Class Size Ratios: Time to Throw Away an Outdated Idea

For years a certain unquestioned assumption has governed North Carolina public schools: smaller class sizes in the lower grades will help boost student achievement. The formal term, “class size reduction” or CSR, determines how money is allotted to Local Education Agencies (LEAs).  In grades K-3, the ratio is one teacher per 18 students and the ratio increases as grade level rises until reaching one teacher per 29 students for grades 10 through 12.

Since 2002-03, the policy has resulted in the hiring of thousands of additional teachers and school staff as well as the construction of new and temporary classrooms. An exact dollar figure is difficult to come by since teacher, staff and facility costs are incurred at both the state and local levels. However, conservative estimates place total costs for class reduction policies at over a billion dollars and growing. In the FY 2011-12 budget, lawmakers approved a provision to hire 1,100 K-3 teachers over two years at a cost of $124 million. The money will be used to reduce classroom teacher ratios for grades 1 through 3 from one teacher per 18 students to one teacher per 17 students.

It’s an expensive proposition, in part driven by the intuitive appeal of smaller class size. Small classes provide greater opportunity for interaction with students; more interaction enhances learning and higher student achievement – so the argument goes.

But does CSR work? Ever-increasing competition for public dollars, dissatisfaction over student achievement levels and new research makes this a good time to review this policy.

First, let’s acknowledge there are instances when targeted class size reduction efforts have produced positive gains in student achievement — especially for minority and disadvantaged students.

Two studies are frequently cited by CSR advocates as evidence that small classes can help to raise academic achievement.  The most famous of these is the Tennessee Student Teacher Achievement Ratio Study (STAR). The study was conducted in the mid 1980s[1]. Beginning in the kindergarten class of 1985, students were either randomly assigned to a class of 22 students or an average class of 15 students.  The study found that on average, students assigned to the smaller classes outperformed students in the larger classes. The difference in performance was equivalent to about three months more schooling. Learning benefits were largest for black and economically disadvantaged students.   Researchers also found that while the gains were evident the first year, they also decreased over time.

About a decade later, the Department of Public Instruction in Wisconsin administered a non-experimental evaluation of the Wisconsin Student Achievement Guarantee in Education (SAGE) program. The program was designed to enhance student achievement by reducing class size per teacher by up to 15 students in high poverty schools and districts. The SAGE study found students in small classrooms in grades 1 through 3 scored significantly higher on Comprehensive Test of Basic Skills than students who attended traditional classrooms. Researchers found gains for African American students were greater than those for other students.[2]

This is when the case for CSR begins to weaken. While it is true that research has shown small class size policies in certain schools or districts may boost scores for disadvantaged populations, does it necessarily follow that implementing CSR policies on a statewide basis will produce uniformly beneficial results? Because students and teachers come to school from different backgrounds and with different skills and aptitudes, isn’t unreasonable to expect a simple, targeted, policy to benefit all students?

That seems to be the conclusion of four major studies on the subject. In an analysis of Connecticut school children in the 1980s and 1990s, Caroline Hoxby studied natural class size variation and found no relationship between class size and achievement.[3]

Matthew Chingos of Brown University examined the impacts of a 2002 amendment to the Florida state constitution which set limits on the number of students in core courses. Between 2004 and 2009, Florida public schools spent about $20 billion to implement class size reduction policies. Chingos found no evidence that students from smaller classes in grades 3 through 8 did better on state math and reading assessment.[4]

Still other research casts further doubts about the benefits of statewide classroom reduction plans. In the late 1990s, California reduced K-3 class size by about 10 students (on average, from 30 to 20 students per classroom). The most rigorous study of the California Classroom Reduction policy found while the program had modest improvements in learning (6 to 11 percent per year), the modest gains were offset by the need to obtain additional classroom facilities and to hire thousands of extra teachers, many of which lacked certification and experience.[5]

Finally, after comparing the test outcomes of a nationally representative database of eighth grade students who had attended different sized classes in different subjects, Thomas Dee and Martin West found, while there was indeed a small positive effect on student noncognitive skills and attitudes about learning, class size had no overall impact on student test scores. [6]

So what has been the general impact of CSR policies in North Carolina? While this is not the place for a full evaluation of outcomes, the general trend lines are clear. Over the last ten years, all student groups – including minorities – have had declining percentages of students who passed both the reading and math tests.  In 2001, the percentage of Black, Hispanic and Economically Disadvantaged students who passed both reading and math End-of-Grade tests was 56, 62 and 59 percent respectively. In 2010-11, the scores of the same groups are 48, 54 and 53 percent. Granted, EOG tests have been revised and the changes impact the validity of long term comparisons. However, if small class reduction policies were having an impact statewide, isn’t it safe to assume that year-to-year improvements would be noticeable? Wouldn’t improvements be noticeable in the achievement gap between the races and lower [bb1] graduation rates among minority populations?

While the evidence suggests CSR fails to deliver on a large scale, there are other problems. As Professor Matthew Chingos has astutely pointed out, when teaching is driven by ratios and not tasks, the ratio ends up determining how money is spent and ultimately what is done and how it is done.

In the end, CSR is a “one-size fits all” policy for students as well as teachers. Students and teachers are treated as interchangeable entities. CSR policies falsely assume that all teachers can effectively teach to students with differing abilities and backgrounds. In doing so, CSR denies what we know about teachers and students.

Could it be that we’ve chosen the wrong route to boost student achievement? Could it be we’ve misdiagnosed the student achievement problem?

In recent years an emerging body of research suggests that teacher quality is the paramount variable in determining student achievement.  Various national reports and legislation, such as No Child Left Behind, echo the importance of teacher quality in influencing student achievement.  Like great artists, athletes or scientists, great teachers have the ability to inspire others. However, like most professions, most of teaching is filled with people that fill the spectrum from good to not-so-good. There is variation in teacher quality even within schools and its impact on student achievement can be startling.

Earlier this year, Eric Hanashek of Stanford University calculated the economic value of effective teaching and its estimated impact on future earnings. Using a creative analysis in which he estimates the economic impacts of effective and ineffective teachers on lifetime earnings, Hanashek found that if the U.S. replaced 5 to 8 percent of all teachers with average teachers, the gains in student achievement levels would boost student achievement levels to that of Canada. In addition, if the U.S. replaced 7 to 12 percent of the least effective teachers with those of average effectiveness, the change would “move the United States to the level of the highest-performing countries in the world, such as Finland.” [7]

There is no denying that small classes in the lower grades can have beneficial impacts for certain students — primarily from disadvantaged or minority backgrounds.    However, there is little evidence to suggest that such programs sustain benefits over time or that when replicated on a large scale, produce uniform benefits for all students.

Boosting student achievement is a goal we can all agree on but simply adding more teachers to maintain small classrooms in most cases is expensive and ineffective. Instead, we need to focus on cultivating quality teachers. Enhancing teacher quality through improved recruitment, retention and compensation policies offers students, taxpayers and parents a proven and more economical route to enhancing student achievement. The task is too important to do otherwise.

Teacher Allotment Ratios and Allotted Positions  FY 2010-11
2010-11 Grade Spans	Teacher Allotment Ratio	Positions Allotted in 2010-11
K-3	1:18	25,100
4-6	1:22	15,533
7-8	1:21	10,391
9	1:24.5	5,163
10-12	1:26.64	11,134

Source: Highlights of the North Carolina Public School Budget,
2011. North Carolina  Department of Public Instruction.