Wait, 20 is the perfect amount for a classroom?

Chris McNutt
December 28, 2018
Outside of the obvious logistic and financial reasoning for larger classrooms, I just couldn’t wrap my head around why we would ever want 20 students as a “great number.”

Modern policy makers point to similar claims: around 20 students is the target of a classroom. I’m perplexed by that number. Since I became a teacher, there is no denying that smaller classes are easier. If I had the option, I would never take 20 students over 15, or 25 over 20. The smallest class I ever had, 8, was by-far one of the best learning communities I developed — which consisted of at-risk students who flourished in a small-scale environment and one-on-one instruction. Outside of the obvious logistic and financial reasoning for larger classrooms, I just couldn’t wrap my head around why we would ever want 20 students as a “great number.”

Education in the United States faces an interesting conundrum: each student per class generates additional income and reduces the number of teachers — increasing the operating budget and allowing for potential better programming options. But what is the cost? The average American classroom has 24 students. Is that too many? Too little? After all, Japan and Korea are well-regarded as high achieving, at least on testing, and their average is a whopping 31 to a room (Snyder, 1993).

A Look Back

In the 1800s, class sizes varied dramatically depending on where one lived. Perhaps you went to a 10 student one room schoolhouse in the West? Or a 33 student classroom in one of the new public schools on the Eastern seaboard? Regardless, as the Industrial Revolution and compulsory schooling took hold, classroom sizes increased as would make sense. Between 1900 and 1930, enrollment of 5 to 19 year olds increased from 52 to 72% of the population, and school boards were concerned at the burgeoning costs of hiring more teachers and building new facilities without any change in funding (Rockoff, n.d.). Early classrooms of the late 1800s and early 1900s had an average of 34 to 37 students to a teacher (Snyder, 1993).

This data was commissioned because school administrators faced the same fears we have today: would increasing the size of classrooms cause reduced achievement? At the time, the answer was a resounding “No.” Research Dora Smith in 1931 concluded,

“‘Experiments following each other in rapid succession from 1920 to the present time have proved to us that, so far as the average measurable achievement of our pupils is concerned, our fears for their progress in large classes are unfounded.’”

And as a result in 1932, the North Central Association of Colleges and Schools — which had provided accreditation for educational facilities across many states — decided that the former limit of 20 students to a room was no longer needed. There was no longer any stipulation on class size (Rockoff, n.d.).

Of course, these giant classes weren’t without critics. As educators bore witness to an onslaught of new students, they questioned how research was gathered. Sure enough, similar to the researchers of today, their definition of achievement was solely standardized test performance — which continues to ignore many principles of what it means to be educated. This debate didn’t amount to much though, as the 1920s and Great Depression saw a drastic decrease in birth rates.

However, the Baby Boom generation was coming — and class sizes rose again, being barred at between 20 and 30 students, seemingly without much data to support these sizes (Rockoff, n.d.). This seems to be the norm since the 1960s. Slight fluctuations have occurred, but average class sizes in the United States have consistently hovered around that number.

Examining the Research

This is not just for class size, but really all educational research: we have to be very careful of how we gauge achievement. If our goal is increased — or at least sustained — standardized test scores, our data is going to be much different than if our goal is increased comfort in the classroom, or increased creative risk or curiosity. The vast majority of research pertains to academic test scores in a bubble and that matters a whole lot. If you go on our website, almost every research article we’ve chosen is not based on test scores — because if our measure of success is doing well on something that ultimately doesn’t matter and isn’t a good gauge of learning anyways — then our research isn’t based in something scientific. It’s an overarching problem that’s pushing a traditional system in all aspects of education and something we must look for whenever people present grand claims on what works and what doesn’t. That being said, much of the research on class size surrounds standardized testing and without better options, this is what we’re left with — so take everything with a grain of salt.

Anyways, here are some of the landmark studies surrounding class size:


By far the most commonly brought up study is the Student Teacher Achievement Ratio, or STAR, which was performed in Tennessee in the 1980s. Students were assigned to either a 15 person or 22 person class, and were measured via testing for students in smaller classrooms to have a 3 month gain in academic knowledge over their peers in a 4 year window. Similar work was done in Texas, California, Florida, Connecticut, and other countries with moderate, mixed, and “no effect” results (Chingos & Whitehurst, 2011.)

These studies are a mixed bag for a variety of reasons: perhaps the style of teaching didn’t change when class sizes were reduced. If I keep my class dead silent, individualized, and have students respond to book prompts, there really isn’t a drastic shift in results — but what if I’m the type of teachers who values interaction and constantly builds rapport? The results would definitely shift. There’s also factors of where the school is and what grade is studied: younger students benefited more obviously from smaller class size, as well as those from less advantaged backgrounds (Chingos & Whitehurst, 2011.)
In addition, classroom size studies are hard to interpret. In wealthier areas, there tend to be more employed teachers leading to smaller class sizes — and we have to be sure that better student outcomes are specifically a result of class size, or maybe it’s due to other privileged factors. However, specifically in the STAR study, students were placed in classrooms at random within a district with an attempt of solving this issue.

