For this testimony as a word document with footnotes, click here.
Testimony before the New York City Council’s Commission on CFE
On the benefits of smaller classes for NYC public schools
October 19, 2004
By Leonie Haimson, Class Size Matters
The research is abundantly clear that class size makes a critical difference in terms of student achievement, particularly among poor and minority students. In 1985, the best-designed, large-scale experiment ever carried out in the history of education was launched: the STAR study in Tennessee.
Over four years, thousands of students were randomly assigned to either small or large classes in grades K-3rd. Those who were assigned to smaller classes did better in every way that could be measured: they received higher test scores and better grades, they had improved attendance records, fewer were held back, and they exhibited much fewer disciplinary problems.
Two other results were even more striking: The students whose performance improved the most were those from poor and minority backgrounds. Recently, Alan Krueger, economist at Princeton, re-analyzed the STAR data, and found that smaller classes in the early grades narrowed the achievement gap between racial and ethnic groups by about 38%..
The other extremely compelling result of the STAR study was that many of these gains lasted long after 3rd grade, when these students were put back into the general population. In 4th, 6th, and 8th grade, students were significantly ahead of their regular-class peers in all subjects,. By the time they reached high school, they had better grades, higher graduation rates, and higher scores on their college entrance exams. For those who attended a smaller class in grades K-3, the difference in rates between black and white students headed towards college was cut in half.
Many of the achievement gains first noted in Tennessee in the STAR study have since been replicated in studies carried out in Wisconsin, California, and elsewhere. (See attached fact sheet, “Class size reduction in California”).
Large-scale studies have also confirmed the importance of class size. David Grissmer at RAND examined student achievement in 44 states, as reflected in test scores on the national exams known as the NAEPs (National Assessment for Education Progress). He concluded that those states that had the lowest class size in the early grades were also those highest levels of student achievement, particularly among poor and minority children.
Although there has been no experimental study carried out on class size reduction in the middle and upper grades similar to the STAR study, there are at least eight studies that show that smaller classes in grades 4-12 are correlated with higher student success and achievement rates. (See attached fact sheet, “The importance of class size in the middle and upper grades.”)
Two of these are most instructive: In an analysis of more than 200 school districts, researcher Harold Wenglinsky of the Educational Testing Service concluded that smaller classes were correlated with a significantly higher level of achievement for fourth and eighth graders. In the eighth grade, smaller classes were also associated with a much improved “school social environment”, as measured by factors such as better attendance and less school vandalism.
An authoritative study put out by the US Department of Education looked at the achievement levels of students in 2,561 schools across the nation, as measured by their performance on the NAEP exams. The data included at least 50 schools in each state, including those from large and small, urban and rural, affluent and poor areas.
After controlling for student background, the only objective factor that was found to be correlated with higher student success was class size, not school size, not teacher qualifications, nor any other variable that the researchers could identify. What was perhaps even more striking is that these achievement gains were more strongly correlated with small classes in the upper rather than the lower grades.
Other proven benefits of smaller classes
Smaller classes have also been shown to have a host of other benefits. (See attached fact sheet, “Non-academic benefits of class size.”) Researchers in Wisconsin, California and New York have found that reduced class size leads to more parent volunteers in the classroom, and more parental involvement overall. When I interviewed teachers about this during the first year of class size reduction in New York City schools, they explained that this results from their ability to keep in closer communication with parents about their children’s educational progress and needs. Because of the overly large classes in the New York City public schools, often parents only have five to ten minutes with their child’s teacher on parent-teacher night. When classes are smaller, teachers explained to me, that time can be extended to fifteen minutes or more. Moreover, when classes are smaller, they told me that they are better able to reach out to parents not just when their child is doing badly, but also when he or she has excelled as well.
As the ETS study suggests, behavior problems are also significantly reduced in smaller classes. In California, students were significantly less disruptive. In Burke County, North Carolina, disciplinary problems and disruptions declined by more than 25% after class size reduction occurred. In New York City, a principal in East Harlem told me that disciplinary referrals had dropped 60% in her school nearly overnight, the first year they instituted smaller classes in the early grades. This same school, PS 198, was subsequently profiled in an article in the NY Daily News, as having increased the number of students reading at grade level from 23% to 69% four years later - a turnaround that its principal attributed specifically to smaller classes at her school. (See attached Daily News article, “Fearing gov's math: Adds up to return of larger classes, parents worry,” February 13, 2003.)
Since smaller classes foster greater academic and social engagement, they also lead to lower dropout rates. A statistical analysis using data from the National Center for Education Statistics reveals that in school districts across the country, graduation rates are highest where student-teacher ratios were lowest, with the gains particularly sharp for among minority students.
Smaller classes and teacher quality
When teachers are surveyed, they respond that the most effective way to improve the quality of teaching would be to lower their class sizes, far above all other methods, including raising salaries or providing more professional development.
In a national survey, principals also responded that reducing class size would be the best way to improve the quality of teaching, over increasing teacher salaries or introducing merit pay.
