (e.g., drugs or weapons), but not including dog sniffs;

  • control access to school grounds during school hours (e.g., locked or monitored gates);

  • use one or more security cameras to monitor the school; and

  • have any security guards, security personnel, or sworn law enforcement officers present at [the principal’s school] at least once a week. 173Id.

Principals responded with a “yes” or “no” to each one of these questions. 174Id.

The dependent variables for my study represented the likelihood that a school principal responded affirmatively to using various combinations of strict security practices. For example, the dependent variables measured the likelihood that a school used metal detectors on students, either by requiring students to pass through metal detectors every day, or by performing one or more random metal detector checks on students (metal detectors); had a security or law enforcement officer on campus (guards); conducted random sweeps for contraband (random sweeps); used security cameras (security cameras); and controlled access to school grounds by locking or monitoring the gates (locked gates).

I examined four different combinations of the strict security practices described above. I began with a core condition that can create an intense environment that may not be conducive to a healthy learning climate, which includes using metal detectors and having a guard or law enforcement officer present at school. 175See supra Part II. I subsequently added additional measures that intensify the surveillance environment. The resulting four combinations are as follows: (a) metal detectors and guards; (b) metal detectors, guards, and random sweeps; (c) metal detectors, guards, random sweeps, and security cameras; and (d) metal detectors, guards, random sweeps, security cameras, and locked gates.

There is one significant limitation of using the SSOCS data to test this study’s hypothesis. In two of the questions, the SSOCS asked principals whether it was a practice of their school to perform random metal detector checks or random sweeps for contraband “one or more” times during the school year. 1762009–2010 SSOCS Questionnaire, supra note 155, at 5. Thus, it is possible that when a principal responded affirmatively to one (or both) of these questions, the school had indeed performed only one random metal detector check or one random sweep for contraband during the school year. And, if a school performed only one random metal detector check or only one random sweep for contraband during the school year, perhaps that action had only a minimal effect on the learning environment. On the other hand, if a school indicated that it was using several of these criminal-justice-oriented security measures in conjunction, which is what this study measured, then it would seem more likely that the school used these security measures more than once during the school year, and quite possibly used them regularly. Further, if a school made a significant investment in metal detectors, it seems unlikely that it would use them only one time during the school year. 177Nevertheless, more precise questions that target how often schools perform these searches on students should be included in future distributions of the SSOCS. Nance, supra note 21, at 429 (recommending that the U.S. Department of Education reformulate many of the questions posed in the SSOCS).

C. Independent Variables

To measure the effect of student race, I used the percentage of the schools’ student population that consisted of African-Americans, Hispanics, Asian/Pacific Islanders, and American Indian/Alaska native students (% minority). 178 See Robers et al., supra note 90, at 112. Racial data for the 2009–2010 SSOCS came from the 2007–2008 CCD school data file. See 2009–2010 Restricted-Use Manual, supra note 165, at 29. Although there was a two-year difference, it is highly unlikely that a school would experience a major shift in student population over a two-year period. A major racial shift in the student population for a school over a two-year period would require an extraordinary event such as a desegregation court order. To measure student poverty levels, I used the percentage of students who were eligible for free or reduced-price lunch. Free and reduced-price lunch is a common proxy for student poverty (% poverty). 179 See, e.g., Federal Education Budget Project, New Am. Found., http://febp.newamerica.net/background-analysis/federal-school-nutrition-programs (last updated July 1, 2013) (“Researchers often use free or reduced price lunch (FRPL) enrollment figures as a proxy for poverty at the school level, because Census poverty data (which is used at the state and district level) is not available disaggregated below the school district level and is not collected annually.”); Michael Heise, Litigated Learning, Law’s Limits, and Urban School Reform Challenges, 85 N.C. L. Rev. 1419, 1441 (2007) (using student eligibility for free and reduced-price lunch as a proxy for student poverty). In addition to race and poverty, I included other student demographics that are consistent with student marginalization, such as the percentage of students enrolled in special education (% special ed.) and the percentage of students who have limited English proficiency (% LEP). 180See Kupchik & Ward, supra note 137, at 15. I also included the percentage of students who scored below the fifteenth percentile on standardized tests (% low test score) because, as explained above, it is possible that the accountability movement has motivated some schools to adopt strict security measures. 181See supra Part I.D. Each of these percentages was reported by school principals. 182See 2009–2010 SSOCS Questionnaire, supra note 155, at 16.

School crime is another factor that might influence whether school officials decide to implement strict security measures. 183See Kelly Welch & Allison Ann Payne, Racial Threat and Punitive School Discipline, 57 Soc. Probs. 25, 27 (2010) (“One factor presumed to be closely associated with school punitiveness and disciplinary practice is the level of school crime and disorder.” (citation omitted)). The SSOCS asks school officials to report the number of incidents of various types of school crime during the school year. 1842009–2010 SSOCS Questionnaire, supra note 155, at 11. Because the severity of the school crimes may influence whether schools implement tighter security measures, 185Kupchik & Ward, supra note 137, at 17–18. I categorized the crimes according to their degree of severity. I included violent incidents such as rape or attempted rape; sexual battery other than rape; robbery with or without a weapon; physical attacks with or without a weapon (violence); 1862009–2010 SSOCS Questionnaire, supra note 155, at 11. threats of physical attack with or without a weapon (threats); 187Id. incidents involving possession of a firearm, explosive device, knife, or other sharp object (weapons); 188Id. incidents of distribution, possession or use of illegal drugs, inappropriate prescription drugs, or alcohol (drugs); 189Id. incidents of theft of items over $10 (theft); 190Id. and incidents of vandalism (vandalism). 191Id.

In addition, school disorder may influence a school’s decision to implement tighter security measures. 192See Kupchik & Ward, supra note 137, at 17; Welch & Payne, supra note 183, at 27. To account for school disorder, I created an index based on responses to several questions posed to school officials in the SSOCS. School officials were asked to rate on a scale of one to five the frequency of occurrences relating to student racial tensions, student bullying, student sexual harassment of other students, student harassment of other students based on sexual orientation or gender identity, disorder in the classroom, student verbal abuse of teachers, student acts of disrespect for teachers other than verbal abuse, student gang activities, and student cult or extremist group activities. 1932009–2010 SSOCS Questionnaire, supra note 155, at 13. Principals indicated whether these types of problems occurred (1) daily, (2) at least once a week, (3) at least once a month, (4) on occasion, or (5) never happens. Id. After I recoded the scale so that higher values indicated greater frequency, I computed the mean value of the principals’ responses to these questions (school disorder). 194See Kupchik & Ward, supra note 137, at 17–18.

I also controlled for the principals’ perception of crime problems near the school. 195Id. at 18 (controlling for the administrators’ perceptions of the level of crime in the neighborhoods around the schools). School officials were asked to describe the crime level in the area where their schools were located on a scale of one to three (high, moderate, or low). 196“Controlling for” means measuring the effect of one variable while taking into account the effect of another variable. See Will G. Hopkins, A New View of Statistics, Sportscience, http://www.sportsci.org/resource/stats/complex.html (last updated June 22, 2002). For example, researchers who wish to study the effect of gender on weight would want to “control for” height because of the known effect height has on weight. See id. I recoded those values so that higher values reflected higher perceptions of crime (neighborhood crime). 197See Kupchik & Ward, supra note 137, at 18.

The involvement of external community groups in schools may also affect the decision to implement strict security measures. 198See id. at 16. Principals were asked if various community and outside groups were involved in their efforts to promote safe, disciplined, and drug-free schools. 1992009–2010 SSOCS Questionnaire, supra note 155, at 7. Those groups included parent groups (parents), social services agencies (social services), juvenile justice agencies (juvenile justice), law enforcement agencies (law enforcement), mental health agencies (mental health), civic organizations/service clubs (civic orgs.), private corporations/businesses (business orgs.), and religious organizations (religious orgs.). 200Id. I included a dummy variable for each of these groups to indicate their participation. 201Creating dummy variables is a way to include categorical predictor variables in estimation models such as logistic regression. See Inst. for Digital Research & Educ., FAQ: What Is Dummy Coding?, UCLA.edu, http://www.ats.ucla.edu/stat/mult_pkg/faq/general/dummy.htm (last visited Aug. 12, 2013) (describing dummy coding).

I also took into account the geographic region of the schools. Prior research suggests that schools across every geographic region have intensified their surveillance and disciplinary measures, but the South has historically maintained a stronger disciplinary approach in schools. 202See Kupchik & Ward, supra note 137, at 1617. The SSOCS database did not contain information regarding the individual states in which each participating school is located. Rather, it only contained information regarding in which geographical region each school resides (Northeast, South, West, Midwest). 203See 2009–2010 Restricted-Use Manual, supra note 165, at 25. I included dummy variables for the Northeast, West, and Midwest to compare schools from these regions to schools in the South.

School urbanicity also may affect whether principals decide to employ strict security measures. 204See Welch & Payne, supra note 183, at 28 (“School urbanicity is one feature associated with increased student punishment . . . .”). The SSOCS database categorized schools by whether the school was located in a city, suburb, town, or rural area. 2052009–2010 Restricted-Use Manual, supra note 165, at 28–29. I created dummy variables for schools located in suburban areas (suburban), towns (town), and rural areas (rural) to compare to schools located in cities. Additionally, student population (student pop.) may influence the use of strict security measures. Schools with larger student populations might be more likely to implement security measures as a quick, efficient means to maintain safety and discipline. I included each school’s total enrollment as reported by school principals. 2062009–2010 SSOCS Questionnaire, supra note 155, at 16.

I also included other control variables such as the building level and whether the school was nontraditional. Regarding building level, as I only examined secondary schools, I included dummy variables for middle schools (middle) and combined schools (combined) to compare to high schools. 207Middle schools are “schools in which the lowest grade is not lower than grade 4 and the highest grade is not higher than grade 9.” Nat’l Ctr. for Educ. Statistics, Crime, Violence, Discipline, and Safety in U.S. Public Schools: Findings from the School Survey on Crime and Safety: 2009–10, at 7 tbl.1 n.5 (2011), available at http://nces.ed.gov/pubs2011/2011320.pdf. High schools are “schools in which the lowest grade is not lower than grade 9 and the highest grade is not higher than grade 12.” Id. Combined schools include “other combinations of grades, including K–12 schools.” Id. Nontraditional schools (nontraditional) included whether the school was a charter school or magnet school and was dummy-coded as a comparison against traditional schools. 2082009–2010 SSOCS Questionnaire, supra note 155, at 17. Finally, I included the school’s average percentage of students attending school daily as reported by the school principals (student attendance), 209Id. which also has been linked to increases in school violence. 210Kevin P. Brady, Sharon Balmer & Deinya Phenix, School–Police Partnership Effectiveness in Urban Schools: An Analysis of New York City’s Impact Schools Initiative, 39 Educ. & Urb. Soc’y 455, 456 (2007).

Because several of the continuous independent variables were positively skewed, I transformed each of the continuous variables by including its natural log. 211See Kupchik & Ward, supra note 137, at 18–19 (transforming continuous variables by including their natural logs to correct for positive skewness). I did this for all the independent variables except the dummy variables (geographic region, urbanicity, building level, nontraditional school, and involvement of various community organizations), the ordinal variable of neighborhood crime, and the school disorder index. 212Id.

D. Empirical Methodology

Each of the four dependent variables represents whether a school employed various combinations of criminal-justice-oriented security measures. If a school employed all of the methods in a certain combination, it was given a value of “1.” If it did not employ all of the methods in a certain combination, it was given a value of “0.” Because each of the dependent variables was dichotomous, I used binary logistic regression to estimate the odds that a school employed a specific combination of strict security measures. 213See generally Joseph F. Hair, Jr. et al., Multivariate Data Analysis 276–81 (5th ed. 1998) (providing an overview of logistic regression analysis). I conducted all empirical analyses using a computer program called SPSS. 214For more information on SPSS, see Inst. for Digital Research & Educ., SPSS Starter Kit, UCLA.edu, http://www.ats.ucla.edu/STAT/spss/sk/default.htm (last visited July 25, 2013). In addition, to produce optimal national estimates, I used the weighted variable provided in the SSOCS database. 215The weighted variable compensates for unequal probabilities of selection, minimizes bias associated with responding and nonresponding schools, reduces sampling error, and calibrates the data to known population characteristics to produce optimal national estimates. See 2009–2010 Restricted-Use Manual, supra note 165, at 13 (describing the specific weighting procedures employed); see also Ibrahim S. Yansaneh, Construction and Use of Sample Weights ch. 5.3 (U.N. Group of Experts Meeting to Review the Draft Handbook on Designing of Household Sample Surveys, Draft U.N. Doc. ESA/STAT/AC.91/5, 2003), available at http://unstats.un.org/unsd/demographic/meetings/egm/Sampling_1203/docs/no_5.pdf. Furthermore, I adjusted the sample weight created by NCES by dividing it by its mean to create a mean weight of one. This is a recommended procedure when employing logistic regression analysis using SPSS. See Marks & Nance, supra note 165, at 14; Patty Glynn, Adjusting or Normalizing Weights “On the Fly” in SPSS, U. of Wash., http://staff.washington.edu/ glynn/adjspss.pdf (last updated July 8, 2004). The sample for this study includes approximately 910 middle schools, 950 high schools, and 110 combined schools, for an approximate total of 1,970 schools. 216Pursuant to the guidelines for presenting results from the restricted-use SSOCS database, raw sample numbers have been rounded to the nearest ten. U.S. Dep’t. of Educ., Restricted-Use Data Procedures Manual 20 (2011), available at http://nces.ed.gov/pubs96/96860rev.pdf.

E. Results

Figure 1 shows the total number of schools in the sample that employed a certain combination of security measures. Figure 2 shows the estimate of the total number of schools nationally that employed a certain combination of security measures. As the data indicate, while only a relatively small percentage of schools applied combinations of strict security measures, the estimated number of schools applying combinations of strict security measures across the country is quite large, potentially affecting hundreds of thousands or even millions of students’ everyday lives. 217For example, the average high school in the United States has between 700 and 800 students. See Overview of Public Elementary and Secondary Schools and Districts: School Year 1999–2000, Nat’l Ctr. for Educ. Stats. (Sept. 2001), http://nces.ed.gov/pubs2001/overview/table05.asp (last visited Aug. 12, 2013). Multiplying 700 students by 3,344 schools produces a rough estimate of 2,340,800 students.

Figure 1

Figure 1. Number of Schools in Sample Using Combinations of Strict Security Measures

Notes: Combination 1 includes metal detectors and guards.

Combination 2 includes metal detectors, guards, and random sweeps.

Combination 3 includes metal detectors, guards, random sweeps, and security cameras.

Combination 4 includes metal detectors, guards, random sweeps, security cameras, and locked gates.

Figure 2

Figure 2. Estimate of the Number of Schools Nationally That Use Combinations of Strict Security Measures

Notes: Combination 1 includes metal detectors and guards.

Combination 2 includes metal detectors, guards, and random sweeps.

Combination 3 includes metal detectors, guards, random sweeps, and security cameras.

Combination 4 includes metal detectors, guards, random sweeps, security cameras, and locked gates.

Population estimates calculated from the sample weights provided by NCES. See 2009–2010 Restricted-Use Manual, supra note 165, at 13 (describing appropriate uses for sample weights provided by NCES).

