Safety from the Inside Out (#)

Fall 2004

Safety from the Inside Out

Rethinking Traditional Approaches

By Alfie Kohn

For many people, the idea of safety in an educational context brings to mind the problem of school violence, and specifically the string of shootings at schools across the country in recent years. Let’s begin, then, by noting that the coverage of these events has obscured several important facts:

* The real horror is that young people die, not where they die. To be sure, there’s something deeply unsettling about the juxtaposition of the words “violence” and “schools.” But keep in mind that the vast majority of young homicide victims are killed at home, on the streets, or somewhere else other than school. During one three-year period in the 1990s, for example, about eighty homicides took place on school grounds . . . while more than eight thousand children were killed elsewhere. This is important to keep in mind both so that we recognize the full extent of the problem and so we don’t exaggerate how dangerous schools really are.

* There is a tendency, upon hearing about stunning cases of school violence, to infer that adolescents are Public Enemy No. 1. But Mike Males, a sociologist, urges us to focus our attention on the “far more common phenomenon of adults killing kids.” He points out that Americans blame teenagers too easily, and usually inaccurately, for what’s wrong with our society.[1]

* When school violence does occur, low-income students of color are disproportionately likely to be the victims, notwithstanding Columbine and other notorious school-shooting incidents. If that fact is surprising, it may be because of the media’s tacit assumption that any problem – crime, drugs, violence — is more newsworthy when white people in the suburbs are affected.


Yet another series of mistaken assumptions comes into play when educators and policy makers try to respond to violence – or to their fear of it. Questionable beliefs often lead to wrongheaded policies.

First, we Americans love to imagine that technical fixes will take care of complicated problems. (Remember the V-chip, which was supposed to be the solution to children’s exposure to violent television programming?) Some people still cling to the hope that schools can be made safer if we just install enough surveillance cameras and metal detectors. In reality, though, it’s simply not feasible to guard every doorway or monitor every screen. The number of cameras at one Washington, D.C. high school was recently doubled, from 32 to 64, but the principal admitted that it’s hard to keep guns “on the outside of the school unless we become armed camps, and I don’t think anyone wants to send their child to an armed camp.” His comments were reported in a newspaper article that was aptly headlined “Trust, Not Cameras, Called Best Prevention.”[2]

Pedro Noguera, who teaches at New York University, put it this way: “Design and staffing of schools are driven by security concerns, but no thought is given to how these designs and atmospheres make students and [teachers] feel. If we use prisons as our models for safe schools – well, prisons are not safe places, right? Safety comes from human relations. I’d say we’d do much better to invest in counselors than armed guards.”[3]

Second, when we do focus on the human element of violence prevention, we often assume that students just need to be taught the appropriate skills.[4] This model is so simple and familiar to us that we don’t even think of it as a model at all. It seems a matter of common sense that if children don’t pay attention to what someone else is saying, they would benefit from some remedial listening skills. If they fail to lend a hand to someone in distress, they need to hone their helping skills. If they’re reluctant to stand up for themselves, they’re candidates for assertiveness training. Thus, by analogy, if violence keeps breaking out, all we need to do is teach students the skills of conflict resolution.

Unfortunately, skills are not enough. Most kids already know how to listen, how to help, and how to assert themselves. The question is why they sometimes lack the disposition to act in these ways. It’s much the same with efforts to raise academic achievement: A skills-based approach has its limits if we ignore the question of how interested students are in what they’re being taught. Such efforts may even do more harm than good if an emphasis on teaching basic skills makes school downright unappealing. The same goes for literacy in particular: Consider how many children know how to read, but don’t. In short, what matters is not only whether people can learn, or act, in a particular way, but whether they have the inclination to do so.

Why, then, do we spend so much time teaching skills? For one thing, this implies that it’s the students who need fixing. If something more complicated than a lack of know-how is involved, we might have to question our own practices and premises, which can be uncomfortable. Moreover, a focus on skills allows us to ignore the structural elements of a classroom (or school or family). If students hurt one other, it’s easier for us to try to deal with each individual’s actions than it is to ask which elements of the system might have contributed to the problem.