Smaller classes go hand in hand with enhanced teacher quality in several ways. First of all, educators are able to give more individualized instruction and feedback to their students in smaller classes, but they themselves are given continual feedback as to how their own methods of instruction are working, allowing them to adjust their classroom techniques over the course of the year to best keep their students focused and engaged.
In small classes, students learn better because less time is spent on discipline and classroom management. Overly large classes are frequently cited by teachers as one of their major impediments to success. In a national survey, 62% of teachers say that overcrowding and large class sizes are an important contributor to discipline problems in schools.
Yet another way in which smaller classes lead to more effective teaching relates to the fact that in large classes, the racial disparity between teacher and student can contribute to a wider achievement gap.
A recent re-analysis of the Tennessee STAR data showed that both white and black students did better when randomly assigned to a teacher of their own race. According to this study, students performed better every year they were taught a same-race teacher, and these benefits were cumulative; others feel further behind every year they had a teacher of a different race. Yet for those students who were assigned to a smaller class, the race of their teacher no longer mattered; and all of them learned more, no matter what their teachers’ racial background. (See attached article, Michael Winerip, “Good Teachers + Small Classes = Quality Education,”The New York Times, May 26, 2004.) One can speculate as to why this might be; it may be that in a smaller class, stereotyping on the part of both teacher and student is less likely to occur. Also, given fewer students, teachers may no longer be forced to ration their time by focusing on certain students rather than others.
Smaller classes will help to improve the supply of qualified, experienced teachers, particularly in high-needs schools, by making it less likely that teachers leave the profession after only a few years, or move to other school districts where working conditions are easier. In California, class size reduction led to significantly improved rates of teacher retention statewide.
And though initially, there were reports in California of teachers migrating from urban schools to teach elsewhere as new teaching slots opened up, follow-up studies showed “virtually no differences in teacher migration rates between urban and non-urban districts” as a result of the program. Instead, migration rates of teachers in dropped after class sizes were reduced, most sharply in high-poverty schools, to levels much lower than before the program began. (See attached chart: Teacher migration rates following class size reduction in California.) This is not surprising, as smaller classes increased teachers’ sense of satisfaction and chance of success, especially in high-needs schools where the need for individualized instruction is greatest.
Here in New York City, we have a problem keeping teachers, though no problem recruiting them. The Mayor recently testified before the CFE special masters panel that we had 75,000 qualified applicants for 6,000 open positions last spring.
Yet like other school districts that serve a primarily poor and minority student population, we do have a crisis in teacher retention. The two-year attrition rate for new teachers is 25%, with 18% of them departing in their first year, compared to a national rate of ten percent.
A survey from the New York City Council Investigation Division revealed that more than 25% of mid-career teachers (with 6-24 years experience) and nearly 30% of newer teachers (with 1-5 years experience) report that they will likely leave the city public school system within the next three years. For these two groups, providing smaller classes is one of the top three changes that they say would cause them to consider remaining longer in their jobs.
Expected cost savings through smaller classes
Though admittedly expensive, class size reduction is likely to bring a host of savings to our schools, in both human and economic terms. Alan Krueger, professor at Princeton and former chief economist of the US Labor Department, has calculated that reducing class size leads to economic benefits that are twice as high as the costs, in terms of the future earnings potential of graduates alone, given their higher rates of achievement. His estimation is conservative, and did not include many of the other ancillary benefits and cost-savings that smaller classes are likely to bring to the system.
For example, if smaller classes raise our rate of teacher retention, this will provide significant savings in terms of recruitment, training and mentoring costs. The Teaching Fellows program is especially expensive, costing more than $25,000 per recruit, and yet the attrition rate is even higher among Fellows than new teachers as a whole.
There will also likely be considerable savings in terms of referrals to special education programs. In the New York City Chancellor’s District, composed of low-achieving elementary schools where class sizes in all grades were reduced to 20 or less, not only did test scores rise significantly, but special education referrals also declined significantly. Just last week, a parent informed me that when his third grade child had entered a large class for the first time this fall, he was diagnosed with a “multiple sensory disorder” that will probably cause him to be referred to a special private school that may cost the system $40,000 or more. Yet this parent believed that in a smaller class, his child’s needs could have been met at much lower cost, and that might never had been diagnosed with this condition at all.
Class size reduction as well as the elimination of overcrowding would also substantially lower the costs of policing our schools, including the need to install multiple safety officers. Even larger savings would come in the area of remedial education, grade retention, and the multiple years it now takes our students to graduate. Every year a child is held back means another year in education costs, of about $11,000 -$12,000 per year. Moreover, thousands of our students now take five, six or even seven years to graduate. If we could reduce these rates substantially, as would likely occur if we provided smaller classes to these students, this would mean millions saved to the system as a whole.
Yet the most costly aspect, both in human and economic terms, of our inadequate education system is the fact that 50% or more of our students never graduate from high school at all. Is it any wonder that, as a recent report from the Community Service Society revealed, in 2003 barely one-half (51.8 percent) of New York’s black men were employed? Given the increase in funding we are likely to receive from the state, it is not merely our opportunity to give our children a chance to succeed by providing them with smaller classes, I would argue that it our duty.