And there were studies that built off the ideas of this program. One was a data analysis by Alan Krueger in 1999 that looked at 11,600 students. He found that in one year, those in smaller classrooms (which was marked as 13–17 students) achieved 4% on standardized testing and continued to improve compared to their peers at 1% per year. Notably, teacher aides or specific practices by teachers in these rooms seemed to have no noticed effect — as in, no matter how a teacher managed their room or even if another adult was present, class size mattered. In addition, minority students and those on free and reduced lunch had even greater gains than their peers (Krueger, 1999.)

Raj Chetty and Lifelong Implications of STAR

In another, and likely one of the most interesting I’ve seen, study in 2010 — Raj Chetty and colleagues examined the lifelong implications — such as earnings, college attendance, and home ownership — of those involved in the STAR program. Essentially they sought an answer to the question: “Do small classrooms sizes change life outcomes?” Remember, the STAR program was randomized so this isn’t really a question of background and privilege. They found that children from small classrooms were more likely to attend college, own homes, be married, and have greater retirement funds. In addition, those with smaller class sizes and an experienced teacher, defined by 10 years in practice, earned higher salaries.

Another tidbit from this study was that researchers were confused on why only elementary students performed better on standardized tests with small classes, yet into adulthood these students were still performing better on their achievement measures than their peers. They noted that “higher quality kindergarten classrooms may build non-cognitive skills that have returns on the labor market but do not improve performance on standardized tests.” Do note that this data isn’t conclusive, but it is worthwhile to consider that structuring our classrooms for the success of our students beyond standardized testing is implied to actually make student lives better — which makes a lot of sense and it’s nice to note there is some data here (Chetty et. al., 2011).

Reducing Class Size Without Preparation

Another study resulted from a legislative push in 1996 in California to reduce class sizes from 30 to 20 students maximum. Researchers Christopher Jepsen and Steven Rivkin examined what happened from this shift. In this study, they found that students in small classrooms had greater math and reading scores as a whole, but some students struggled more. You see — in response to adding more classrooms, California needed a lot of new teachers, so they hired people without any formal training or even certification. Students placed in these classrooms had extreme difficulty, especially at-risk students. This is a concerning practice to consider when we talk about reducing classroom size, especially for the implications it has for programs like Teach for America, which are situationally good ideas, but often lead to misqualified people with our children (Jepsen & Rivkin, 1996.)

As a quick aside, there is an obvious financial issue to maintain small class sizes. If every classroom in the United States had one fewer student, it would cost the United States $12 billion, and the reverse is true as well (Harris, 2009.) I try to stay clear of multifaceted political leanings, but to be clear, by cutting the US military budget by $120 billion (18%), we would have an average classroom size of 14. In my opinion, that shift would be an incredible sight to behold.

So what?

It is difficult to prescribe anything in education without tackling multiple issues at once. We can’t simply fund schools more, lower class sizes, and expect a transformative change. Although there would be better learning outcomes, according to research as well as anecdotal evidence from most teachers I’ve worked with, it wouldn’t be as substantial as lower class sizes with new opportunities. And we must be sure to train well-prepared educators, experts in their field, to meet these demands. But imagine an individualized, project-centered classroom — a small learning community where students had the resources, time, energy, and guidance to find out what matters to them. A place where a teacher had the time to seriously get to know a small group of students and push them to succeed. Imagine if higher education rejected the use of 100+ attendee lecture halls in return for individualized research opportunities?

In the same regard, we must be careful of how research results are determined and shared. It is easy for any of us to come to a simple conclusion that fits our narrative, and these findings are one in the same. However, that does not conclude that class sizes are irrelevant — as is shared by traditionalists. As shown above, our way of measurement is often misleading.

I wonder if the populous desire of Americans will lead to greater organization and pushback for our schools. In an era desperate for anti-corruption, sustainable policies that reflect the lives of average people — when teachers are organizing for better pay and respect — will our inevitable next step be fighting for our children’s futures? Professional educators continue to bolster their progressive claims with data and their students earnest respect for their classes. But Internet fame and book sales aren’t enough for radical change. In the vein of Paulo Freire, it makes sense for us to reach out to coworkers — envision a better tomorrow for our classrooms — and fight to make that happen. Without a barricade of respectable teachers on the frontlines, policies which damage education will always creep their way into the forefront. It’s up to us to make a difference and use our knowledge to change the world.

Chris McNutt
Chris McNutt is the co-founder and executive director of Human Restoration Project, a nonprofit organization focused on student engagement, well-being, and motivation. His work centers on realizing systems-based change, examining how progressive pedagogical shifts (e.g. PBL, ungrading) reimagine school to best suit the needs of students and teachers alike. He was a public high school digital media & design educator who focused on experiential learning, portfolio-driven assessment, and community involvement.
The YouTube symbol. (A play button.)

watch now