A skills-based approach is also compatible with behaviorism, whose influence over our schools – and, indeed, over all of American society — is difficult to overstate. Behaviorism dismisses anything that can’t be reduced to a discrete set of observable and measurable behaviors. This dogma lies behind segmented instructional techniques, as well as many of the most popular approaches to character education, classroom management, and our practices with students who have special needs.

When we’re preoccupied with behaviors, we’re less likely to dig deep in order to understand the reasons, values, and motives that give rise to those behaviors. We end up embracing superficial responses, such as trying to improve the climate of a school by forcing students to dress alike. (Among other limitations of a policy like this, our assumption seems to be that we can reduce aggression by borrowing an idea from the military.) But any time we talk about changing students’ “behaviors,” we run the risk of ignoring the students who are doing the behaving. We lose the human beings behind the actions. Thus, we may come to see students as computers that can be reprogrammed, or pets that can be retrained, or empty receptacles that can be refilled – all dangerously misleading metaphors. We offer behavioral instruction in more appropriate ways to express anger, but the violence continues because we haven’t gotten anywhere near where the problem is.


It often doesn’t work very well, then, to employ technical fixes or to teach skills. But there’s a third response that isn’t merely ineffective; it’s actively counterproductive. I have in mind the policies that follow from assuming we can stamp out violence – or create safety – by coercive means. In her book The Peaceable School, Vicky Dill remarked that while it can be bad to have no plan for dealing with school violence, “it can be much worse to have a simplistic, authoritarian policy.”[5]

A reliance on old-fashioned discipline, with threats of punishments for offenders, not only distracts us from dealing with the real causes of aggression, but in effect models bullying and power for students. Many school officials fail to understand this fact and end up throwing fuel on the fire by responding to signs of student distress with ever-harsher measures. Consistent with the tendency to ignore the structural causes of problems, they seem to think sheer force will make the bad stuff go away; if students are made to suffer for doing something wrong, they will see the error of their ways. When that proves ineffective, it’s assumed that more punishment — along with tighter regulations and less trust — will do the trick.[6]

The shootings at Columbine provoked a general panic in which hundreds of students across the country were arrested, while “countless others were suspended or expelled for words or deeds perceived as menacing.”[7] The fear here is understandable: Administrators wondered whether their district, too, might be incubating killers. But we need to understand the difference between overreaction, such as closing down a school to search for bombs after a student makes an off-hand joke, and destructive reactions, such as coercive policies.

A particularly egregious example of the latter is the so-called “zero tolerance” approach, which is based on the premise that harsh punishment will work better if it’s meted out indiscriminately – indeed, in robotic fashion. It took a few years before this strategy began to attract critical attention in the media.[8] Research, meanwhile, has been accumulating to confirm that it makes no sense at all. One study discovered that students in schools with such a policy “actually report feeling less safe . . . than do students in schools with more moderate policies.”[9] That subjective impression is supported by objective evidence: Another analysis showed that “even after schools with zero tolerance policies had implemented them for more than four years, those schools were still less safe than schools without such policies.”[10] Moreover, zero tolerance doesn’t affect everyone equally: African-American and Latino students are more likely than their white counterparts to be targeted by this sort of punitive discipline.[11] As a society, we seem to have a lot more tolerance for the misbehavior of white children.

The finding that schools become less safe as a result of adopting zero-tolerance policies will sound paradoxical only to those readers who believe that threats and punishment can create safety. In reality, safety is put at risk by such an approach. A safe school environment is one where students are able to really know and trust – and be known and trusted by – adults.[12] Those bonds, however, are ruptured by a system that’s about doing things to students who act inappropriately rather than working with them to solve problems. “The first casualty” of zero-tolerance policies “is the central, critical relationship between teacher and student, a relationship that is now being damaged or broken in favor of tough-sounding, impersonal, uniform procedures.”[13]

Zero-tolerance is bad enough, but the situation becomes even worse when the punishments in question are so harsh that students are turned into criminals. Across the country, the New York Times reported in early 2004, “schools are increasingly sending students into the juvenile justice systems for the sort of adolescent misbehavior that used to be handled by school administrators.”[14] Apart from the devastating effects that turning children over to the police can have on their lives, the school’s climate is curdled because administrators send the message that a student who does something wrong may be taken away in handcuffs and, in effect, exiled from the community. Here we see the reductio ad absurdum of trying to improve schools by relying on threats and fear.