Table 1 contains the results of the logistic regression analysis. It displays the exponentiated coefficients (Exp(B)) for the independent variables in each of the models. 218The variance inflation factors (VIF), a common statistic to detect multicollinearity, indicated that multicollinearity was not a problem for the models. In addition, the Hosmer & Lemeshow test, a statistical test to assess a lack of fit between the data and the model, indicated a goodness of fit for each of the models. An exponentiated coefficient estimates the change in odds of a school using a combination of certain security practices for each one-unit increase in an independent variable. 219See Raymond E. Wright, Logistic Regression, in Reading and Understanding Multivariate Statistics 217, 223 (Laurence G. Grimm & Paul R. Yarnold eds., 1995) (“The odds ratio estimates the change in the odds of membership in the target group for a one-unit increase in the predictor.”). For example, hypothetically, if the odds ratio for the independent variable “% minority” were two, then the odds of a school employing a certain combination of strict security measures would be twice greater for schools having a minority population of fifty percent than for schools having a minority population of forty-nine percent. Table 1 also displays whether the effects of the independent variables are statistically significant. 220Statistical significance measures the likelihood of whether the regression coefficient, or the effect of the independent variable, is different from zero in the population of schools from which the sample was drawn. Id. at 226–27.

Table 1. Logistic Regression Model Predicting Odds of School Using Combination of Security Practices (Exp(B) Reported)

Comb. 1

Comb. 2

Comb. 3

Comb. 4

School Crime

Violence (ln)

.837

.925

.944

.908

Weapons (ln)

.995

.821

.938

1.048

Thefts (ln)

.818*

.850

.805

.835

Drugs (ln)

1.141

1.137

1.153

.988

Vandalism (ln)

1.034

1.071

1.112

1.152

Threats (ln)

.978

.916

.883

.825

School Disorder

1.141

1.236

1.337

1.366

Neighborhood Crime

1.036

.842

.764

.572**

Cmty. Involvement

Parents

1.168

1.708*

1.574

1.594

Social Services

1.115

1.221

1.169

1.548

Juvenile Justice

1.878

1.074

1.292

1.220

Law Enforcement

1.118

1.092

.907

.980

Mental Health

1.340

1.588

1.322

1.310

Civic Orgs.

.819

.734

.831

.539*

Business Orgs.

1.108

1.293

1.274

1.325

Rel. Orgs.

1.158

1.124

1.148

1.484

Geographic Regiona

Northeast

1.382

.730

.789

.516

Midwest

.872

.540

.538

.417*

West

.171***

.203***

.119

.135***

Urbanicityb

Suburban

.394***

.376**

.375

.358**

Town

.511*

.519

.516

.305*

Rural

.411**

.589

.624

.312**

Building Levelc

Middle

.865

1.200

.963

1.176

Combined

.705

1.236

.668

.939

Nontraditional

1.356

1.544

1.263

1.174

Stud. Attendance (ln)

.667

1.022

.881

.623

Special Ed. (%) (ln)

1.106

1.014

.999

1.076

LEP (%) (ln)

.660***

.699***

.711***

.637***

Low Test Score (%) (ln)

1.231

1.436**

1.313*

1.222

Student Pop. (ln)

2.771***

3.133***

2.423***

2.618***

Minority (%) (ln)

1.877***

2.853***

2.603***

3.359***

Poverty (%) (ln)

7.734***

2.236***

2.311**

3.412**

*p < 0.05, **p < 0.01, ***p < 0.001

Notes: Combination 1 includes metal detectors and guards.

Combination 2 includes metal detectors, guards, and random sweeps.

Combination 3 includes metal detectors, guards, random sweeps, and security cameras.

Combination 4 includes metal detectors, guards, random sweeps, security cameras, and locked gates.

a. Southern states were the comparison group.

b. Urban schools were the comparison group.

c. High schools were the comparison group.

All four models demonstrate that a school’s percentage of minority students is a strong predictor of whether the school uses a combination of strict security measures. Stated another way, as the school’s percentage of minority students increases, the odds of using combinations of security measures also increases. This is true even after controlling for other important factors that may influence the decision to employ strict security measures, such as school crime, neighborhood crime, and school disorder. In addition, in all four models, student poverty is a strong predictor for using a combination of strict security measures. These findings suggest that schools serving high percentages of low-income students or minority students are more inclined to rely on heavy-handed, justice-oriented measures to control crime and maintain order than other schools that confront similar crime and discipline issues.

Another interesting finding is that, in all four models, the odds of relying on strict security measures were greater for schools with larger student populations than for schools with smaller student populations. While the precise reasons for this finding are unclear, it suggests that schools with large student populations, especially those that serve large populations of low-income or minority students, may lack the resources needed to address school crime and disorder in alternative ways. Perhaps school officials are simply overwhelmed by the large number of students and believe that they must rely on shortcut methods such as metal detectors and random sweeps to combat student crime. Indeed, educators in large schools may find it more difficult to develop meaningful relationships with students and to watch for signs of trouble. 221See Ascher, supra note 120, at 3 (“The inability of teachers and other school staff to make meaningful connections with students in large schools and increasingly large classes has become a key safety issue.”). Further, large schools may not be in a financial position to hire additional personnel to properly implement the alternative programs described below, but instead rely on money from federal and state governments to purchase strict security measures. 222See infra Part IV.B.

The analysis also reveals that geographic region is a strong predictor of whether a school uses a combination of strict security measures. In three of the models, the odds for using strict security measures were greater for schools in the South than for schools in the West. In one of the models, the odds were greater for schools in the South than for schools in the Midwest. This provides support for the claim that schools in the South continue to maintain a strong disciplinary approach in schools. 223See Kupchik & Ward, supra note 137, at 1617.

In addition, in three of the four models, the odds of using a combination of security measures were greater for urban schools than suburban schools. Furthermore, in two of the models, the odds of using a combination of measures were greater for urban schools than for schools located in suburban, town, and rural areas. These findings confirm what has been observed anecdotally—that large, urban schools that serve low-income students or minority students are more likely to create intense surveillance environments for students than other schools. 224See, e.g., Brady, Balmer, & Phenix, supra note 210, at 45657 (“An increasing fear of school violence coupled with the public’s misperceptions of [school safety] . . . has caused school officials, especially those located in urban areas, to implement more punitive-based . . . policies . . . .”); Noguera, supra note 122, at 206 (noting that most urban high schools “rely upon coercion or excessive forms of control” to ensure school safety). Again, it is important to emphasize that these findings hold true even when controlling for other important conditions including school crime, disorder, and neighborhood crime.

The percentage of students who have limited English proficiency also was a strong predictor in all of the models, but it was negatively related to the use of security measures. That is, as the percentage of students who have limited English proficiency increased, the odds of using combinations of measures decreased. A possible explanation for this relationship is that school officials perceive students who have limited English proficiency to be less threatening.

In general, other school characteristics, such as average student attendance, being a charter or magnet school, the percentage of special education students, and building level were not strong predictors. But there was one interesting exception. In two of the models, the percentage of students receiving low test scores was a significant predictor. However, because causal relationships and temporal order cannot be detected from this dataset, one cannot discern whether (a) applying security measures had a negative impact on the learning environment, (b) officials applied strict security measures to push low-performing students out of school, or (c) there is no causal relationship at all, but only a positive correlational relationship.

The involvement of external community organizations in schools’ efforts to promote safe and disciplined schools generally was not a strong predictor for employing combinations of security measures. Nevertheless, two notable exceptions emerged. First, although parent involvement generally was not a strong predictor in the models, it was in one model. Relative to the combination that included metal detectors, guards, and random sweeps, schools indicating that parents were involved in their efforts to create safe schools were more likely to impose that combination. This suggests that, at least in some cases, parents wanted schools to employ such methods to keep children safe. Second, the involvement of civic organizations was a negative predictor in one model.

Finally, and interestingly, the data generally do not indicate any relationships between the number of criminal incidents reported and the use of security measures. But there was one exception. The number of thefts was negatively related to the combination that included metal detectors and guards. Nevertheless, because causality and temporal order cannot be detected from the data, the data do not allow one to conclude that the strict security measures actually caused a decrease in the number of thefts in the schools. 225See Kupchik & Ward, supra note 137, at 24–25.

IV. Discussion and Recommendations

The U.S. Supreme Court interprets the Fourth Amendment as providing students with almost no protection from strict security measures, especially from measures considered to be “minimally intrusive.” 226See supra Part I.A. Further, federal and state laws have actively encouraged the use of strict security measures by providing schools with money to implement these measures and enacting accountability laws that appear to motivate school officials to push low-performing students out of school. 227See supra Parts I.C–D. This empirical analysis reveals what has emerged nationally inside of this framework. Even when taking into account other important variables that might influence schools to adopt strict security measures such as school crime, school disorder, and neighborhood crime, schools serving low-income students and minority students are much more likely to adopt measures that create an intense school climate than schools serving other students. Thus, while school crime, neighborhood crime, and school disorder no doubt influence some schools to adopt strict security measures, this study finds that those factors do not fully explain school officials’ decision to do so. Rather, this study suggests that the racial and socioeconomic makeup of the students may also account for this decision. As new funding for school security measures now becomes available for schools after Newtown, there is a danger that these disparities may continue or even worsen. But disparity in the use of strict security measures along racial and economic lines is only one important reason to divert government funding away from strict security measures. Another important reason is that there are better ways to curb school and community violence—ways that will enhance rather than degrade the learning environment.

Part IV consists of three subsections. Section A discusses the implications of the empirical findings. Section B describes alternative measures that schools can adopt to promote school safety without deteriorating the learning environment. Finally, section C proposes other recommendations to address the disproportionate use of strict security measures on minorities and low-income students.

A. The Empirical Findings Raise Serious Concerns

The empirical findings provide support for the theory that the primary mission of schools that serve low-income and minority students may “not [be] to educate, but to ensure ‘custody and control.’” 228See Wacquant, supra note 138, at 108. The findings suggest that school officials’ mindset in these schools may be to counter violence with force; to curb school crime by creating intense environments designed to coerce students into compliance; and to create safe schools by identifying, apprehending, and excluding students that have the potential to be disruptive. 229See Hirschfield, supra note 110, at 45; Noguera, supra note 122, at 192. The findings also suggest that schools serving primarily affluent or white students implement alternative ways to curb school crime. This is not to say, however, that schools serving affluent and white students do not rely on any security measures at all. They do. However, schools working with these student populations tend to rely on different, less intrusive and more covert measures, such as surveillance cameras. 230See Robers et al., supra note 90, at 164 (disaggregating the use of security measures by racial and socioeconomic categories of students). As Paul Hirschfield observes, “[C]riminalization in middle class schools is less intense and more fluid than in the inner-city . . . . In short, the gated community may be a more apt metaphor to describe the security transformation of affluent schools, while the prison metaphor better suits that of inner-city schools.” 231Hirschfield, supra note 150, at 84. Recent data from the U.S. Department of Education also show that schools that serve affluent or white populations tend to rely more on drug-sniffing dogs than other schools. See Robers et al., supra note 90, at 165. But, as the empirical analysis presented here demonstrates, schools serving low-income minority students are much more likely to use multiple methods that create an intense surveillance environment than other schools. See supra Part III. Educational scholar Pedro Noguera sums up his extensive knowledge of the public school system as follows: “I frequently visit schools in suburban communities and private schools that serve affluent students and see quite clearly that poor children in the inner city are more likely to receive an education that places greater emphasis on order and control than academic rigor.” 232See Noguera, supra note 89, at 348.

Noguera offers a compelling explanation for why schools serving disadvantaged students tend to adopt strict security measures. He observes that schools traditionally carry out three primary functions: (1) to sort children based on their academic abilities and put them on trajectories that will influence their future economic roles and occupations, (2) to socialize children by teaching them values and norms that are central to our democratic society, and (3) to provide an important custodial function by caring for and protecting children while they are away from home. 233Id. at 344. While each function is important and challenging, schools cannot accomplish the first two functions without maintaining order and control. 234Id.; see also Levin, supra note 21, at 164849. During the sorting process, children soon learn where they stand and develop certain expectations regarding their future economic roles and occupations. 235Noguera, supra note 89, at 344. Students eventually begin to understand that some of them will assume leadership roles in private and public institutions and will prosper; others will achieve at least minimum economic security; but others will fill dead-end jobs, receive low wages, and become subordinated. 236Id. Students who begin to comprehend that the educational process is not working for them—that they most likely will not enter college or have a promising career—typically are the students who are most disruptive to the educational process. 237Id.; see also Steinberg, Allensworth & Johnson, supra note 125, at 27–31 (finding that the academic skills of the students are strongly related to school safety); Paul Willis, Learning to Labor: How Working Class Kids Get Working Class Jobs 72 (1977) (arguing that for students who believe that knowledge and credentials acquired in schools are irrelevant, “the teachers’ authority becomes increasingly the random one of the prison guard, not the necessary one of the pedagogue”). They question the educational system, push back against the mandatory attendance policies, disrupt classroom activities, and seek other, often illegitimate, means to establish their identity, self-worth, and status in their communities and among their peers. 238See Steinberg, Allensworth & Johnson, supra note 125, at 46 (maintaining that lower-achieving students are less likely to be engaged in school, more likely to feel frustrated by their performance, more likely to act out, and less likely to respond to academic punishments).

School officials generally understand that students on a dead-end educational path typically cause more trouble in schools. 239Noguera, supra note 89, at 344. However, instead of focusing their efforts on meeting students’ challenging needs, inspiring them, or helping them to discover an alternative path, schools will often focus on maintaining order and discipline. 240Id. Thus, schools with disproportionate numbers of academically unsuccessful students, many of which are in the inner city and serve low-income, minority students, often will use extreme forms of discipline and control that include implementing strict security measures. 241Id. at 345. As Noguera observes, “Such schools often operate more like prisons than schools. They are more likely to rely on guards, metal detectors, and surveillance cameras to monitor and control students, restrict access to bathrooms, and attempt to regiment behavior by adopting an assortment of rules and restrictions.” 242Id.

The empirical findings presented here provide support for Noguera’s theory. They demonstrate that large, urban schools serving historically disadvantaged student populations are more likely to create intense environments than other schools. 243See supra Part III, Table 1 and accompanying discussion. Indeed, low-income and minority students attending large, inner-city schools often face unique, difficult challenges in schools and, too often, those schools do not meet their needs. No doubt many educators teaching in these environments genuinely try to help troubled students understand the benefits of receiving an education. 244See Hirschfield, supra note 150, at 92. However, there is evidence that some educators and school administrators in impoverished areas believe that troubled students are prison-bound, 245Id. or at least there is evidence that students perceive that their teachers view them as such. 246Michelle Fine et al., Civics Lessons: The Color and Class of Betrayal, 106 Tchrs. C. Rec. 2193, 2204–05 (2004) (finding that students believed that their teachers considered them to be “animals,” “inmates,” or “killers”); Pedro A. Noguera, The Trouble with Black Boys: The Role and Influence of Environmental and Cultural Factors on the Academic Performance of African American Males, 38 Urb. Educ. 431, 448 (2003) (finding that black students were much less likely than white students to believe that their teachers support and care about them). Furthermore, although many educators may wish to reverse the downward trajectory and help troubled students succeed in school, they believe that their schools lack the resources to assist those students, 247Ascher, supra note 120, at 4 (describing the pressures that many urban educators teaching disadvantaged students face, making those educators feel frustrated, overworked, stressed, cynical, and burned out). especially without compromising the quality of education for other students. 248See Hirschfield, supra note 150, at 92. And, lamentably, while educators in these environments may lack the resources necessary to reduce class sizes, adequately train teachers, and provide counselors, mentors, mental health experts, and behavior specialists, 249See Ascher, supra note 120, at 3–4 (describing the inadequate resources of many urban school serving disadvantaged populations). the government provides resources to pay for strict security measures. 250See supra Part I.C. Thus, school authorities often adopt measures that institute order and control and willingly remove students who contribute to a disorderly environment.

