“Zero Tolerance.” Causes, Consequences, and Alternatives

Zero Tolerance policies, as practiced in school districts around the country, are now widely ridiculed and condemned. Rightfully so. Thoughtless and indiscriminate application of such policies has led to members of high school baseball teams being expelled for having baseball bats in their trunks, and to five‐​year‐​olds being expelled from Kindergarten for making handgestures in the shape of guns.¹

As disciplinary referrals to the principal’s office have increased, so has the frequency of outof‐ school suspensions. You will hear from other speakers today the harm that these can do to the suspended student. Fortunately, there are much better discipline policies available to us, and I would like to begin my testimony by describing one such hypothetical alternative:

Imagine a school district that resolved not to expel students or use out‐​of‐​school suspensions. Instead, let’s say it vigorously and consistently enforced a clear code of conduct, giving detentions for small violations and in‐​school-suspensions for more serious transgressions like starting fights.

These in‐​school‐​suspensions would assign a host of duties intended to discourage repeat offenses and encourage civilized behavior. Suspended students might write reflective essays about their behavior and why it was inappropriate. They could be assigned clean‐​up duties around the school. They could also be required to write a letter of apology to their fellow classmates, teacher, and principal, and this letter could be read out to the class or even at a school assembly.

Since disruptive students are often behind academically, they could be required to attend Saturday morning classes to help them catch up. And as a way of illustrating that their behavior was beneath the standards expected of students of their age, they might be assigned to a different class at a lower grade level during the period of their in‐​school suspension, and required to do all the work assigned in that class.

The interesting thing about this hypothetical school district is that it is not hypothetical and it is not a school district. The policies I’ve just described are those that have been in place for a decade at the American Indian Model charter schools, often abbreviated as “AIM” schools. The name of this small network of three charter schools is vestigial: the student body today is primarily, South‐​East Asian, Hispanic, and black. Virtually all the students qualify for free or reduced price lunches, and the schools are all located in the heart of Oakland, California, one of the most violent and crime‐​ridden cities in America.

Oakland’s Public School district has its own armed police force, one of its elementary schools suspended 97 students for acts of violence in a single recent year,² and that isn’t the district’s worst school. It’s typical for multiple students to be shot each year, with several fatalities. Bullying and fights on school grounds are a daily occurrence in the district.

But the American Indian Charter Schools are different. I’ve visited them, interviewed their students, teachers, and administrators, and studied their academic achievement.³ This is what I’ve found:

The atmosphere at these schools is orderly and studious. Attendance rates are around 99 percent. There are no metal detectors and no on‐​campus police. Violence is almost unheard of. The average number of fights across all three schools combined is about 3 per year. Sixth grade teachers—those teaching students experiencing the American Indian Model for the first time— hand out detentions for any behavior that disrupts the class. Talking with those younger middleschool students, it’s obvious that many of them chafe at the relentlessness of the schools’ discipline policy. They’re kids. It’s natural for them to push the boundaries of what’s acceptable. But those same students who roll their eyes at all the detentions their middle school hands out, are quick to report how different their school is from the district schools they recently left behind, and which many of their friends still attend. They tell stories, in shocked and dismayed voices, of the bullying, fighting, and drug‐​dealing that routinely go on in the district schools, and they are very happy that these things are incredibly rare at their AIM school.

By the time they reach high school, AIM students not only behave with great maturity, they have excellent study habits and skills. Teachers at the high school level spend virtually no time on discipline. Because they don’t have to. They spend all their time teaching. The students are self‐​motivated, and proud of their academic success. And they are very academically successful. When I studied the performance of all of California’s 68 charter school networks last year, I found that AIM schools were the highest performing by a wide margin. AIM students are just as far ahead of students at the well‐​respected KIPP charter schools as KIPP students are ahead of students at regular public schools. That is after controlling for the race and socio‐​economic status of the students, as well as peer effects. In fact, low‐​income Hispanic and African American students at AIM schools outperform the state wide average for wealthier white and Asian students.