There are many explanations for this deeply disturbing trend, including the loss of school-based mental health services due to budget cuts. But Mark Soler of the Youth Law Center, a public interest group that protects at-risk children, observes that these days “zero tolerance is fed less by fear of crime and more by high-stakes testing. Principals want to get rid of kids they perceive as trouble” because doing so may have an advantageous effect on their school’s overall test results.[15]

School safety is at risk, that is, not merely because some educators wrongly believe that stricter or more consistent application of punitive discipline will help, but because of the pressure to raise test scores. What’s more, that same pressure, which leads some people to regard students in trouble as disposable commodities, also has the effect of squeezing out efforts to help them avoid getting into trouble in the first place. Programs to promote conflict resolution and to address bullying and other sorts of violence are being eliminated because educators are themselves being bullied into focusing on standardized test results to the exclusion of everything else. Scott Poland, a school psychologist and expert in crisis intervention, writes: “School principals have told me that they would like to devote curriculum time to topics such as managing anger, violence prevention and learning to get along with others regardless of race and ethnicity, but . . . [they are] under tremendous pressure to raise academic scores on the state accountability test.”[16]

Thus, argues Margaret McKenna, the president of Lesley University, “Some of the most important lessons of Columbine have been all but forgotten — left behind, so to speak, in no small measure because of . . . the No Child Left Behind Act. The law’s narrow focus on yearly improvement in test scores has [made schools] . . . even less conducive to teachers’ knowing their students well.” To drive home the point that our priorities have become skewed, she observes that “the test scores at Columbine High were among the highest in Colorado.”[17]


Even in those cases where a student’s actions pose a significant risk to the safety of others, the first question for an educator should not be, “Have we used sufficient force to stamp out this threat?” but “What have we done to address the underlying issues here? How can we transform our schools into places that meet students’ needs so there is less chance that someone will be moved to lash out in fury?”

Here’s a different way to look at it: We need to stop talking primarily about creating peaceful schools, which is not a particularly ambitious or meaningful goal. Schools, after all, are completely peaceful at 3 a.m. Similarly, a classroom full of docile, unquestioning students may be peaceful, even if they aren’t learning much of value, don’t care much about one another, and would rather be someplace else. What we need to work for is the creation of schools that are peaceable – that is, committed to the value of peace and to helping students feel safe, in all senses of that word.[18]

Physical safety, the most obvious kind, has understandably been the top priority, particularly where it seems to be in short supply. But intellectual and emotional safety matter, too – in their own rights and also because they’re related to physical safety. Bullying and other violent acts are less likely to happen in a school that feels like a caring community, a place where children experience a sense of connection to one another and to adults, a place where they come to think in the plural and feel a sense of belonging. That’s the polar opposite of a school where kids are picked on for being different or uncool, to the point that they fear entering certain hallways or sections of the cafeteria. Caring school communities don’t let that happen: They regard any evidence of nasty cliques or hurtful exclusion as serious problems to be addressed. They do everything possible so that no one fears being laughed at, picked on, or humiliated.