These findings also provide support for the growing consensus that racial status is strongly related to punitive measures taken in schools. 251Welch & Payne, supra note 183, at 28. In a report issued in August 2012 by UCLA’s Civil Rights Project, Daniel Losen and Jonathan Gillespie analyzed data publicized by the OCR. 252Daniel J. Losen & Jonathan Gillespie, The Civil Rights Project, Opportunities Suspended: The Disparate Impact of Disciplinary Exclusion from School 6 (2012), available at http://civilrightsproject.ucla.edu/resources/projects/center-for-civil-rights-remedies/school-to-prison-folder/federal-reports/upcoming-ccrr-research/losen-gillespie-opportunity-suspended-2012.pdf. That report reveals that one out of every six black students enrolled in K–12 public schools has been suspended at least once, but only one out of twenty white students has been suspended. 253Id. Worse, one out of every four disabled black children was suspended during the 2009–2010 school year. 254Id. at 7.

Losen and Gillespie explain that because the OCR data do not contain the reasons for suspensions, their report does not provide definitive answers for why these disparities exist. 255Id. at 32–33. Nevertheless, other studies demonstrate that black students are indeed punished disproportionately relative to their offenses. 256See Ciolfi & Ryan, supra note 10, at 327–28 (citing studies demonstrating that students of color are punished disproportionality relative to their behavior). For example, using a national sample of 294 schools, scholars Kelly Welch and Allison Ann Payne found that schools with greater percentages of black students were likely to impose harsher penalties for misbehavior such as automatic suspension, expulsion, and notifying the police. 257Welch & Payne, supra note 183, at 36. They also found that schools with more black students and more low-income students were less likely to use milder forms of punishment, such as referrals to visit school counselors or oral reprimands. 258Id. at 36–37. Furthermore, they found that schools with more black students were less supportive of alternative forms of discipline, such as community service and restorative justice programs. 259Id. at 37. Other studies show similar racial disparities with respect to school discipline. 260See, e.g., Tony Fabelo et al., Breaking Schools’ Rules: A Statewide Study of How School Discipline Relates to Students’ Success and Juvenile Justice Involvement 45 (2011), available at http://justicecenter.csg.org/files/Breaking_Schools_Rules_Report_Final.pdf (showing that in a statewide Texas study that blacks were disproportionately likely to be disciplined for “discretionary reasons” and more likely to be removed from class); Catherine P. Bradshaw et al., Multilevel Exploration of Factors Contributing to the Overrepresentation of Black Students in Office Disciplinary Referrals, 102 J. Educ. Psychol. 508, 508 (2010) (finding that after controlling for teacher ratings of students’ behavior problems, black students were more likely than white students to receive office disciplinary referrals); Daniel J. Losen, The Civil Rights Project, Discipline Policies, Successful Schools, and Racial Justice 4 (2011), available at http://nepc.colorado.edu/files/NEPC-SchoolDiscipline.pdf. Thus, my study illuminates only a portion of the widespread racial inequalities that exist in our nation’s public schools. 261See supra note 8 and accompanying text. Yet, disturbingly, there is currently almost no legal recourse to address these racial inequalities. See Glennon, supra note 10, at 120–24 (describing the limitations of addressing racial inequalities in public schools under the Fourteenth Amendment and Title VI of the Civil Rights Act of 1964); Skiba, Eckes & Brown, supra note 10, at 1089–1100 (describing the limitations of addressing racial inequalities in public schools under the Fourteenth Amendment and Title VI of the Civil Rights Act of 1964). In a subsequent research project, I intend to propose an alternative framework to address the disparate use of strict security measures on minorities.

B. Alternative Measures More Effectively Reduce Violence

The disproportionate use of strict security measures on minority and low-income students is a compelling reason for policymakers not to make funds available to schools to purchase these measures. But another compelling reason is that school officials can do more to reduce violence in their schools and communities by adopting alternative measures. While strict security measures are visible, tangible responses to address school violence that might provide superficial comfort, 262See Ascher, supra note 120, at 4 (maintaining that metal detectors and other mechanical devices only provide a false sense of safety). they do not address what is truly needed: to help children become ethical, well-rounded, productive members of our society who can resolve conflicts without resorting to violence. 263Id. at 7 (describing the benefits for children of adopting alternative methods to reduce school violence). At a recent conference for community members in New Jersey, “Safe and Secure Schools: Perspectives After Newtown,” keynote speaker Maurice Elias reminded attendees that “our children cannot learn, and our teachers cannot teach, in schools that are unsafe, unsupportive, uncaring, uncivil or lacking in intellectual challenge . . . . These are the ultimate sources of security to children and in ways that are more lasting than metal detectors.” 264Joe Green, South Jersey Schools Discuss Safety Following ‘Perspectives After Newtown,NJ.com (Jan. 23, 2013, 8:00 AM), http://www.nj.com/gloucester-county/index.ssf/2013/01/south_jersey_schools_to_talk_s.html (internal quotation marks omitted). Similarly, after the Columbine shootings, the U.S. Secret Service and the Department of Education issued a joint report concluding that a fundamental component for reducing school violence is to improve the school’s climate and strengthen trust and communication among students and educators. 265See Robert A. Fein et al., U.S. Secret Serv. & U.S. Dep’t of Educ., Threat Assessment in Schools: A Guide to Managing Threatening Situations and to Creating Safe School Climates 1112 (2002), available at http://www.secretservice.gov/ntac/ssi_guide.pdf. This report was recently emphasized by the Obama administration after the Newtown shootings. See The White House, Now Is the Time 1213 (2013), available at http://www.whitehouse.gov/sites/default/files/docs/wh_now_is_the_time_full.pdf. As that report emphasized,

In educational settings that support climates of safety, adults and students respect each other. A safe school environment offers positive personal role models in its faculty. It provides a place for open discussion where diversity and differences are respected; communication between adults and students is encouraged and supported; and conflict is managed and mediated constructively. 266Fein et al., supra note 265, at 11; see also Fedders, Langberg & Story, supra note 89, at 6 (“Positive relationships among students, families, teachers, administrators, and staff are the most effective tools in creating a safe school environment.”).

Indeed, after conducting a comprehensive study of school safety in the Chicago public school system, scholars Matthew Steinberg, Elaine Allensworth, and David Johnson concluded that even in schools serving large populations of students from areas of high crime and high poverty, “it is the quality of relationships between staff and students and between staff and parents that most strongly defines safe schools.” 267Steinberg, Allensworth, & Johnson, supra note 125, at 1.

We certainly cannot prevent all acts of violence at school or elsewhere. But we can do more to help students develop social and emotional stability, improve their attitudes about themselves and others, decrease their levels of emotional stress, teach them how to resolve conflicts peacefully, and help them develop positive relationships with their teachers, other adults, and their peers. 268As scholar Jim Ryan reminds us,From the birth of the common school movement through early desegregation cases, schools were seen not simply as places where students learned how to read and write but also as places where they learned how to become better citizens. Indeed, for a long time, the socializing or civic mission of schools was considered by many to be just as important as the academic mission.James E. Ryan, The Supreme Court and Voluntary Integration, 121 Harv. L. Rev. 131, 143 (2007) (footnote omitted). Such initiatives will do more to curb violence in schools and in communities—and help students to become more well-rounded and prepared for the workforce—than strict security measures ever could. This section will describe specific, alternative methods that schools can implement to achieve these results.

First, a program called Positive Behavioral Interventions and Supports (PBIS) is a well-respected, data-driven initiative that provides strategies for defining, teaching, and supporting appropriate behavior to create strong learning environments for an entire district or school. 269See Fedders, Langberg & Story, supra note 89, at 8 (recommending PBIS as a proven measure to improve school safety); Losen & Gillespie, supra note 252, at 43; Mayer & Leone, supra note 88, at 13 (explaining that PBIS is an approach that “transforms the school environment to support overall student success, behaviorally, socially, and academically”); School-Wide PBIS, Positive Behav. Interventions & Supports, http://www.pbis.org/school/default.aspx (last visited Aug. 13, 2013) (describing school-wide PBIS). PBIS is a decision-making framework that helps educators select and implement evidence-based practices to improve academic and behavioral outcomes for all students. 270See What Is School-Wide Positive Behavioral Interventions & Supports?, Positive Behav. Interventions & Supports, http://www.pbis.org/school/what_is_swpbs.aspx (lasted visited Aug. 13, 2013). Under this program, educators develop a set of behavior interventions and supports, use data to solve problems, modify the environment to prevent problems from developing and occurring, teach and support appropriate skills and behaviors, implement the social and behavioral practices with consistency and accountability, and continually monitor progress. 271Id.

The program takes a multi-tiered approach. The first tier consists of providing a system-wide set of prevention strategies for all students, staff, and settings that include defining, teaching, and rewarding appropriate behavior. 272Id.; see also Is School-Wide Positive Behavior Support an Evidence-Based Practice?, Positive Behav. Interventions & Supports (March 2009), http://www.pbis.org/research/default.aspx. Second-tier strategies consist of intensive prevention initiatives directed at specialized groups of students who did not respond positively to the system-wide approach. 273See What Is School-Wide Positive Behavioral Interventions & Support?, supra note 270. Core elements of the second-tier strategies include screenings, progress monitoring, systems for increasing structure and predictability, contingent adult feedback, and increasing home and school communication. 274See Is School-Wide Positive Behavior Support an Evidence-Based Practice?, supra note 272. Finally, third-tier strategies are directed at students who did not respond positively to the second-tier prevention strategies. 275Id. Those strategies include developing team-based, highly individualized plans to assist students with their needs. 276Id. In sum, this program creates learning environments that are more inclusive, productive, and engaging, and of course, safer for both students and educators. 277Id.; see also Mayer & Leone, supra note 88, at 13 (explaining the benefits of implementing PBIS for students). It has been successful in all settings, including urban schools and in the juvenile justice system. 278See Frequently Asked Questions, Positive Behav. Interventions & Supports, http://www.pbis.org/school/primary_level/faqs.aspx (last visited Aug. 13, 2013); see also Domenici & Forman Jr., supra note 107, at 290 (employing successfully a modified version of PBIS in a school inside of a juvenile justice facility to encourage students to behave according to school values).

Second, the practice of restorative justice is another alternative approach to top-down, authoritarian discipline regimes that rely on strict security measures. Restorative justice practices are dispute resolution tools that involve both the victims and offenders. 279N.Y. Civ. Liberties Union, Safety with Dignity: Alternatives to the Over-Policing of Schools 8 (2009), available at http://www.nyclu.org/files/Safety_with_Dignity.pdf. These practices “focus on repairing the harm, engaging victims, establishing accountability, developing a community, and preventing future actions.” 280Thalia N. C. González & Benjamin Cairns, Moving Beyond Exclusion: Integrating Restorative Practices and Impacting School Culture in Denver Public Schools, in Justice for Kids: Keeping Kids Out of the Juvenile Justice System, supra note 10, at 241, 241. Through formal and informal conferences, victims share with offenders how they have been hurt by the offenders’ actions, and the offenders have opportunities to make apologies. 281Id.; see also Laura Mirsky, Building Safer, Saner Schools, 69 Educ. Leadership, Sept. 2011, at 45, 4549 (2011); Laura Mirsky, SaferSanerSchools: Transforming School Culture with Restorative Practices, Restorative Pracs. Eforum 1 (May 20, 2003), http://www.iirp.edu/iirpWebsites/web/uploads/article_pdfs/ssspilots.pdf. During the conferences, both the victims and offenders devise remedies for the wrong committed. 282See Mirsky, Building Safer, Saner Schools, supra note 281, at 4549; Mirsky, SaferSanerSchools, supra note 281, at 1; cf. Catherine J. Ross, A Place at the Table: Creating Presence and Voice for Teenagers in Dependency Proceedings, 6 Nev. L.J. 1362, 1365 (2006) (“We have known for decades that involving young people in decision-making promotes more positive outcomes.”). Restorative justice initiatives teach all students to share feelings in response to undesirable behavior from other students, which has the effect of humanizing victims and changing the dynamics of those involved in the incidents. 283See Mirsky, Building Safer, Saner Schools, supra note 281, at 4549.

The following example is provided to illustrate the benefits of restorative justice. After a student broke a window in the Humanities Preparatory Academy in Manhattan, New York, the student participated in a restorative justice conference. 284See N.Y. Civ. Liberties Union, supra note 279, at 18. During the session, participants discovered that the day before the incident, the student had received notice that his family was being removed from its shelter and had no place to go. 285Id. Because this did not excuse his behavior, the offender and the committee members decided that he needed to give back to the school community. 286Id. Understanding that the student could not afford to replace the window, the session participants jointly decided that the offender would help answer an office phone after school for a month. 287Id. During that time, his advisor and a social worker contacted the student’s family to offer support. 288Id.

While breaking a window is less serious than an assault or, worse, a shooting, a central theme to the practice of restorative justice is to teach students early about the consequences of their actions and to allow them to restore their integrity, reputation, and self-esteem as they seek to make amends for their wrongful actions. 289See González & Cairns, supra note 280, at 243 (“Central to the restorative practice is the maintenance of individuals’ dignity and sense of self-worth.”). By participating in restorative justice initiatives, students will avoid serious offenses later in their lives. But even when students engage in more serious offenses, restorative justice practices can help put students back on track to avoid even more serious offenses in the future. Schools that have implemented restorative justice practices have shown tremendous results. For example, West Philadelphia High School, formerly known as one of Philadelphia’s most dangerous schools, posted strong results after implementing restorative justice initiatives. 290Int’l Inst. for Restorative Practices, Improving School Climate: Findings from Schools Implementing Restorative Practices 6 (2009), available at http://www.iirp.edu/pdf/IIRP-Improving-School-Climate.pdf. After the first year, violent acts and serious incidents decreased by fifty-two percent; after the second year, violent acts and serious incidents decreased by an additional forty percent. 291Id. at 7. Several other schools implementing restorative justice practices also have improved school safety. 292Id. at 931; see also González & Cairns, supra note 280, at 252–53 (describing the benefits of restorative justice initiatives); cf. Martha Minow, Education for Co-Existence, 44 Ariz. L. Rev. 1, 5 (2002) (“Experts believe that teaching students how to negotiate and communicate, and how to mediate conflicts, can enhance students’ capacities to cooperate and to employ self-control, thereby reducing incidents of aggression at school.”). While West Philadelphia High School and perhaps other schools using restorative justice programs have not completely abandoned using strict security measures, implementing alternative measures to curb violence is an encouraging sign.