Their entire graduating classes are generally accepted to multiple 4 year colleges, often quite prestigious ones. I’ve interviewed AIM school alumni currently attending or having graduated from colleges like Berkeley and Dartmouth, I know a number of others are currently enrolled at Stanford. The AIM model shows that it is possible to design discipline policies vastly better for students than the cavalier use of expulsions and out‐​of‐​school suspensions. What’s more, there are ways of systematically encouraging the adoption of similarly effective discipline practices. And I will discuss those in a moment.

But it is also painfully clear that we are not there yet. Today, policies like the ones in use at the AIM schools are rare exceptions. So if of out‐​of‐​school suspensions were curtailed tomorrow in districts like Oakland, they would not be instantly replaced with highly effective alternatives. Knowing that, it is crucial to ask: what would they be replaced with? What would happen if principals facing extensive discipline problems in conventional public schools suddenly curtailed their use of out‐​of‐​school suspensions?

That’s not a rhetorical question. In fact, it has already been answered in a forthcoming study in the journal International Economics Review.⁴ In that study, Rochester University professor Joshua Kinsler discovered that cutting out‐​of‐​school suspensions in schools with many disruptive students lowers overall student achievement.

Why is that? As we know, out of school suspensions do no good for the suspended student academically, but Kinsler found that they do appear to benefit the rest of the school, presumably by making it easier for teachers to teach the non‐​disruptive children.

Professor Kinsler’s findings reminded me of an essay I came across recently, dealing with school violence. It reflects on bullying suffered by the author when he was a boy, and how it was dealt with by his school. I’d like to share a brief quote with you:

These reflections were written by Judge Teske. He makes an eloquent case that adults in our school and justice systems must defend innocent children from bullies. His argument is compelling, and it applies just as much to children’s education as to their physical safety.

Yes, out‐​of‐​school suspensions are far from the ideal disciplinary strategy. But until superior strategies, like those of the AIM model, have been widely adopted, curtailing out‐​of‐​school suspensions will likely have the perverse result of compromising the education of millions of innocent children.

There is a bitter irony here. A key concern with Zero Tolerance policies is the harm they do to African American children, because African American students are more likely to be referred to the principal’s office and, as a result, more likely to be suspended. But only a small fraction of black students are actually suspended. The vast majority are not disruptive. They are simply trying to get an education. They are, like the victims of bullying described by Judge Teske, innocent. What Kinsler’s research shows, is that in public schools with discipline problems, it hurts those innocent African American children academically to keep disruptive students in the classroom. According to Kinsler’s findings, significantly cutting out‐​of‐​school suspensions in those schools widens the black‐​white academic achievement gap.

Clearly we must find a better solution. The existence of highly successful disciplinary and academic models like the American Indian charter schools in Oakland proves that we can do better. The challenge is to figure out why such successful models are so rare today, and how we can replicate them.

Those are tough questions, and there isn’t the time or space in this testimony to do them justice, but let me share two points that I think help to point the way forward. First, ask yourselves why the nation’s public schools have so widely adopted such badly‐​designed Zero Tolerance policies? An especially clear answer to that question comes from an Associated Press story from 2001, back when public school officials still vividly remembered the years before Zero Tolerance became widespread. Let me quote to you briefly from that article:

Today, Zero Tolerance policies are faulted for applying discipline rules blindly and mechanically, with no consideration for extenuating circumstances. But that is precisely why those policies were adopted in the first place. Prior to their adoption, education officials at every level of government were inundated with lawsuits and complaints of disciplinary bias. Elected officials were pressured to do something. Officials sought to reduce this flood of lawsuits and complaints by automatically ejecting students for violating the letter of a Zero Tolerance policy.

In short, they adopted these policies because it seemed in their own interests to do so—not because they thought it was in the interests of students. I don’t say that to fault these officials. They were people just like the rest of us, and they were influenced by the incentives of their workplace, just as we all are. It would be unrealistic to expect otherwise. If we want better policies to be adopted, we have to change the incentives in the system. For instance, consider a system in which administrators who keep more students in school, maintain orderly classrooms, and achieve higher graduation rates are recognized and rewarded for their achievements. What if administrators’ and teachers’ job security and pay were tied to these desirable outcomes?