These efforts take place in individual classrooms and also as a matter of school policy. Proactive efforts to build community and resolve conflicts are important, but so too must educators focus on what gets in the way of safety and community. Thus, teachers not only hold class meetings on a regular basis so that students can participate in making decisions; they also use these meetings to address troubling things that may be going on. One teacher spoke up after a math lesson, for example, to talk with her students about something I don’t like and I don’t want to hear because it makes me feel bad, and if it makes me feel bad it probably makes someone else in here feel bad. It’s these two words. (She writes “That’s easy” on the chalkboard and draws a circle around the phrase.) . . . . When I am struggling and trying so hard, [hearing that phrase] makes me feel kind of dumb or stupid. Because I am thinking, gosh, if it’s so easy why am I having so much trouble with it? . . . . And what’s one of our rules in here? It’s to be considerate of others and their feelings.[19]

Such an intervention may be motivated not only by a general commitment to making sure that students don’t feel bad, but also by a desire to promote high-quality learning. There are intellectual costs when students don’t feel safe to take risks. A classroom where kids worry if their questions will be thought silly is a classroom where unselfconscious engagement with ideas is less likely to take place. (Of course, students often are unwilling to ask questions or acknowledge that they’re struggling for fear of the reaction from the teacher, not just from their classmates.)

On a schoolwide level, intellectual and emotional safety require that students are freed from being rated and ranked, freed from the public pressure to show how smart they are – or, even worse, how much smarter they are than everyone else. Awards assemblies and honor rolls are very effective ways to destroy the sense of safety that supports a willingness to learn. Some schools that pride themselves on their commitment to high standards and achievement have created a climate that really isn’t about learning at all – let alone about caring. Such places are more about results than about kids. Their students often feel as though they’re in a pressure cooker, where some must fail in order that others can succeed. The message students get is that other people are potential obstacles to their own success.[20]

There is much more to be said, of course, about how and why to build community, to meet kids’ needs, to create a culture of safety and caring.[21] The benefits of doing so are most pronounced in schools that have more low-income students,[22] yet such schools are often distinguished instead by punitive discipline and a climate of control. However, schools in affluent areas may also feel unsafe in various ways. Columbine High School was reportedly a place where bullying was common and a sharply stratified social structure was allowed to flourish, one in which athletes were deified. (Some of these sports stars taunted other students mercilessly “while school authorities looked the other way.”[23]) In some suburban schools, the curriculum is chock-full of rigorous Advanced Placement courses and the parking lot glitters with pricey S.U.V.s, but one doesn’t have to look hard to find students who are starving themselves, cutting themselves, or medicating themselves, as well as students who are taking out their frustrations on those who sit lower on the social food chain.

Even in a school free of weapons, children may feel unsafe and unhappy. And that’s reason enough to rethink our assumptions, redesign our policies, and redouble our commitment to creating a different kind of educational culture.


1. Mike Males, “Who’s Really Killing Our Schoolkids?” Los Angeles Times, May 31, 1999. Also see other writings by Males, including his book The Scapegoat Generation: America’s War on Adolescents (Monroe, ME: Common Courage Press, 1996).

2. The article, by Debbi Wilgoren, appeared in the Washington Post on Feburary 3, 2004, p. A-7.

3. Pedro A. Noguera, “School Safety Lessons Learned: Urban Districts Report Progress,” Education Week, May 30, 2001, p. 15.

4. This section is adapted from my article “The Limits of Teaching Skills” (Reaching Today’s Youth, Summer 1997).

5. Vicky Schreiber Dill, A Peaceable School: Creating a Culture of Non-Violence (Bloomington, IN: Phi Delta Kappa, 1997), p. 24. Also see Irwin A. Hyman and Pamela A. Snook, Dangerous Schools (San Francisco: Jossey-Bass, 1999), an excerpt from which appeared in the March 2000 issue of Phi Delta Kappan.

6. This section is adapted from my article “Constant Frustration and Occasional Violence: The Legacy of American High Schools,” (American School Board Journal, September 1999). For more on the counterproductive effects of – and some alternatives to – punitive “consequences” and rewards, see my book Beyond Discipline: From Compliance to Community (Alexandria, VA: Association for Supervision and Curriculum Development, 1996).

7. Caroline Hendrie, “In Schools, A Sigh of Relief as Tense Spring Draws to a Close,” Education Week, June 23, 1999.

8. For example, see Dirk Johnson, “Schools’ New Watchword: Zero Tolerance,” New York Times, December 1, 1999; Jesse Katz, “Taking Zero Tolerance to the Limit,” Los Angeles Times, March 1, 1998.