Third, educators can enhance safety and order in their schools by improving the strength and quality of classroom activities. 293Losen & Gillespie, supra note 252, at 36; see also Michael Eskenazi, Gillian Eddins & John M. Beam, Equity or Exclusion: The Dynamics of Resources, Demographics, and Behavior in the New York City Public Schools 2 (2003) (finding that teacher qualifications had a strong effect on student behavior). Having well-planned lessons and a varied instructional approach that includes hands-on learning activities, clear and well-defined behavioral expectations, teaching strategies that are targeted to meet students’ individual needs, and an empathetic approach focused on engaging and continually re-engaging students are vital to a safe learning environment. 294Eskenazi, Eddins & Beam, supra note 293, at 2; Losen & Gillespie, supra note 252, at 36; see also Fedders, Langberg, & Story, supra note 89, at 8. Such an environment provides students with a sense of purpose, commitment, and personal responsibility. It can help students feel that the educational process will work for them if they trust their teachers and commit themselves to the process. And above all, students need to feel and understand that teachers care about them, that teachers want them to succeed, that teachers believe that they can succeed, and that teachers are willing to do what it takes to help them succeed. 295See Fedders, Langberg, & Story, supra note 89, at 8 (explaining that proven preventive measures to improve school safety include having “staff who are positive, compassionate, nurturing, caring, and respectful; model appropriate behaviors; create a climate of emotional support; and are committed to maintaining strong, positive relationships with all students”); Steinberg, Allenworth & Johnson, supra note 125, at 1 (finding that the quality of relationships between educators and students most strongly defines safe schools).

Fourth, policymakers can promote safety by creating smaller schools where students can receive individualized attention and develop personal relationships with adults and other students. 296See Fedders, Langberg, & Story, supra note 89, at 8 (maintaining that a proven method to bolster school safety is “small classes and schools that students and staff experience as communities and where students and staff know each other well and feel responsible for one another”); Diane Walker, Smaller Schools: A Safer Alternative, About.com Secondary Educ., http://712educators.about.com/library/weekly/aa032901a.htm (last visited Aug. 13, 2013). Where this is not possible, large schools can divide children into smaller groups that remain together for an extended period of time and are taught by the same group of teachers. 297See Philip J. Cook et al., School Crime Control and Prevention, 39 Crime & Justice 313, 377 (2010). Research suggests that these smaller learning environments provide students with a better sense of community and reduce school crime. 298Id.

Fifth, schools can implement initiatives targeted to help students develop emotional and social stability. Social and emotional learning strategies enable students to “recognize and manage emotions, set and achieve positive goals, appreciate the perspectives of others, establish and maintain positive relationships, make responsible decisions, and handle interpersonal situations constructively.” 299Joseph A. Durlak et al., The Impact of Enhancing Students’ Social and Emotional Learning: A Meta-Analysis of School-Based Universal Interventions, 82 Child Dev. 405, 406 (2001). These programs provide instruction in understanding, integrating, and applying social and emotional skills in many contexts, which helps prevent violence, bullying, and substance abuse among students. 300Id.

Implementing these alternative programs to reduce school violence is not easy, but it is possible. Six New York City schools serving at-risk students all have successfully created safe, constructive learning environments without relying on strict security measures. 301See N.Y. Civ. Liberties Union, supra note 279, at 7–8. Common characteristics these schools share include efforts to promote dignity and respect among all school members; strong and compassionate leadership; open lines of communication among students, teachers, and school officials; and establishing clear, fair rules and disciplinary procedures. 302Id. at 7. Notably, none of these schools use a metal detector. 303Id. Since this report was published, one of these schools was forced to install a metal detector, which had a profound, negative result on the students. See supra pp. 17–18 and accompanying notes. The results truly have been remarkable. Each school enjoys above-average attendance, graduation rates, and substantially lower than average incidents of crime and suspension rates. 304See N.Y. Civ. Liberties Union, supra note 279, at 7.

C. Further Recommendations

While school-led reform is the most important component for instituting change, federal and state agencies must also play a role. 305See Glennon, supra note 10, at 107 (arguing that the federal government should take the lead in aligning educational policy with child development and encourage schools to employ positive behavior interventions that foster communities of trust rather than environments of distrust). An obvious place for federal and state agencies to begin is to stop providing grants to schools for strict security measures and make those funds available to support alternative initiatives. This is particularly important now that federal and state governments are considering providing money to schools for strict security measures after the Newtown tragedy. 306See supra note 3; see also The White House, supra note 265, at 11 (proposing to provide $150 million to school districts and law enforcement agencies to hire, among other individuals, more law enforcement officers to work in schools). I urge policymakers to use those funds to encourage schools to implement PBIS, restorative justice, and social and emotional learning programs; to provide training to teachers and school officials; to grant students access to additional mental health services; and to help schools hire more behavioral counselors and other specialists.

In addition, I encourage the U.S. Department of Education to provide grants to researchers to study more closely the harmful effects of strict security measures on students, especially when those measures are applied disproportionately to low-income and minority students. The Department of Education could disseminate the results of these studies to school districts nationally. Further, it could recommend that schools refrain from using strict security measures in favor of alternative, more effective methods and provide training to school officials. State boards of education could do the same for the schools in their respective states.

Finally, I recommend that the Department of Education’s Office of Civil Rights play a more active role in addressing the disproportionate use of strict security measures on minority students. The regulations to Title VI promulgated by the OCR recognize disparate impact as a form of discrimination. 307See, e.g., Peters v. Jenney, 327 F.3d 307, 319 (4th Cir. 2003). While a private plaintiff cannot seek relief under a Title VI regulation, the OCR can enforce those regulations. 308Id. As a policy matter, the Obama administration seems to be moving in the direction of allowing the Department of Education to address the disparate impact of school policies. See Exec. Order No. 13,621 § 2(b)(3)(vi), 77 Fed. Reg. 45471, 45472 (July 26, 2012); see also Thomas E. Perez, Assistant Attorney Gen. for Civil Rights, U.S. Dep’t of Justice, Remarks at the Civil Rights and School Discipline: Addressing Disparities to Ensure Equal Educational Opportunities Conference (Sept. 27, 2010) available at http://www.justice.gov/crt/speeches/perez_eosconf_speech.php. However, on the day the OCR released data demonstrating disparities in disciplinary procedures, Russlyn Ali, assistant secretary for the OCR, reminded the public that data on disparities do not necessarily prove a civil rights violation. See Losen & Gillespie, supra note 252, at 34. Recently, the OCR began investigating the discipline practices of approximately twenty school districts to examine any racial disparities, including Oakland Unified School District (OUSD). 309See Nirvi Shah, Feds: ‘Unprecedented’ Deal in Oakland on Black Student Suspensions, Educ. Week (Sept. 28, 2012, 12:44 PM), http://blogs.edweek.org/edweek/District_Dossier/2012/09/feds_unprecedented_deal_in_oak.html. Those investigations were part of the Obama administration’s efforts to address the overrepresentation of some racial groups in suspensions and expulsions from school. 310See Mary Ann Zehr, ‘Disparate Impact’ Discipline Policy Criticized, 20 Educ. Week, Feb. 23, 2011, at 27, 27 (2011). The investigations have proved effective thus far. On September 28, 2012, the Department of Education announced the voluntary resolution of its investigation of OUSD regarding whether African-American students were disciplined more often and more severely than white students. 311See Press Release, U.S. Dep’t of Educ., U.S. Department of Education Announces Voluntary Resolution of Oakland Unified School District Civil Rights Investigation (Sept. 28, 2012), available at http://www.ed.gov/news/press-releases/us-department-education-announces-voluntary-resolution-oakland-unified-school-di. OUSD made several commitments to address this problem. Those commitments included agreeing to collaborate with experts to develop positive school climates, identifying at-risk students and providing them with support services, revising its disciplinary policies, and providing training to school officials and educators. 312Id. I encourage the OCR to play a similar role to address the disproportionate use of strict security measures on minority students. 313This recommendation will be the subject of a future article. In addition, in subsequent scholarly articles, I intend to encourage courts to adopt an alternative framework to evaluate students’ rights under the Fourth Amendment.

Following these recommendations will help rectify the disproportionate use of strict security measures on minorities, reduce the overall number of schools that rely on strict security measures, provide a better learning environment for all students, and prepare children who will be more emotionally and socially balanced and less prone to commit violent acts.

Conclusion

The horrific massacre at Sandy Hook Elementary School has rocked our nation. It has caused us to think deeply about violence, guns, student safety, and school security measures. A natural response to this tragedy is to invest millions of dollars in strict security measures to prevent further violence from occurring. But this strategy is shortsighted and, indeed, misguided. The empirical evidence set forth here suggests that these actions may lead to further inequalities along racial and economic lines in our school systems. Further, there are more effective ways to address school crime that do not harm the educational climate. Such programs may not be easy to apply in urban schools that face challenging conditions, but the existence of successful schools in challenging environments provides tangible evidence that it is possible. 314See Noguera, supra note 122, at 207. In many of these schools, including inner-city schools, students view their experience as too special to be spoiled by crime or violence and too important to risk suspension or expulsion. These schools have an ethos of trust and belonging. 315Powell, supra note 100. Children want to attend these schools because they feel that they are part of a special community—a community whose members care for one another, protect one another, desire the best for one another, and expect the best from one another. 316Id. These are the schools that make a difference—a real difference—in the lives of students.

Footnotes

Assistant Professor of Law, University of Florida Levin College of Law. I would like to thank the following individuals for the helpful feedback and assistance they provided for this Article: Nancy Dowd, Paul Gugliuzza, Catherine Hancock, Lyrissa Lidsky, Tom C. W. Lin, Richard Peterson, Sharon Rush, and Michael Wolf. I also appreciate the helpful comments provided by faculty members during presentations of earlier drafts of this article at the 2012 Southeast Law Schools Jr./Sr. Faculty Workshop at Tulane Law School, the Southeastern Association of Law Schools 2012 Annual Conference, and 2013 Junior Scholars Conference at the University of Florida Levin College of Law’s Criminal Justice Center. I received outstanding research assistance from Olga Balderas and Dane Ullian. I am grateful for the summer research grant provided by the University of Florida Levin College of Law. I would also like to thank the U.S. Department of Education for providing access to the restricted-use data for the 2009–2010 School Survey on Crime and Safety. Some of the material in this article was presented as an excerpt in the Stanford Law Review Online. See Jason P. Nance, 65 Stan. L. Rev. Online 103 (2013).

1See, e.g., Tom Raum & Jennifer Agiesta, Poll: Americans Angrier About Sandy Hook than 9/11 Attacks, CNSNews (Jan. 16, 2013), http://cnsnews.com/news/article/poll-americans-angrier-about-sandy-hook-shooting-911-attacks (reporting the anger Americans feel over the shootings); Jeanette Rundquist, Surprise Security Drills Coming to N.J. Schools, NJ.com (Jan. 16, 2013, 9:07 PM), http://www.nj.com/news/index.ssf/2013/01/surprise_school_security_drills.html (“The tragedy in Connecticut was the school community’s 9/11. This has touched the very soul of the country, no less the school community.” (internal quotation marks omitted)).

2See, e.g., Stephanie Banchero & Caroline Porter, In Pursuit of Safety, Schools’ Paths Diverge: Districts Face Tight Budgets, Worries over Armed Security on Campus as They Respond to the Massacre at Sandy Hook, Wall St. J. (Jan. 17, 2013, 7:49 PM) http://online.wsj.com/article/SB10001424127887323783704578248113754373942.html (“Schools are running security drills, planning to install security cameras and bullet-proof glass, and hiring safety consultants.”); Lynh Bui, Montgomery County Parents Ask for More School Security, Teacher Training During Budget Hearing, Wash. Post (Jan. 11, 2013, 3:15 PM) http://www.washingtonpost.com/blogs/maryland-schools-insider/post/montgomery-county-parents-ask-for-more-school-security-teacher-training-during-budget-hearing/2013/01/11/e8d3dcf4-5aab-11e2-9fa9-5fbdc9530eb9_blog.html (explaining the increased concerns parents have for their children’s safety and their demand for increased on-campus security); Motoko Rich, School Officials Look Again at Security Measures Once Dismissed, N.Y. Times (Dec. 18, 2012) http://www.nytimes.com/2012/12/19/education/after-newtown-shootings-schools-consider-armed-security-officers.html?pagewanted=1&_r=1 (describing the reactions of policymakers to the Newtown shootings).

3See, e.g., H.B. 612, 2013 Leg., Reg. Sess. (Ala. 2013) (proposing funding to school districts for “surveillance cameras, metal detectors, and other safety equipment”); S.B. 2230, 2013 Leg., Reg. Sess. (N.Y. 2013) (funding security aid to provide schools with metal detectors, electronically operated partitions, and security cameras); S.B. 2267, 63d Leg. Assemb. (N.D. 2013) (proposing funding to equip schools with alarms, cameras, electronic door locks, emergency response call buttons, intercom systems, key or pass card systems, and metal detectors); H. 2343, Public School Security Equipment Grant Act of 2013, ch. 608 (Va. 2013) (providing funding to schools to for purchasing security equipment); see also Doug Finke, State Board of Education Calls for $874M Increase in School Funding, PJSTAR (Jan. 24, 2013, 9:02 PM), http://www.pjstar.com/news/x1503807290/State-Board-of-Education-calls-for-874M-increase-in-school-funding (“The Illinois State Board of Education called for an $874 million increase in state spending on elementary and secondary education on Thursday, including $20 million that would be available for school security measures.”); Mary Wilson, Pa. Senate Leader Suggests Steep Increase in School Security Funding, Newsworks (Jan 22, 2013), http://www.newsworks.org/index.php/local/item/49955-pa-senate-leader-suggests-steep-increase-in-school-security-funding?linktype=hp_topstorylist (“The leader of Pennsylvania’s Senate wants to increase grant funding twentyfold for school security, including armed guards.”); Liz Hayes, Burrell Board Focuses on School Safety, TribLive (May 7, 2013, 12:51 AM), http://triblive.com/neighborhoods/yourallekiskivalley/yourallekiskivalleymore/3970294-74/burrell-security-wagner#axzz2THQXVCDR (discussing a Pennsylvania School District that is considering using metal detectors more regularly, installing X-ray equipment to screen student backpacks, and making windows “more bulletproof”); Tom Barton, Board Members Unsure Metal Detectors Needed at Beaufort County Schools, Island Packet.com (May 8, 2013), http://www.islandpacket.com/2013/05/08/2495164/board-members-unsure-metal-detectors.html (discussing a South Carolina School Board’s consideration of adding metal detectors in schools).

4See, Ending the School-to-Prison Pipeline Before the Subcomm. on the Constitution, Civil Rights, and Human Rights, 112th Cong. (2012), available at http://www.judiciary.senate.gov/hearings/hearing.cfm?id=b61e5f08eadf22b2ec4ab964fc64ae9f; Susan Ferriss, ‘School to Prison Pipeline’ Hit on Capitol Hill, Ctr. for Pub. Integrity (Dec. 13, 2012, 6:00 AM), http://www.publicintegrity.org/2012/12/13/11921/school-prison-pipeline-hit-capitol-hill.

5 See, e.g., Advancement Project et al., Federal Policy, ESEA Reauthorization, and the School-to-Prison Pipeline 2 (2011), available at http://b.3cdn.net/advancement/ceb35d4874b0ffde10_ubm6baeap.pdf; NAACP Legal Def. & Educ. Fund, Inc., School to Prison Pipeline, NAACPLDF, http://www.naacpldf.org/case/school-prison-pipeline (last visited Aug. 12, 2013).