My second observation on the way forward begins with a question. What kind of school is most likely to implement a successful discipline policy? As you can imagine, a lot of factors are involved, but there’s good evidence that a cornerstone of these successful schools is consistency. When students understand that the expectations for their behavior are the same from grade to grade and from classroom to classroom, that everyone in the school is on the same page, it has been shown to lead to more studious, orderly schools.⁷ Consistency is a hallmark of discipline policy at the American Indian Model schools.

Wonderful as it is to know that, it begs the further question: how do you cultivate such consistency and build a strong sense of shared mission and understanding among school staff. There’s actually good evidence on this question as well, reaching back decades. One study, published in the journal Sociology of Education, compared the attitudes of teachers in two different groups of schools.⁸ Their results are reported in Figure 1, and reveal that teachers in Group A schools were substantially more positive about their school’s atmosphere in every one of the nine areas measured.

Essentially identical results were found by Harvard sociologist Susan Moore Johnson, who studied a different set of schools, but also ones falling in Group A or Group B.⁹ Professor Johnson reported that teachers in Group A schools consistently “expressed clearer notions of their schools’ goals and purposes; they identified the values that they shared with others in the schools’ they explained how these understandings were grounded in their schools’ histories and were reinforced and expressed in their traditions.” Group B teachers, she found, were “offend perplexed” by the same questions about their schools’ culture and values. A typical Group B teacher responded that his school “probably does have some unifying culture and I’m just not aware of it.” Apparently not seeing the irony in that statement.

Here’s how professor Johnson summed up her findings:

Neither of the soociological studies I’ve mentioned attributed the differences between Grroup A and Grroup B schools to the students they served. Instead, they attributed them to the organizational structure of those schools. As some of you have no doubt already guessed, what I’ve called Group A and Group B are in fact private and public schools.

For decades, education economists have reported findings consistent with those of their sociologist colleagues. Controlling for student and family background, graduation and college acceptance rates are higher in independent schools than in public schools, whereas crime rates are lower—especially for urban African American students.¹⁰ The District of Columbia’s own school voucher program, overseen by Congress, has a significantly higher graduation rate than the district’s vastly‐​higher‐​spending public schools.

From a policy standpoint, these findings are problematic. Under our present system, the people with the least access to independent schools are low‐​income families—precisely those who are more likely to live in higher crime neighborhoods with troubled public schools; the very people most desperately in need of better, safer alternatives.

I do not present this evidence to encourage Congress to enact nation‐​wide private school choice legislation. Even if the Constitution permitted such a program, which it does not, evidence from other nations suggests it would be more effective to implement such policies at the state level. But Congress can encourage the adoption of such state‐​level programs by virtue of the public prominence of its Senators and Representatives. Congress can also nurture and expand the DC Opportunity Scholarship program, as an example to states of what is possible. And above all, Congress can avoid instituting new regulations and programs that would impede states’ efforts to bring safe, responsive independent schools within reach of all children, and can discontinue federal programs that have proven themselves ineffective and that consume funds that could more effectively be spent by the states and the people.