9. This quotation is from Robert Blum of the University of Minnesota. The study, to which he contributed, was published in the Journal of School Health and summarized in Darcia Harris Bowman, “School ‘Connectedness’ Makes for Healthier Students, Study Suggests,” Education Week, April 24, 2002, p. 16.

10. John H. Holloway, “The Dilemma of Zero Tolerance,” Educational Leadership, December 2001 / January 2002, p. 84. The analysis summarized here was published by the National Center for Education Statistics in 1998. Also see an excellent review of the effects of such policies: Russ Skiba and Reece Peterson, “The Dark Side of Zero Tolerance: Can Punishment Lead to Safe Schools?” Phi Delta Kappan, January 1999, pp. 372-76, 381-82.

11. A report by a civil rights group called The Advancement Project, based on an analysis of federal statistics, was described in Kenneth J. Cooper, “Group Finds Racial Disparity in Schools’ ‘Zero Tolerance,’” Washington Post, June 15, 2000.

12. For example, see Deborah Meier, In Schools We Trust (Boston: Beacon, 2002).

13. William Ayers and Bernadine Dohrn, “Have We Gone Overboard with Zero Tolerance?” Chicago Tribune, November 21, 1999.

14. Sara Rimer, “Unruly Students Facing Arrest, Not Detention,” New York Times, January 4, 2004, p. A-1.

15. That explanation also makes sense to Augustina Reyes of the University of Houston: “If teachers are told, ‘Your [test] scores go down, you lose your job,’ all of a sudden your values shift very quickly. Teachers think, ‘With bad kids in my class, I’ll have lower achievement on my tests, so I’ll use discretion and remove that kid.’” Both Reyes and Soler are quoted in Annette Fuentes, “Discipline and Punish,” The Nation, December 15, 2003, pp. 17-20.

16. “The Non-Hardware Side of School Safety,” NASP [National Association of School Psychologists] Communique, vol. 28, no. 6, March 2000. Poland made the same point while testifying at a Congressional hearing on school violence in March 1999 – a month before the shootings at Columbine.

17. Margaret A. McKenna, “Lessons Left Behind,” Washington Post, April 20, 2004, p. A-19.

18. The distinction between peaceful and peaceable was popularized by Bill Kreidler, who worked with Educators for Social Responsibility and wrote several books about conflict resolution. He died in 2000 at the unripe age of 48.

19. Paul Cobb, Erna Yackel, and Terry Wood, “Young Children’s Emotional Acts While Engaged in Mathematical Problem Solving.” In Affect and Mathematical Problem Solving: A New Perspective, edited by D. B. McLeod and V. M. Adams (New York: Springer-Verlag, 1989), pp. 130-31.

20. Our culture’s uncritical acceptance of the ideology of competition is such that even people who acknowledge the damaging effects of an “excessive” emphasis on winning may continue to assert that competition, per se, is inevitable or productive. If this assertion is typically unaccompanied by evidence, that’s probably because the available data support exactly the opposite position – namely, that a win/lose arrangement tends to hold us back from doing our best work and from optimal learning. I’ve reviewed some of that evidence in No Contest: The Case Against Competition (Boston: Houghton Mifflin, 1986).

21. See my article “Caring Kids: The Role of the Schools,” Phi Delta Kappan, March 1991, pp. 496-506; and chapter 7 (“The Classroom as Community”) of Beyond Discipline, op. cit. Many other writers, of course, have also addressed this question.

22. Victor Battistich, Daniel Solomon, Dong-il Kim, Marilyn Watson, and Eric Schaps, “Schools as Communities, Poverty Levels of Student Populations, and Students’ Attitudes, Motives, and Performance: A Multilevel Analysis,” American Educational Research Journal 32 (1995), pp. 627-58.

23. Lorraine Adams and Dale Russakoff, “Dissecting Columbine’s Cult of the Athlete,” Washington Post, June 12, 1999, p. A-1.

To be notified whenever a new article or blog is posted on this site, please enter your e-mail address at