6Id.

7See Press Release, U.S. Dep’t of Educ., New Data from U.S. Department of Education Highlights Educational Inequalities Around Teacher Experience, Discipline, and High School Rigor (Mar. 6, 2012), available at http://www.ed.gov/news/press-releases/new-data-us-department-education-highlights-educational-inequities-around-teache.

8Id.

9Id.

10See e.g., Catherine Y. Kim et al., The School-to-Prison Pipeline: Structuring Legal Reform 53–54 (2010) (describing the overrepresentation of children of color in restrictive special education programs); Angela A. Ciolfi & James E. Ryan, Race and Response-to-Intervention in Special Education, 54 How. L.J. 303, 326–27 (2011) (explaining that African-American students are overrepresented in restrictive educational settings such as alternative schools and other disciplinary programs); Nancy E. Dowd, What Men? The Essentialist Error of the “End of Men, 93 B.U. L. Rev. (forthcoming) (describing the physical marginalization, the pattern of disproportionate discipline, and the overidentification and misidentification of learning disabilities of black males in school); Theresa Glennon, Knocking Against the Rocks: Evaluating Institutional Practices and the African American Boy, 5 J. Health Care L. & Pol’y 10, 11 (2002) (explaining that black males are more likely to be identified as disabled and are more likely to be placed in educational programs that exert greater control and provide fewer services than other children); Theresa Glennon, Looking for Air: Excavating Destructive Educational and Racial Policies to Build Successful School Communities, in Justice for Kids: Keeping Kids Out of the Juvenile Justice System 107, 110–11 (Nancy E. Dowd ed., 2011) (citing studies demonstrating that minority students are disproportionately disciplined); Russell J. Skiba, Suzanne E. Eckes & Kevin Brown, African American Disproportionality in School Discipline: The Divide Between Best Evidence and Legal Remedy, 54 N.Y.L. Sch. L. Rev. 1071, 1086–89 (2009/2010) (discussing the vast empirical evidence of racial disproportionality in school discipline).

11See infra Part III.

12See infra Part II.

13See infra Part III.

14See infra Part I.A.

15See infra Part I.B.

16See infra Part I.D.

17See infra Part I.D.

18See infra Part I.C.

19See infra Part IV.B.

20In a subsequent scholarly article, I also intend to describe an alternative framework that courts should employ to evaluate students’ Fourth Amendment rights in schools. See infra note 313.

21See Jason P. Nance, Random, Suspicionless Searches of Students’ Belongings: A Legal, Empirical, and Normative Analysis, 84 U. Colo. L. Rev. 367 (2013); see also James E. Ryan, The Supreme Court and Public Schools, 86 Va. L. Rev. 1335, 1415 (2000) (explaining that “the Court’s decisions regarding student searches rest on the value-laden view that maintaining discipline is necessary to preserve the educational process of schools”). Scholars have long debated how far students’ constitutional rights should extend in schools. But, as Catherine Kim recently observed, “both sides of the debate share a common starting point: such restrictions must be justified, if at all, by pedagogical goals.” Catherine Y. Kim, Policing School Discipline, 77 Brook. L. Rev. 861, 866 (2012); see also, e.g., Anne Proffitt Dupre, Should Students Have Constitutional Rights? Keeping Order in the Public Schools, 65 Geo. Wash. L. Rev. 49, 53 (1996) (applauding the Supreme Court’s recent jurisprudence that restricts students’ Fourth Amendment rights because doing so enhances the ability of school officials to provide students with a serious education); Betsy Levin, Educating Youth for Citizenship: The Conflict Between Authority and Individual Rights in the Public School, 95 Yale L.J. 1647, 1648–49 (1986) (describing the need to find an equilibrium between respecting students’ constitutional rights so that students learn to value those rights, and restricting students’ rights to help school officials maintain an orderly environment conducive to learning); Ryan, supra, at 1338–43, 1424–26 (maintaining that when a school engages in its core, universal function of imparting academic and vocational skills, it should be treated more deferentially and given some constitutional leeway to achieve its goals).

22See Vernonia School District 47J v. Acton, 515 U.S. 646, 653–54 (1995); New Jersey v. T.L.O., 469 U.S. 325, 340–41 (1985).

23See infra Part I.B.

24See infra Part I.C.

25See Nance, supra note 21, at 376–94.

26T.L.O., 469 U.S. at 328.

27Id.

28Id. at 328–29.

29Id. at 329.

30See id. at 332–33.

31Id. at 339–40.

32See id. at 337–38 (“A search of a child’s person or of a closed purse or other bag carried on her person, no less than a similar search carried out on an adult, is undoubtedly a severe violation of subjective expectations of privacy.” (footnote omitted)).

33Id. at 339 (citing Nat’l Inst. of Educ., U.S. Dep’t of Health, Educ. & Welfare, Violent Schools––Safe Schools: The Safe School Study Report to Congress 1 (1978)).

34See id. at 340–41.

35Id. at 337.

36Id. at 341 (quoting Terry v. Ohio, 392 U.S. 1, 20 (1968)).

37Id. at 343.

38See id. at 342 n.8 (“We do not decide whether individualized suspicion is an essential element of the reasonableness standard we adopt for searches by school authorities.”).

39Vernonia School District 47J v. Acton, 515 U.S. 646 (1995).

40Id. at 653.

41Id. at 650.

42Id.

43See Robert M. Bloom, The Story of Pottawatomie County v. Lindsay Earls: Drug Testing in the Public Schools, in Education Law Stories 337, 346 (Michael A. Olivas & Ronna Greff Schneider eds., 2008) (“At the time, James said, ‘I was like one of the smartest kids in class. I never got a referral (to the principal’s office) and I thought that was probably enough for them to see I wasn’t taking drugs.’”).

44Vernonia, 515 U.S. at 651.

45Id. at 651, 664–65.

46Id. at 653 (“The school search we approved in T.L.O., while not based on probable cause, was based on individualized suspicion of wrongdoing. As we explicitly acknowledged, however, ‘the Fourth Amendment imposes no irreducible requirement of such suspicion.’” (quoting New Jersey v. T.L.O., 469 U.S. 325, 342 n.8 (1985)) (internal quotation marks omitted)).

47Id. at 654, 658, 660.

48Id. at 656.

49See id.

50Id. at 658.

51Id. at 661–63.

52Id. at 681 (O’Connor, J., dissenting).

53Id. at 682.

54Id. (all but the first alteration in original) (quoting Tr. 9 (Apr. 29, 1992)).

55Bd. of Educ. v. Earls, 536 U.S. 822 (2002).

56Id. at 825.

57Id. at 849 (Ginsburg, J., dissenting).

58Id. at 826–27 (majority opinion).

59Id. at 827.

60Id. at 824–25.

61Id. at 830–38; see also Nance, supra note 21, at 384–87.

62Earls, 536 U.S. at 834–35.

63Id. at 834.

64See infra Part III. But see Nance, supra note 21, at 391–94 (arguing that the Fourth Amendment requires school officials to have particularized evidence of a substance use or weapons problem before performing suspicionless searches that are “highly intrusive,” such as a search through a student’s personal belongings).More recently, the Supreme Court addressed the constitutionality of a student search in Safford Unified School District Number 1 v. Redding, 557 U.S. 364 (2009). In Redding, the Court did not evaluate the constitutionality of a random, suspicionless search as it did in Vernonia and Earls; instead, the Court evaluated the legality of a strip search performed on a thirteen-year-old female student who was accused of bringing unauthorized prescription and over-the-counter drugs to school. Id. at 368. Relying on the same two factors the Court presented in T.L.O., the Court concluded that the search violated the Constitution because it was excessively intrusive in light of the age of the student and the nature of the school violation. Id. at 378 (citing New Jersey v. T.L.O., 469 U.S. 325, 342 (1985)). Nevertheless, the Court refused to hold the school officials accountable because the school officials were protected under the qualified immunity doctrine. Id. at 378–79. As some scholars have noted, this case demonstrates how difficult it is for students to recover for unlawful searches. E.g., Barry C. Feld, T.L.O. and Redding’s Unanswered (Misanswered) Fourth Amendment Questions: Few Rights and Fewer Remedies, 80 Miss. L.J. 847, 947–54 (2011) (explaining the impediments for students to recover for unlawful searches).

65Cases that permit the use of metal detectors include Hough v. Shakopee Pub. Sch., 608 F. Supp. 2d 1087, 1104 (D. Minn. 2009); In re F.B., 726 A.2d 361, 366 (Pa. 1999); In re Latasha W., 70 Cal. Rptr. 2d 886, 886–87 (Ct. App. 1998); and State v. J.A., 679 So. 2d 316, 319–20 (Fla. Dist. Ct. App. 1996).

66Cases that uphold random locker searches include State v. Jones, 666 N.W.2d 142, 150 (Iowa 2003); In re Isiah B., 500 N.W.2d 637, 641 (Wis. 1993); and In re Patrick Y., 746 A.2d 405, 414–15 (Md. 2000). For an extended discussion on the disagreement among courts regarding whether students possess an expectation of privacy in their lockers, see Kim et al., supra note 10, at 115–17; Feld, supra note 64, at 933–37; and Nance, supra note 21, at 411–12 and accompanying notes. Further, there is no rational basis to conclude that students should lose their expectation of privacy in their personal belongings simply because they place them in their lockers. See Nance, supra note 21, at 411–12 and accompanying notes.

67See, e.g., United States v. Taketa, 923 F.2d 665, 677 (9th Cir. 1991) (“Videotaping of suspects in public places, such as banks, does not violate the [F]ourth [A]mendment . . . .”). In certain locations, however, video surveillance—particularly surreptitious video surveillance—inappropriately intrudes on liberties that are fundamental to human dignity. See Brannum v. Overton Cnty. Sch. Bd., 516 F.3d 489, 499 (6th Cir. 2008) (holding that surreptitious video surveillance of a student locker room violates the Fourth Amendment because the surveillance intrudes on a student’s inherent right to bodily privacy).

68See infra Part III.

69See infra Part IV. The fact that the cumulative effect of using several strict measures together arguably amounts to a substantial invasion of students’ privacy provides justification for courts to apply a modified framework under the Fourth Amendment. I will explore this concept in greater detail in a subsequent scholarly article.

70See Office of Cmty. Oriented Policing Servs., 2011 Secure Our Schools Program, U.S. Dep’t of Justice, http://www.cops.usdoj.gov/pdf/2011AwardDocs/CSPP-SOS-CHP/SOSMethodology.pdf (last visited Aug. 12, 2013).

71Id.

72See Press Release, U.S. Dep’t of Justice, Office of Cmty. Oriented Policing Servs., U.S. Department of Justice COPS Office Announces over $13 Million in School Safety Grants (Sept. 8, 2011), available at http://www.cops.usdoj.gov/Default.asp?Item=2599.

73See Office of Cmty. Oriented Policing Servs., Fact Sheet: Secure Our Schools, U.S. Dep’t of Justice (Sept. 2011), http://www.cops.usdoj.gov/pdf/2011AwardDocs/CSPP-SOS-CHP/2011-SOS-Post-FactSheet.pdf.

7420 U.S.C. §§ 7101–7165 (2006).

75Id. § 7115(b)(2)(E)(ii). Notably, the grant application procedures make clear that the Department of Education views school security equipment as only one component to an overall strategy designed to create a safe and healthy learning environment. See U.S. Dep’t of Educ. et al., Safe Schools/Healthy Students: Information and Application Procedures for Fiscal Year 2009, at 22 (2009), available at http://www2.ed.gov/programs/dvpsafeschools/2009-184l.pdf. In the 2009 application, no more than ten percent of the grant could be used to fund security equipment. Id. The remainder of the grant must be used for programs designed to address student safety and health holistically by providing behavioral, social, and emotional support; mental health services; early childhood social and emotional learning programs; and alcohol, tobacco, and other drug prevention activities for students, their families, and the community. Id. at 22–24. This funding initiative currently is suspended. U.S. Dep’t of Educ., Safe Schools-Healthy Students Initiative: Application Information, ED.gov, http://www2.ed.gov/programs/dvpsafeschools/applicant.html (last modified Nov. 4, 2011).

76See Ala. Code § 41-15B-2.2 (2012) (“School Safety Enhancement Programs eligible for grants shall be designed to prevent or reduce violence in the schools . . . . The programs shall relate to one or more of the following: . . . (v) Safety plans involving the use of metal detectors, other security devices, uniforms, school safety resource officers, or other personnel employed to provide a safe school environment.”); Ga. Code Ann. § 20-2-1185 (West 2012) (“A public school may request funding assistance from the state for the installation of safety equipment including, but not limited to, video surveillance cameras, metal detectors, and other similar security devices.”); 24 Pa. Stat. Ann. § 13-1302-A (West 2012) (“[T]he office is authorized to make targeted grants to school entities to fund programs which address school violence, including . . . metal detectors, protective lighting, surveillance equipment . . . and training in the use of security-related technology.”); S.C. Code Ann. § 59-66-30 (2012) (“Using funds appropriated by the General Assembly, each public middle, junior high, and high school in the State must be equipped with one hand-held metal detector.”).

77See supra note 3.

78Pub. L. No. 107-110, 115 Stat. 1425 (2002) (codified as amended in scattered sections of 20 U.S.C.).

79Torin Monahan & Rodolfo D. Torres, Introduction to Schools Under Surveillance: Cultures of Control in Public Education 1, 5 (Torin Monahan & Rodolfo D. Torres, eds. 2010); U.S. Dep’t of Educ., Testing: Frequently Asked Questions, ED.gov, http://www2.ed.gov/nclb/accountability/ayp/testing-faq.html (last visited on Aug. 12, 2013).

80Monahan & Torres, supra note 79, at 5.

81See, e.g., Advancement Project, Test, Punish, and Push Out: How “Zero Tolerance” and High-Stakes Testing Funnel Youth into the School-to-Prison Pipeline 28–33 (2010), available at http://b.3cdn.net/advancement/d05cb2181a4545db07_r2im6caqe.pdf; Linda Darling-Hammond, Race, Inequality, and Educational Accountability: The Irony of ‘No Child Left Behind, 10 Race, Ethnicity & Educ. 245, 252–55 (2007); Deborah Gordon Klehr, Addressing the Unintended Consequences of No Child Left Behind and Zero Tolerance: Better Strategies for Safe Schools and Successful Students, 16 Geo. J. on Poverty L. & Pol’y 585, 602–03 (2009); Nance, supra note 21, at 397; James E. Ryan, The Perverse Incentives of the No Child Left Behind Act, 79 N.Y.U. L. Rev. 932, 969–70 (2004). Under the Federal Gun-Free Schools Act, states receiving federal education funds must have a state law that requires schools to expel students for at least one year for bringing a firearm to school. See 20 U.S.C. § 7151(b)(1). Although the NCLBA reenacted a softened version of this law by permitting superintendents to modify the expulsion requirement on a case-by-case basis, see id.; see also Federal Law on Guns in Schools, L. Center to Prevent Gun Violence (May 21, 2012), http://smartgunlaws.org/federal-law-on-guns-in-schools/, many states and schools adopted laws and policies modeled after the original Gun-Free Schools Act of 1994 by creating strict rules that impose predetermined consequences for certain acts, irrespective of the surrounding circumstances or intent of the students. See Klehr, supra, at 589. These so-called zero tolerance policies have pushed more students out of schools and have increased referrals to the juvenile justice system. Id. at 590.