Notes

¹http://​www​.kndu​.com/​g​l​o​b​a​l​/​s​t​o​r​y​.​a​s​p​?​s​=​1​1​9​79866
²http://​www​.sfgate​.com/​b​a​y​a​r​e​a​/​a​r​t​i​c​l​e​/​O​a​k​l​a​n​d​-​W​h​e​n​-​s​c​h​o​o​l​-​b​u​l​l​i​e​s​-​g​e​t​-​o​u​t​-​o​f​-​h​a​n​d​-​3​2​1​6​5​7​8.php
³Andrew J. Coulson, “The Other Lottery: Are Philanthropists Backing the Best Charter Schools?” Cato Institute Policy Analysis no. 677, June 6, 2011. https://​www​.cato​.org/​p​u​b​s​/​p​a​s​/​P​A​6​7​7.pdf
⁴Joshua Kinsler, “School Discipline: A Source or Salve for the Racial Achievement Gap? ‚” International Economic Review, forthcoming. Draft: http://​www​.econ​.rochester​.edu/​p​e​o​p​l​e​/​K​i​n​s​l​e​r​/​j​k​i​n​s​l​e​r​_​s​u​s​p​e​n​s​i​o​n​s.pdf. “A rule based discipline code that requires all schools to use the same suspension policies closes the discipline gap but results in a significant widening of the achievement gap. Schools serving predominantly black students shorten suspension lengths, which reduces the deterrent effect of discipline, increases misbehavior, and decreases achievement” (p. 7). See also: Joshua Kinsler, “Understanding the black–white school discipline gap,” Economics of Education Review, vol. 30 (2011), no. 6, p. 1370–1383.
⁵Steven Teske, “In Defense of Self‐​Defense — Are Schools Safer with Zero Tolerance?” Juvenile Justice Information Exchange, JJIE​.org, February 10, 2011, http://​jjie​.org/​j​u​d​g​e​-​s​t​e​v​e​-​t​e​s​k​e​-​d​e​f​e​n​s​e​-​o​f​-​s​e​l​f​d​e​f​e​n​s​e​-​s​c​h​o​o​l​s​-​s​a​f​e​r​z​e​r​o​-​t​o​l​e​r​a​n​c​e​/9692
⁶Anne Gearan, “ABA may oppose ‘Zero Tolerance’,” Sun Journal, Lewiston Maine, February 15, 2001. Text in square brackets added. http://​news​.google​.com/​n​e​w​s​p​a​p​e​r​s​?​n​i​d​=​1​9​1​4​&​d​a​t​=​2​0​0​1​0​2​1​5​&​i​d​=​7​j​F​H​A​A​A​A​I​B​A​J​&​s​j​i​d​=​6​P​M​M​A​A​A​A​I​B​A​J​&​p​g=132 9,2369349
⁷See this overview by the U.S. Department of Education: http://​www2​.ed​.gov/​p​u​b​s​/​T​o​o​l​s​f​o​r​S​c​h​o​o​l​s​/​c​m​c​d​.html, and also: H. Jerome Freiberg, Chris A. Huzinec, and Stacey M. Templeton, “Classroom Management—a Pathway to Student Achievement: A Study of Fourteen Inner‐​City Elementary Schools,” The Elementary School Journal, vol. 110, no. 1 (September 2009), pp. 63–80.
⁸Valerie E. Lee, Robert F. Dedrick, and Julia B. Smith, “The Effect of the Social Organization of Schools on Teachers’
Efficacy and Satisfaction,” Sociology of Education, vol. 64 (1991), no. 3, p. 190–208.
⁹Susan Moore Johnson, Teachers at Work: Achieving Success in Our Schools (New York: Basic Books, 1990).
¹⁰Several of the graduation rate, college acceptance rate, and college graduation rate comparison studies are collected and tabulated in: Andrew J. Coulson, “Comparing Public, Private, and Market Schools: The International Evidence,” Journal of School Choice, vol. 3 (2009), vol. 1, p. 31–54. https://​www​.cato​.org/​p​u​b​s​/​a​r​t​i​c​l​e​s​/​c​o​u​l​s​o​n​_​c​o​m​p​a​r​i​n​g​_​p​u​b​l​i​c​_​p​r​i​v​a​t​e​_​m​a​r​k​e​t​_​s​c​h​o​o​l​s​_​j​s​c.pdf. An interesting comparative study of criminal behavior is: David N. Figlio and Jens Ludwig, “Sex, Drugs, and Catholic Schools: Private Schooling and Non‐​Market Adolescent Behaviors,” German Economic Review, 2012, vol. 13 (2012), issue 4, p. 385–415.