82Ryan, supra note 81, at 969.

83Id. at 970.

84See Timeline: School Violence in the U.S., CNN.com (Dec. 14, 2012, 12:11 PM), http://www.cnn.com/2012/12/14/us/timeline-school-violence/index.html (detailing the number of violent acts in schools committed by students and outsiders).

85Cf. Amanda Terkel, Columbine High School Had Armed Guard During Massacre in 1999, Huffington Post (Dec. 21, 2012, 11:07 PM), http://www.huffingtonpost.com/2012/12/21/columbine-armed-guards_n_2347096.html (describing the Columbine High School shooting tragedy that occurred in the presence of an armed security officer).

86See id. (noting presence of armed guards); Marcus Wright, Experts Say Intrusive Security at Public Schools Reproduces Social Inequality, Mich. Citizen (Nov. 15, 2012), http://michigancitizen.com/dps-eaa-tighten-security/ (noting presence of metal detectors).

87See Arne Duncan, Resources for Schools to Prepare for and Recover from Crisis, Homeroom (Dec. 17, 2012), http://www.ed.gov/blog/2012/12/resources-for-schools-to-prepare-for-and-recover-from-crisis/ (explaining that not all tragedies in schools can be prevented).

88See, e.g., Monahan & Torres, supra note 79, at 2–3 (explaining that although schools continue to be the safest places for children, the threat of another school shooting haunts the social imagery); Matthew J. Mayer & Peter E. Leone, School Violence and Disruption Revisited: Equity and Safety in the School House, 40 Focus on Exceptional Child. Sept. 2007, at 1, 6 (“[M]edia coverage of school violence has shaped the public’s beliefs, and in many cases has led to a distorted perception of violence in schools, as well as adolescent violence more generally.” (citations omitted)); Randy Borum et al., What Can Be Done About School Shootings? A Review of the Evidence, 39 Educ. Researcher 27, 27 (2010) (“[S]chool shootings receive such intense publicity, and are such inherently disturbing events, that they generate an inflated perception of danger.” (citation omitted)).

89See Duncan, supra note 87 (“Schools are among the safest places for children and adolescents in our country, and, in fact, crime in schools has been trending downward for more than a decade.”); see also Barbara Fedders, Jason Langberg & Jennifer Story, School Safety in North Carolina: Realities, Recommendations & Resources 4 (May 2013), available at http://www.legalaidnc.org/public/learn/media_releases/2013_MediaReleases/school-safety-in-north-carolina.pdf (“School violence that results in death is extremely rare. Young people are much more likely to be harmed in the home or on the streets than they are in schools.” (footnote omitted)); Randall R. Beger, The “Worst of Both Worlds”: School Security and the Disappearing Fourth Amendment Rights of Students, 28 Crim. Just. Rev. 336, 338 (2003) (“Contrary to popular belief, schools remain among the safest places for children.”); Borum et al., supra note 88, at 27 (explaining that the number of homicides that occur on school grounds represents less than one percent of the annual homicides of children ages five to eighteen, and that “any given school can expect to experience a student homicide about once every 6,000 years”); Pedro A. Noguera, Schools, Prisons, and Social Implications of Punishment: Rethinking Disciplinary Practices, 42 Theory into Prac. 341, 343 (2003) (“Despite surveys that suggest a growing number of teachers and students fear violence in school, schools in the United States are generally safe places.” (citation omitted)).

90See Simone Robers et al., Indicators of School Crime and Safety: 2011, at iv (2012), available at http://nces.ed.gov/pubs2012/2012002rev.pdf.

91See Simone Robers et al., Indicators of School Crime and Safety: 2010, at iv (2010). But see id. (“In 2008, students ages 12 to 18 were victims of about 1.2 million nonfatal crimes (theft plus violent crime) at school, compared to about 1 million nonfatal crimes away from school.” (footnotes omitted)).

92Daryl Khan, Perps or Pupils? Safety Policy Creates Friction in New York City Schools, Center for Pub. Integrity (Sept. 21, 2012, 6:00 AM), http://www.publicintegrity.org/2012/09/21/10950/perps-or-pupils-safety-policy-creates-friction-new-york-city-schools.

93Id.

94Id.

95Id.

96Id.

97Id.

98Id.

99Id.

100Michael Powell, In a School Built on Trust, Metal Detectors Inject Fear, N.Y. Times (Sept. 17, 2012), http://www.nytimes.com/2012/09/18/nyregion/in-a-brooklyn-school-metal-detectors-inject-fear.html.

101Id.

102Id.

103Id. (internal quotation marks omitted).

104Id.

105Id. (internal quotation marks omitted).

106Id.

107See David Domenici & James Forman Jr., What It Takes to Transform a School Inside a Juvenile Justice Facility: The Story of the Maya Angelou Academy, in Justice for Kids: Keeping Kids Out of the Juvenile Justice System, supra note 10, at 283, 289 (“High-achieving schools are places where a culture of trust dominates.”); Roger D. Goddard, Megan Tschannen-Moran & Wayne K. Hoy, A Multilevel Examination of the Distribution and Effects of Teacher Trust in Students and Parents in Urban Elementary Schools, 102 Elementary Sch. J. 3, 3 (2001) (explaining that trusting in others is an important element of the learning process); Megan Tschannen-Moran & Wayne K. Hoy, A Multidisciplinary Analysis of the Nature, Meaning, and Measurement of Trust, 70 Rev. Educ. Res. 547, 547 (2000) (explaining that trusting in others is an important element of the learning process).

108See Donna Lieberman, Over-Policing in Schools on Students’ Education and Privacy Rights, NYCLU, http://www.nyclu.org/content/over-policing-schools-students-education-and-privacy-rights (last visited Aug. 12, 2013).

109 See Martin R. Gardner, Student Privacy in the Wake of T.L.O.: An Appeal for an Individualized Suspicion Requirement for Valid Searches and Seizures in the Schools, 22 Ga. L. Rev. 897, 943 (1988).

110Paul Hirschfield, School Surveillance in America: Disparate and Unequal, in Schools Under Surveillance: Cultures of Control in Public Education, supra note 79, at 38, 46.

111Id.

112Gardner, supra note 109, at 943.

113Id.

114Id. (footnotes omitted).

115Jen Weiss, Scan This: Examining Student Resistance to School Surveillance, in Schools Under Surveillance: Cultures of Control in Public Education, supra note 79, at 213, 227.

116Id. at 213–14.

117Id. at 213.

118See Lieberman, supra note 108.

119Id.

120Carol Ascher, Gaining Control of Violence in the Schools: A View from the Field, ERIC Digest, Sept. 1994, at 1, 4 (1994) (internal quotation marks omitted), available at http://www.eric.ed.gov/PDFS/ED377256.pdf (“Rather than offering reassurance, metal detectors and other mechanical devices, as well as security forces, are seen as providing a false sense of safety, if not a harsh symbol of the failure to create safe schools.”).

121See Advancement Project, Education on Lockdown: The Schoolhouse to Jailhouse Track 8 (2005), available at http://www.advancementproject.org/sites/default/files/publications/FINALEOLrep.pdf (explaining that while strict security measures “produce a perception of safety, there is little or no evidence that they create safer learning environments or change disruptive behaviors”); John Blosnich & Robert Bossarte, Low-Level Violence in Schools: Is There an Association Between School Safety Measures and Peer Victimization?, 81 J. Sch. Health 107, 107 (2011) (concluding that school security measures did not reduce violent behaviors related to bullying); Abigail Hankin, Marci Hertz & Thomas Simon, Impacts of Metal Detector Use in Schools: Insights from 15 Years of Research, 81 J. Sch. Health 100, 105 (2011) (concluding that there is insufficient evidence to demonstrate whether metal detectors reduce school violence); Matthew J. Mayer & Peter E. Leone, A Structural Analysis of School Violence and Disruption: Implications for Creating Safer Schools, 22 Educ. & Treatment Child. 333, 350, 352 (1999) (finding that student disorder and victimization were higher in schools using strict security measures than in schools that did not use such measures); Richard E. Redding & Sarah M. Shalf, The Legal Context of School Violence: The Effectiveness of Federal, State, and Local Law Enforcement Efforts to Reduce Gun Violence in Schools, 23 Law & Pol’y 297, 319 (2001) (“It is hard to find anything better than anecdotal evidence” to demonstrate that strict security measures such as metal detectors and guards reduce violence in schools). But see Ctrs. for Disease Control & Prevention, Violence-Related Attitudes and Behaviors of High School Students—New York City, 1992, 42 Morbidity & Mortality Wkly. Rep. 773 (1993), available at http://www.cdc.gov/mmwr/preview/mmwrhtml/00022011.htm (showing that students who attended schools using metal detectors were less likely to carry a weapon inside a school (7.8% versus 13.6%), but the presence of metal detectors did not reduce overall school violence); Renee Wilson-Brewer & Howard Spivak, Violence Prevention in Schools and Other Community Settings: The Pediatrician as Initiator, Educator, Collaborator, and Advocate, 94 Pediatrics 623, 626–27 (1994) (stating that one school system in New York City reported that after the school security staff began using hand-held metal detectors to conduct unannounced lobby searches of students at the beginning of the school day, weapon-related incidents decreased in thirteen of fifteen schools); Rachana Bhatt & Tomeka Davis, The Impact of Random Metal Detector Searches on School Violence, Contraband Possession, and Perceptions of Safety 22–23 (Dec. 2012) (unpublished manuscript), available at http://www2.gsu.edu/~ecorrb/index_files/RandomSearch.pdf (finding that, when comparing two geographically adjacent school districts in Florida, the school district that used random metal detector searches reduced the probability of a student bringing a weapon to school). Research on the effectiveness of school security measures is extremely limited, especially causal research demonstrating the effects of strict security measures. See Mayer & Leone, supra note 88, at 12.

122See Ascher, supra note 120, at 5 (maintaining that strict security measures “increase, rather than alleviate, tension in schools”); Beger, supra note 89, at 340 (citing several studies demonstrating that “aggressive security measures produce alienation and mistrust among students”); Michael Easterbrook, Taking Aim at Violence, 32 Psychol. Today, July–Aug. 2009 at 52, 56 (providing evidence that strict security measures alienate students); Clifford H. Edwards, Student Violence and the Moral Dimensions of Education, 38 Psychol. Schs. 249, 250 (2001) (arguing that “intrusive strategies are likely to undermine the trust needed to build cooperative school communities capable of really preventing violence”); Mayer & Leone, supra note 121, at 350, 352 (finding that student disorder and victimization were higher in schools using strict security measures than in schools that did not use such measures); Pedro A. Noguera, Preventing and Producing Violence: A Critical Analysis of Responses to School Violence, 65 Harv. Educ. Rev. 189, 190–91 (1995) (arguing that a “get tough” approach does not create a safe environment because coercive measures create mistrust and resistance among the student body); accord Aaron Kupchik, Homeroom Security: School Discipline in an Age of Fear 7, 15–18 (2010) (explaining that student misbehavior is likely to increase rather than decrease when students perceive they are treated unfairly and with disrespect).

123See Mayer & Leone, supra note 121, at 349.

124Id. at 351.

125Matthew P. Steinberg, Elaine Allensworth & David W. Johnson, Student and Teacher Safety in Chicago Public Schools: The Roles of Community Context and School Social Organization 15 (2011) (emphasis removed), available at http://ccsr.uchicago.edu/sites/default/files/publications/SAFETY%20IN%20CPS.pdf.

126Id. (emphasis removed).

127Id. at 2.

128See Sara Neufeld & Sumathi Reddy, Violent Week Renews Metal Detector Debate, Balt. Sun, Oct. 14, 2006, at 1A, available at http://articles.baltimoresun.com/2006-10-14/news/0610140131_1_metal-detectors-school-students-park-elementary-school.

129Id.

130See Wright, supra note 86.

131Neufeld & Reddy, supra note 128 (internal quotation marks omitted).

132Noguera, supra note 122, at 193; see also Ascher, supra note 120, at 5 (“[T]hose few students intent on bringing in weapons are inevitably a step ahead of the security devices, which means that enforcement activities alone cannot create a safe school.”); Neufeld & Reddy, supra note 128 (explaining that students interviewed in focus groups maintained that it was “easy to get around” metal detectors).

133 See Crystal A. Garcia, School Safety Technology in America: Current Use and Perceived Effectiveness, 14 Crim. Just. Pol’y Rev. 30, 40 (2003).

134See infra Part IV.B.

135See supra notes 121–22.

136See, e.g., Hankin, Hertz, & Simon, supra note 121, at 101 (“The Cleveland public school system estimated a cost of $3.7 million to incorporate walk-through metal detectors and X-ray scanners in each of 111 public schools.”); id. at 105 (“Metal detector programs are expensive, and funds spent on metal detectors would not be available for other programs and strategies that have been shown to be effective at reducing youth risk for violence and promoting pro-social behaviors.”); Dakarai I. Aarons, Mayor Herenton Vows to Add More Officers, Detectors to City’s Public Middle and High Schools, Com. Appeal (Feb. 13, 2008, 10:48 AM), http://www.commercialappeal.com/news/2008/feb/13/board-member-says-memphis-city-schools-should-be-s/ (explaining that a “feasibility study conducted by school administrators said it would cost the 115,000-student school district $4.5 million in equipment and staffing to conduct metal-detector screenings daily in middle and high schools”).

137See Aaron Kupchik & Geoff K. Ward, Reproducing Social Inequality Through School Security: Effects of Race and Class on School Security Measures 3–10 (unpublished manuscript), available at http://www.edweek.org/media/kupchikward-02security.pdf.

138Loïc Wacquant, Deadly Symbiosis: When Ghetto and Prison Meet and Mesh, 3 Punishment & Soc’y 95, 108 (2001).

139Id.

140See Kupchik & Ward, supra note 137, at 7 (“Marginalized youth are presumed to be young criminals and treated as such through exposure to criminal justice oriented practices (e.g., police surveillance and metal detectors), while youth with social, political and cultural capital are presumed to be well-behaved, treated as such, and empowered to be productive citizens.”).

141See Hirschfield, supra note 110, at 40 (arguing that the disproportionate use of strict security measures prepares urban minority students for certain positions in the post-industrial order, such as those of prisoners, soldiers, or service sector workers).

142Kupchik & Ward, supra note 137, at 6–7.

143Khan, supra note 92.

144Id. (internal quotation marks omitted).

145Id.; accord Victor M. Rios, Punished: Policing the Lives of Black and Latino Boys, at xiv, 133–38 (2011) (finding that low-income youth and young men of color disproportionately experience surveillance, and that there is a “system in which schools, police, probation officers, families, community centers, the media, businesses, and other institutions systematically treat young people’s everyday behaviors as criminal activity”).

146Khan, supra note 92.

147See Glennon, supra note 10, at 112 (arguing that schools cannot promote a positive school climate for learning when students are aware of the disparate application of discipline along racial lines).

148See Khan¸ supra note 92.

149Id.

150Id. (internal quotation marks omitted). Paul Hirschfield’s interview of a former inner-city student also illustrates the sense of unfairness that students perceive at schools that rely on strict security measures. This student explained:That school was run more like a prison than a high school. It don’t have to be nothing illegal about it. But you’re getting arrested. No regard for if a college going to accept you with this record. No regard for none of that, because you’re not expected to leave this school and go to college. You’re not expected to do anything.Paul J. Hirschfield, Preparing for Prison?: The Criminalization of School Discipline in the USA, 12 Theoretical Criminology 79, 79 (2008).

151See Constance Flanagan et al., School and Community Climates and Civic Commitments: Patterns for Ethnic Minority and Majority Students, 99 J. Educ. Psychol. 421, 423 (2007) (stating that studies have shown that minority groups have reported “a lower sense of school belonging than . . . their European American peers”); Rosa Hernández Sheets, Urban Classroom Conflict: Student–Teacher Perception: Ethnic Integrity, Solidarity, and Resistance, 28 Urb. Rev. 165, 175–76 (1996) (reporting in a study on classroom conflict that minority students believed that teachers did not respect them or care about them, that teachers abused their authority, that they were purposely pushed to the edge and were expected and encouraged to be hostile, and that teachers expected and wanted them to act out to justify disciplinary actions); Susan Rakosi Rosenbloom & Niobe Way, Experiences of Discrimination Among African American, Asian American, and Latino Adolescents in an Urban High School, 35 Youth & Soc’y 420, 434 (2004) (reporting that “[w]hen African American and Latino students were asked about their experiences with discrimination, they described hostile relationships with adults in positions of authority such as . . . teachers in school”); cf. Noguera, supra note 122, at 201 (describing the sentiment in many black communities that black children are being treated unfairly in schools).

152Khan¸ supra note 92. Many are concerned that post-Newtown legislation that could provide more funds for school security measures will exacerbate the school-to-prison pipeline. See, e.g., Jamelle Bouie, How Obama Might Make the School-to-Prison Pipeline Worse, Am. Prospect (Jan. 18, 2013), http://prospect.org/article/how-obama-might-make-school-prison-pipeline-worse; Julianne Hing, After Newtown, School Communities Brace for Another Influx of Police, Colorlines (Jan. 10, 2013), http://colorlines.com/archives/2013/01/militarized_post-newtown-schools.html; Simon McCormack, Gun Control Legislation Must Not Include More Cops in Schools: ACLU Letter to Biden, Huffington Post (Jan. 15, 2013), http://www.huffingtonpost.com/2013/01/14/gun-control-legislation_n_2474188.html.

153See, e.g., David Firestone, After Shootings, Nation’s Schools Add to Security, N.Y. Times, Aug. 13, 1999, at A1, available at http://www.nytimes.com/1999/08/13/us/after-shootings-nation-s-schools-add-to-security.html?pagewanted=all&src=pm; John D. Sutter, Columbine Massacre Changed School Security, CNN.com (Apr. 20, 2009, 9:30 AM), http://articles.cnn.com/2009-04-20/living/columbine.school.safety_1_metal-detectors-and-security-national-school-safety-school-security?_s=PM:LIVING.

154See, e.g., Julie Kiernan Coon, Security Technology in U.S. Public Schools (2007); Kupchik & Ward, supra note 137.

155See Nat’l Ctr. for Educ. Stats., School Survey on Crime and Safety Principal Questionnaire: 2009–10 School Year 5 (2010) [hereinafter 2009–2010 SSOCS Questionnaire], available at http://nces.ed.gov/surveys/ssocs/pdf/SSOCS_2010_Questionnaire.pdf (last visited Aug. 12, 2013).

156See Kupchik & Ward, supra note 137, at 14.

157Id. at 2.

158Id.

159Id. In fact, the percentage of racial/ethnic minority students was actually a negative predictor for use of drug-sniffing dogs in high schools. Id. at 42.

160Id. at 14.

161See id. at 41–42.

162See 2009–2010 SSOCS Questionnaire, supra note 155.

163See Statistical Standard Program: Getting Started, Nat’l Ctr. for Educ. Stats., http://nces.ed.gov/statprog/instruct_gettingstarted.asp (last visited Aug. 12, 2013). The restricted-use data “have a higher level of detail in the data compared to public-use data files.” Id. Although the restricted-use datasets are not available to the general public, datasets that contain less sensitive data for prior school years are currently available. Id. Those datasets can be downloaded at http://nces.ed.gov/surveys/ssocs/data_products.asp.

164NCES provides restricted-use datasets to certain researchers in qualified organizations. Id. To qualify, an organization must provide a justification for access to the restricted-use data, submit the required legal documents, agree to keep the data safe from unauthorized disclosures at all times, and to participate fully in unannounced, unscheduled inspections of the researcher’s office to ensure compliance with the terms of the License and the Security Plan form.Id.; see also Statistical Standards Program: Applying for a Restricted-Use Data License, Nat’l Ctr. for Educ. Stats., http://nces.ed.gov/statprog/instruct_apply.asp?type=rl (last visited Aug. 12, 2013) (providing guidelines for applying to receive restricted-use datasets).

165The Common Core of Data “is an NCES annual census system that collects fiscal and nonfiscal data on all public schools, public school districts, and state education agencies in the United States.” Nat’l Ctr. for Educ. Stats., 2009–2010 School Survey on Crime and Safety (SSOCS): Restricted-Use Data File User Manual 8 (2011) [hereinafter 2009–2010 Restricted-Use Manual] (on file with author); see also Helen M. Marks & Jason P. Nance, Contexts of Accountability Under Systemic Reform: Implications for Principal Influence on Instruction and Supervision, 43 Educ. Admin. Q. 3, 10–11 (2007) (describing the Common Core of Data). The CCD includes regular schools, charter schools, and schools that have magnet programs in the United States. It excludes schools in the U.S. territories and outlying areas, such as American Samoa, Guam, the Commonwealth of the Northern Mariana Islands, the U.S. Virgin Islands, and Puerto Rico, as well as overseas Department of Defense schools, newly closed schools, home schools, Bureau of Indian Education schools, nonregular schools, ungraded schools, and schools with a high grade of kindergarten or lower. 2009–2010 Restricted-Use Manual, supra at 8.

166A “sampling frame” is a list of units that could be selected for study. See Richard L. Scheaffer et al., Elementary Survey Sampling 43 (5th ed. 1996).

167See 2009–2010 Restricted-Use Manual, supra note 165, at 8.

168See id. at 9. The sample was stratified by instructional level (e.g., elementary school, middle school, high school), locale (e.g., rural, suburb, urban), enrollment size, and region (e.g., Northeast, Midwest, South, and West). In addition, the sample frame was stratified by percent of combined student population as black/African-American, Hispanic/Latino, Asian, native Hawaiian/other Pacific Islander, or American Indian/Alaska native. Id.

169Id. at 10.

170Id. at 1, 9–13. A response rate of seventy-six percent is very good and reduces bias in the data. Earl Babbie, The Practice of Social Research 256 (9th ed. 2001). NCES notes that some schools were more likely than others to respond to the survey. For example, schools more likely to respond included rural schools, schools with fewer students, combined schools, or those with a low percentage of combined black/African-American, Hispanic/Latino, Asian, native Hawaiian/other Pacific Islander, and American Indian/Alaska native students. 2009–2010 Restricted-Use Manual, supra note 165, at 9–10. While no category had a response rate lower than sixty-nine percent, see id. at 13, using a sample weight to analyze the data helped ameliorate the effects of the discrepancies in the response rates. See infra note 215.

1712009–2010 Restricted-Use Manual, supra note 165, at 1.

1722009–2010 SSOCS Questionnaire, supra note 155, at 5, 8.

173Id.

174Id.

175See supra Part II.

1762009–2010 SSOCS Questionnaire, supra note 155, at 5.

177Nevertheless, more precise questions that target how often schools perform these searches on students should be included in future distributions of the SSOCS. Nance, supra note 21, at 429 (recommending that the U.S. Department of Education reformulate many of the questions posed in the SSOCS).

178 See Robers et al., supra note 90, at 112. Racial data for the 2009–2010 SSOCS came from the 2007–2008 CCD school data file. See 2009–2010 Restricted-Use Manual, supra note 165, at 29. Although there was a two-year difference, it is highly unlikely that a school would experience a major shift in student population over a two-year period. A major racial shift in the student population for a school over a two-year period would require an extraordinary event such as a desegregation court order.

179 See, e.g., Federal Education Budget Project, New Am. Found., http://febp.newamerica.net/background-analysis/federal-school-nutrition-programs (last updated July 1, 2013) (“Researchers often use free or reduced price lunch (FRPL) enrollment figures as a proxy for poverty at the school level, because Census poverty data (which is used at the state and district level) is not available disaggregated below the school district level and is not collected annually.”); Michael Heise, Litigated Learning, Law’s Limits, and Urban School Reform Challenges, 85 N.C. L. Rev. 1419, 1441 (2007) (using student eligibility for free and reduced-price lunch as a proxy for student poverty).

180See Kupchik & Ward, supra note 137, at 15.

181See supra Part I.D.

182See 2009–2010 SSOCS Questionnaire, supra note 155, at 16.

183See Kelly Welch & Allison Ann Payne, Racial Threat and Punitive School Discipline, 57 Soc. Probs. 25, 27 (2010) (“One factor presumed to be closely associated with school punitiveness and disciplinary practice is the level of school crime and disorder.” (citation omitted)).

1842009–2010 SSOCS Questionnaire, supra note 155, at 11.

185Kupchik & Ward, supra note 137, at 17–18.

1862009–2010 SSOCS Questionnaire, supra note 155, at 11.

187Id.

188Id.

189Id.

190Id.

191Id.

192See Kupchik & Ward, supra note 137, at 17; Welch & Payne, supra note 183, at 27.

1932009–2010 SSOCS Questionnaire, supra note 155, at 13. Principals indicated whether these types of problems occurred (1) daily, (2) at least once a week, (3) at least once a month, (4) on occasion, or (5) never happens. Id.

194See Kupchik & Ward, supra note 137, at 17–18.

195Id. at 18 (controlling for the administrators’ perceptions of the level of crime in the neighborhoods around the schools).

196“Controlling for” means measuring the effect of one variable while taking into account the effect of another variable. See Will G. Hopkins, A New View of Statistics, Sportscience, http://www.sportsci.org/resource/stats/complex.html (last updated June 22, 2002). For example, researchers who wish to study the effect of gender on weight would want to “control for” height because of the known effect height has on weight. See id.

197See Kupchik & Ward, supra note 137, at 18.

198See id. at 16.

1992009–2010 SSOCS Questionnaire, supra note 155, at 7.

200Id.

201Creating dummy variables is a way to include categorical predictor variables in estimation models such as logistic regression. See Inst. for Digital Research & Educ., FAQ: What Is Dummy Coding?, UCLA.edu, http://www.ats.ucla.edu/stat/mult_pkg/faq/general/dummy.htm (last visited Aug. 12, 2013) (describing dummy coding).

202See Kupchik & Ward, supra note 137, at 1617.

203See 2009–2010 Restricted-Use Manual, supra note 165, at 25.

204See Welch & Payne, supra note 183, at 28 (“School urbanicity is one feature associated with increased student punishment . . . .”).

2052009–2010 Restricted-Use Manual, supra note 165, at 28–29.

2062009–2010 SSOCS Questionnaire, supra note 155, at 16.

207Middle schools are “schools in which the lowest grade is not lower than grade 4 and the highest grade is not higher than grade 9.” Nat’l Ctr. for Educ. Statistics, Crime, Violence, Discipline, and Safety in U.S. Public Schools: Findings from the School Survey on Crime and Safety: 2009–10, at 7 tbl.1 n.5 (2011), available at http://nces.ed.gov/pubs2011/2011320.pdf. High schools are “schools in which the lowest grade is not lower than grade 9 and the highest grade is not higher than grade 12.” Id. Combined schools include “other combinations of grades, including K–12 schools.” Id.

2082009–2010 SSOCS Questionnaire, supra note 155, at 17.

209Id.

210Kevin P. Brady, Sharon Balmer & Deinya Phenix, School–Police Partnership Effectiveness in Urban Schools: An Analysis of New York City’s Impact Schools Initiative, 39 Educ. & Urb. Soc’y 455, 456 (2007).

211See Kupchik & Ward, supra note 137, at 18–19 (transforming continuous variables by including their natural logs to correct for positive skewness).

212Id.

213See generally Joseph F. Hair, Jr. et al., Multivariate Data Analysis 276–81 (5th ed. 1998) (providing an overview of logistic regression analysis).

214For more information on SPSS, see Inst. for Digital Research & Educ., SPSS Starter Kit, UCLA.edu, http://www.ats.ucla.edu/STAT/spss/sk/default.htm (last visited July 25, 2013).

215The weighted variable compensates for unequal probabilities of selection, minimizes bias associated with responding and nonresponding schools, reduces sampling error, and calibrates the data to known population characteristics to produce optimal national estimates. See 2009–2010 Restricted-Use Manual, supra note 165, at 13 (describing the specific weighting procedures employed); see also Ibrahim S. Yansaneh, Construction and Use of Sample Weights ch. 5.3 (U.N. Group of Experts Meeting to Review the Draft Handbook on Designing of Household Sample Surveys, Draft U.N. Doc. ESA/STAT/AC.91/5, 2003), available at http://unstats.un.org/unsd/demographic/meetings/egm/Sampling_1203/docs/no_5.pdf. Furthermore, I adjusted the sample weight created by NCES by dividing it by its mean to create a mean weight of one. This is a recommended procedure when employing logistic regression analysis using SPSS. See Marks & Nance, supra note 165, at 14; Patty Glynn, Adjusting or Normalizing Weights “On the Fly” in SPSS, U. of Wash., http://staff.washington.edu/ glynn/adjspss.pdf (last updated July 8, 2004).

216Pursuant to the guidelines for presenting results from the restricted-use SSOCS database, raw sample numbers have been rounded to the nearest ten. U.S. Dep’t. of Educ., Restricted-Use Data Procedures Manual 20 (2011), available at http://nces.ed.gov/pubs96/96860rev.pdf.

217For example, the average high school in the United States has between 700 and 800 students. See Overview of Public Elementary and Secondary Schools and Districts: School Year 1999–2000, Nat’l Ctr. for Educ. Stats. (Sept. 2001), http://nces.ed.gov/pubs2001/overview/table05.asp (last visited Aug. 12, 2013). Multiplying 700 students by 3,344 schools produces a rough estimate of 2,340,800 students.

218The variance inflation factors (VIF), a common statistic to detect multicollinearity, indicated that multicollinearity was not a problem for the models. In addition, the Hosmer & Lemeshow test, a statistical test to assess a lack of fit between the data and the model, indicated a goodness of fit for each of the models.

219See Raymond E. Wright, Logistic Regression, in Reading and Understanding Multivariate Statistics 217, 223 (Laurence G. Grimm & Paul R. Yarnold eds., 1995) (“The odds ratio estimates the change in the odds of membership in the target group for a one-unit increase in the predictor.”). For example, hypothetically, if the odds ratio for the independent variable “% minority” were two, then the odds of a school employing a certain combination of strict security measures would be twice greater for schools having a minority population of fifty percent than for schools having a minority population of forty-nine percent.

220Statistical significance measures the likelihood of whether the regression coefficient, or the effect of the independent variable, is different from zero in the population of schools from which the sample was drawn. Id. at 226–27.

221See Ascher, supra note 120, at 3 (“The inability of teachers and other school staff to make meaningful connections with students in large schools and increasingly large classes has become a key safety issue.”).

222See infra Part IV.B.

223See Kupchik & Ward, supra note 137, at 1617.

224See, e.g., Brady, Balmer, & Phenix, supra note 210, at 45657 (“An increasing fear of school violence coupled with the public’s misperceptions of [school safety] . . . has caused school officials, especially those located in urban areas, to implement more punitive-based . . . policies . . . .”); Noguera, supra note 122, at 206 (noting that most urban high schools “rely upon coercion or excessive forms of control” to ensure school safety).

225See Kupchik & Ward, supra note 137, at 24–25.

226See supra Part I.A.

227See supra Parts I.C–D.

228See Wacquant, supra note 138, at 108.

229See Hirschfield, supra note 110, at 45; Noguera, supra note 122, at 192.

230See Robers et al., supra note 90, at 164 (disaggregating the use of security measures by racial and socioeconomic categories of students).

231Hirschfield, supra note 150, at 84. Recent data from the U.S. Department of Education also show that schools that serve affluent or white populations tend to rely more on drug-sniffing dogs than other schools. See Robers et al., supra note 90, at 165. But, as the empirical analysis presented here demonstrates, schools serving low-income minority students are much more likely to use multiple methods that create an intense surveillance environment than other schools. See supra Part III.

232See Noguera, supra note 89, at 348.

233Id. at 344.

234Id.; see also Levin, supra note 21, at 164849.

235Noguera, supra note 89, at 344.

236Id.

237Id.; see also Steinberg, Allensworth & Johnson, supra note 125, at 27–31 (finding that the academic skills of the students are strongly related to school safety); Paul Willis, Learning to Labor: How Working Class Kids Get Working Class Jobs 72 (1977) (arguing that for students who believe that knowledge and credentials acquired in schools are irrelevant, “the teachers’ authority becomes increasingly the random one of the prison guard, not the necessary one of the pedagogue”).

238See Steinberg, Allensworth & Johnson, supra note 125, at 46 (maintaining that lower-achieving students are less likely to be engaged in school, more likely to feel frustrated by their performance, more likely to act out, and less likely to respond to academic punishments).

239Noguera, supra note 89, at 344.

240Id.

241Id. at 345.

242Id.

243See supra Part III, Table 1 and accompanying discussion.

244See Hirschfield, supra note 150, at 92.

245Id.

246Michelle Fine et al., Civics Lessons: The Color and Class of Betrayal, 106 Tchrs. C. Rec. 2193, 2204–05 (2004) (finding that students believed that their teachers considered them to be “animals,” “inmates,” or “killers”); Pedro A. Noguera, The Trouble with Black Boys: The Role and Influence of Environmental and Cultural Factors on the Academic Performance of African American Males, 38 Urb. Educ. 431, 448 (2003) (finding that black students were much less likely than white students to believe that their teachers support and care about them).

247Ascher, supra note 120, at 4 (describing the pressures that many urban educators teaching disadvantaged students face, making those educators feel frustrated, overworked, stressed, cynical, and burned out).

248See Hirschfield, supra note 150, at 92.

249See Ascher, supra note 120, at 3–4 (describing the inadequate resources of many urban school serving disadvantaged populations).

250See supra Part I.C.

251Welch & Payne, supra note 183, at 28.

252Daniel J. Losen & Jonathan Gillespie, The Civil Rights Project, Opportunities Suspended: The Disparate Impact of Disciplinary Exclusion from School 6 (2012), available at http://civilrightsproject.ucla.edu/resources/projects/center-for-civil-rights-remedies/school-to-prison-folder/federal-reports/upcoming-ccrr-research/losen-gillespie-opportunity-suspended-2012.pdf.

253Id.

254Id. at 7.

255Id. at 32–33.

256See Ciolfi & Ryan, supra note 10, at 327–28 (citing studies demonstrating that students of color are punished disproportionality relative to their behavior).

257Welch & Payne, supra note 183, at 36.

258Id. at 36–37.

259Id. at 37.

260See, e.g., Tony Fabelo et al., Breaking Schools’ Rules: A Statewide Study of How School Discipline Relates to Students’ Success and Juvenile Justice Involvement 45 (2011), available at http://justicecenter.csg.org/files/Breaking_Schools_Rules_Report_Final.pdf (showing that in a statewide Texas study that blacks were disproportionately likely to be disciplined for “discretionary reasons” and more likely to be removed from class); Catherine P. Bradshaw et al., Multilevel Exploration of Factors Contributing to the Overrepresentation of Black Students in Office Disciplinary Referrals, 102 J. Educ. Psychol. 508, 508 (2010) (finding that after controlling for teacher ratings of students’ behavior problems, black students were more likely than white students to receive office disciplinary referrals); Daniel J. Losen, The Civil Rights Project, Discipline Policies, Successful Schools, and Racial Justice 4 (2011), available at http://nepc.colorado.edu/files/NEPC-SchoolDiscipline.pdf.

261See supra note 8 and accompanying text. Yet, disturbingly, there is currently almost no legal recourse to address these racial inequalities. See Glennon, supra note 10, at 120–24 (describing the limitations of addressing racial inequalities in public schools under the Fourteenth Amendment and Title VI of the Civil Rights Act of 1964); Skiba, Eckes & Brown, supra note 10, at 1089–1100 (describing the limitations of addressing racial inequalities in public schools under the Fourteenth Amendment and Title VI of the Civil Rights Act of 1964). In a subsequent research project, I intend to propose an alternative framework to address the disparate use of strict security measures on minorities.

262See Ascher, supra note 120, at 4 (maintaining that metal detectors and other mechanical devices only provide a false sense of safety).

263Id. at 7 (describing the benefits for children of adopting alternative methods to reduce school violence).

264Joe Green, South Jersey Schools Discuss Safety Following ‘Perspectives After Newtown,NJ.com (Jan. 23, 2013, 8:00 AM), http://www.nj.com/gloucester-county/index.ssf/2013/01/south_jersey_schools_to_talk_s.html (internal quotation marks omitted).

265See Robert A. Fein et al., U.S. Secret Serv. & U.S. Dep’t of Educ., Threat Assessment in Schools: A Guide to Managing Threatening Situations and to Creating Safe School Climates 1112 (2002), available at http://www.secretservice.gov/ntac/ssi_guide.pdf. This report was recently emphasized by the Obama administration after the Newtown shootings. See The White House, Now Is the Time 1213 (2013), available at http://www.whitehouse.gov/sites/default/files/docs/wh_now_is_the_time_full.pdf.

266Fein et al., supra note 265, at 11; see also Fedders, Langberg & Story, supra note 89, at 6 (“Positive relationships among students, families, teachers, administrators, and staff are the most effective tools in creating a safe school environment.”).

267Steinberg, Allensworth, & Johnson, supra note 125, at 1.

268As scholar Jim Ryan reminds us,From the birth of the common school movement through early desegregation cases, schools were seen not simply as places where students learned how to read and write but also as places where they learned how to become better citizens. Indeed, for a long time, the socializing or civic mission of schools was considered by many to be just as important as the academic mission.James E. Ryan, The Supreme Court and Voluntary Integration, 121 Harv. L. Rev. 131, 143 (2007) (footnote omitted).

269See Fedders, Langberg & Story, supra note 89, at 8 (recommending PBIS as a proven measure to improve school safety); Losen & Gillespie, supra note 252, at 43; Mayer & Leone, supra note 88, at 13 (explaining that PBIS is an approach that “transforms the school environment to support overall student success, behaviorally, socially, and academically”); School-Wide PBIS, Positive Behav. Interventions & Supports, http://www.pbis.org/school/default.aspx (last visited Aug. 13, 2013) (describing school-wide PBIS).

270See What Is School-Wide Positive Behavioral Interventions & Supports?, Positive Behav. Interventions & Supports, http://www.pbis.org/school/what_is_swpbs.aspx (lasted visited Aug. 13, 2013).

271Id.

272Id.; see also Is School-Wide Positive Behavior Support an Evidence-Based Practice?, Positive Behav. Interventions & Supports (March 2009), http://www.pbis.org/research/default.aspx.

273See What Is School-Wide Positive Behavioral Interventions & Support?, supra note 270.

274See Is School-Wide Positive Behavior Support an Evidence-Based Practice?, supra note 272.

275Id.

276Id.

277Id.; see also Mayer & Leone, supra note 88, at 13 (explaining the benefits of implementing PBIS for students).

278See Frequently Asked Questions, Positive Behav. Interventions & Supports, http://www.pbis.org/school/primary_level/faqs.aspx (last visited Aug. 13, 2013); see also Domenici & Forman Jr., supra note 107, at 290 (employing successfully a modified version of PBIS in a school inside of a juvenile justice facility to encourage students to behave according to school values).

279N.Y. Civ. Liberties Union, Safety with Dignity: Alternatives to the Over-Policing of Schools 8 (2009), available at http://www.nyclu.org/files/Safety_with_Dignity.pdf.

280Thalia N. C. González & Benjamin Cairns, Moving Beyond Exclusion: Integrating Restorative Practices and Impacting School Culture in Denver Public Schools, in Justice for Kids: Keeping Kids Out of the Juvenile Justice System, supra note 10, at 241, 241.

281Id.; see also Laura Mirsky, Building Safer, Saner Schools, 69 Educ. Leadership, Sept. 2011, at 45, 4549 (2011); Laura Mirsky, SaferSanerSchools: Transforming School Culture with Restorative Practices, Restorative Pracs. Eforum 1 (May 20, 2003), http://www.iirp.edu/iirpWebsites/web/uploads/article_pdfs/ssspilots.pdf.

282See Mirsky, Building Safer, Saner Schools, supra note 281, at 4549; Mirsky, SaferSanerSchools, supra note 281, at 1; cf. Catherine J. Ross, A Place at the Table: Creating Presence and Voice for Teenagers in Dependency Proceedings, 6 Nev. L.J. 1362, 1365 (2006) (“We have known for decades that involving young people in decision-making promotes more positive outcomes.”).

283See Mirsky, Building Safer, Saner Schools, supra note 281, at 4549.

284See N.Y. Civ. Liberties Union, supra note 279, at 18.

285Id.

286Id.

287Id.

288Id.

289See González & Cairns, supra note 280, at 243 (“Central to the restorative practice is the maintenance of individuals’ dignity and sense of self-worth.”).

290Int’l Inst. for Restorative Practices, Improving School Climate: Findings from Schools Implementing Restorative Practices 6 (2009), available at http://www.iirp.edu/pdf/IIRP-Improving-School-Climate.pdf.

291Id. at 7.

292Id. at 931; see also González & Cairns, supra note 280, at 252–53 (describing the benefits of restorative justice initiatives); cf. Martha Minow, Education for Co-Existence, 44 Ariz. L. Rev. 1, 5 (2002) (“Experts believe that teaching students how to negotiate and communicate, and how to mediate conflicts, can enhance students’ capacities to cooperate and to employ self-control, thereby reducing incidents of aggression at school.”). While West Philadelphia High School and perhaps other schools using restorative justice programs have not completely abandoned using strict security measures, implementing alternative measures to curb violence is an encouraging sign.

293Losen & Gillespie, supra note 252, at 36; see also Michael Eskenazi, Gillian Eddins & John M. Beam, Equity or Exclusion: The Dynamics of Resources, Demographics, and Behavior in the New York City Public Schools 2 (2003) (finding that teacher qualifications had a strong effect on student behavior).

294Eskenazi, Eddins & Beam, supra note 293, at 2; Losen & Gillespie, supra note 252, at 36; see also Fedders, Langberg, & Story, supra note 89, at 8.

295See Fedders, Langberg, & Story, supra note 89, at 8 (explaining that proven preventive measures to improve school safety include having “staff who are positive, compassionate, nurturing, caring, and respectful; model appropriate behaviors; create a climate of emotional support; and are committed to maintaining strong, positive relationships with all students”); Steinberg, Allenworth & Johnson, supra note 125, at 1 (finding that the quality of relationships between educators and students most strongly defines safe schools).

296See Fedders, Langberg, & Story, supra note 89, at 8 (maintaining that a proven method to bolster school safety is “small classes and schools that students and staff experience as communities and where students and staff know each other well and feel responsible for one another”); Diane Walker, Smaller Schools: A Safer Alternative, About.com Secondary Educ., http://712educators.about.com/library/weekly/aa032901a.htm (last visited Aug. 13, 2013).

297See Philip J. Cook et al., School Crime Control and Prevention, 39 Crime & Justice 313, 377 (2010).

298Id.

299Joseph A. Durlak et al., The Impact of Enhancing Students’ Social and Emotional Learning: A Meta-Analysis of School-Based Universal Interventions, 82 Child Dev. 405, 406 (2001).

300Id.

301See N.Y. Civ. Liberties Union, supra note 279, at 7–8.

302Id. at 7.

303Id. Since this report was published, one of these schools was forced to install a metal detector, which had a profound, negative result on the students. See supra pp. 17–18 and accompanying notes.

304See N.Y. Civ. Liberties Union, supra note 279, at 7.

305See Glennon, supra note 10, at 107 (arguing that the federal government should take the lead in aligning educational policy with child development and encourage schools to employ positive behavior interventions that foster communities of trust rather than environments of distrust).

306See supra note 3; see also The White House, supra note 265, at 11 (proposing to provide $150 million to school districts and law enforcement agencies to hire, among other individuals, more law enforcement officers to work in schools).

307See, e.g., Peters v. Jenney, 327 F.3d 307, 319 (4th Cir. 2003).

308Id. As a policy matter, the Obama administration seems to be moving in the direction of allowing the Department of Education to address the disparate impact of school policies. See Exec. Order No. 13,621 § 2(b)(3)(vi), 77 Fed. Reg. 45471, 45472 (July 26, 2012); see also Thomas E. Perez, Assistant Attorney Gen. for Civil Rights, U.S. Dep’t of Justice, Remarks at the Civil Rights and School Discipline: Addressing Disparities to Ensure Equal Educational Opportunities Conference (Sept. 27, 2010) available at http://www.justice.gov/crt/speeches/perez_eosconf_speech.php. However, on the day the OCR released data demonstrating disparities in disciplinary procedures, Russlyn Ali, assistant secretary for the OCR, reminded the public that data on disparities do not necessarily prove a civil rights violation. See Losen & Gillespie, supra note 252, at 34.

309See Nirvi Shah, Feds: ‘Unprecedented’ Deal in Oakland on Black Student Suspensions, Educ. Week (Sept. 28, 2012, 12:44 PM), http://blogs.edweek.org/edweek/District_Dossier/2012/09/feds_unprecedented_deal_in_oak.html.

310See Mary Ann Zehr, ‘Disparate Impact’ Discipline Policy Criticized, 20 Educ. Week, Feb. 23, 2011, at 27, 27 (2011).

311See Press Release, U.S. Dep’t of Educ., U.S. Department of Education Announces Voluntary Resolution of Oakland Unified School District Civil Rights Investigation (Sept. 28, 2012), available at http://www.ed.gov/news/press-releases/us-department-education-announces-voluntary-resolution-oakland-unified-school-di.

312Id.

313This recommendation will be the subject of a future article. In addition, in subsequent scholarly articles, I intend to encourage courts to adopt an alternative framework to evaluate students’ rights under the Fourth Amendment.

314See Noguera, supra note 122, at 207.

315Powell, supra note 100.

316Id.