Does Class Size Matter? What Decades of Research Show

Every parent knows this feeling. You find out your child’s class has 28 students. The teacher is one person. The math seems obvious: fewer kids, more attention per kid, better outcomes. So you sign the petition, you show up at the school board meeting, you push for smaller classes.

The research says you’re right — up to a point. Then it says something more complicated and more useful.

Class size reduction does produce measurable academic gains, but only when classes drop below roughly 20 students, and primarily in kindergarten through third grade. Dropping from 28 to 23 produces essentially no detectable benefit. And across multiple large-scale studies, teacher quality generates three to four times larger effects on student outcomes than class size does — at a fraction of the cost. The policy math is not what most parent advocates expect.

Understanding what the research actually shows doesn’t mean stopping advocacy. It means knowing what to advocate for.

The Small Class Intuition

The intuition that smaller classes produce better outcomes is reasonable on its face. A teacher with fewer students can give more individual attention, notice struggling learners faster, manage behavior more effectively, and personalize instruction more readily. None of this is wrong in principle.

The problem is that the size thresholds matter enormously. The intuitive benefits of individualization don’t scale linearly with class size reduction. Going from 30 students to 25 feels significant to parents but is largely invisible in outcome data. The benefits appear sharply and specifically when classes drop to 15-18 students — a threshold that is very expensive to achieve at scale and that most class size initiatives fall short of.

There’s also a substitution problem. Reducing class sizes requires hiring more teachers. More teachers means drawing from a larger pool — which, when high-quality teacher pipelines are limited, means average teacher quality may decline. This tradeoff has been documented explicitly in multiple analyses of large-scale class size reduction policies. California’s 1996 statewide class size reduction initiative, one of the most ambitious and expensive in U.S. history, produced mixed results in part because the rapid expansion required bringing in large numbers of less-experienced teachers.

Understanding this tradeoff is not a reason to oppose smaller classes. It’s a reason to think carefully about when and where the investment is worth making.

What the Research Actually Says

The most rigorous evidence base for class size effects comes from the Tennessee STAR study (Student-Teacher Achievement Ratio), which ran from 1985 to 1989 and remains the gold standard in education research on this question. Frederick Mosteller’s 1995 analysis of STAR data in The Future of Children summarized the findings: students randomly assigned to small classes of 13-17 students in kindergarten through third grade outperformed students in regular classes of 22-26 students by significant margins on standardized tests of reading and math.

Critically, the gains were lasting. Raj Chetty and colleagues at Harvard, in a 2011 paper published in The Quarterly Journal of Economics, tracked STAR participants into adulthood and found that early small-class assignment predicted higher college attendance rates and higher earnings in young adulthood — not just short-term test score improvements. The effect was concentrated in the K-3 window and was strongest for students from low-income families.

But Chetty’s paper found something else that received considerably less attention in the policy press: within the STAR study, variation in teacher quality produced substantially larger outcome differences than variation in class size. The quality of the teacher assigned to a small class mattered as much or more than the size of the class itself.

Eric Hanushek of Stanford, who has studied education production functions for decades, has been the most prominent critic of class size reduction as a policy tool. His 2003 analysis in The Economic Journal examined the cost-effectiveness of class size reduction versus alternative investments and found that the per-pupil cost of class size reduction is extremely high relative to the benefit produced — particularly when compared to investments in teacher quality. Hanushek’s estimates suggest class size reduction costs approximately $30,000 per student affected per year (in 2003 dollars) to produce gains equivalent to what moving from an average teacher to a good teacher produces for free.

Swedish research extended the STAR findings to a different educational system. Peter Fredriksson and colleagues’ 2013 study in The Quarterly Journal of Economics, using Swedish school data with a regression discontinuity design that created near-random variation in class size, found significant positive effects of smaller class size in grades 4-6 on both short-term achievement and long-term educational attainment. The Swedish findings support class size effects, but the effect sizes were modest.

Steven Rivkin, Hanushek, and John Kain’s 2005 study in Econometrica, using Texas data on millions of students, provided the most precise direct comparison of teacher quality versus class size effects. Their estimates: moving from a 15th-percentile teacher to a 85th-percentile teacher produces annual test score gains equivalent to what you’d see if class size were reduced by 10 students. Teacher quality effects dwarf class size effects by a factor of three to four.

Factor	Effect Size (SD units, approximate)	Cost per student per year	Evidence Quality
Class size reduction (K-3, to <18)	0.2–0.3 SD	Very high (~$30K+)	Strong (STAR RCT)
Class size reduction (4-8, to <20)	0.1–0.15 SD	High	Moderate (observational)
Class size reduction (>20 to 25)	~0	High	Strong null result
High vs. low teacher quality	0.6–0.9 SD per year	Low (same salary cost)	Strong (multiple studies)
Intensive early literacy intervention	0.4–0.7 SD	Moderate	Strong
High-quality tutoring (1:1 or 1:3)	0.3–0.5 SD	Moderate-high	Strong

The table is not an argument against small classes. It’s a calibration tool. The evidence clearly shows that class size reduction in early grades below the 20-student threshold produces real, lasting gains. The question is what that investment trades off against, and whether the trade is favorable for a given school or district.

What to Actually Do

Advocate for the right grade levels and size thresholds

The STAR study and subsequent replications show class size effects clearly in K-3 and under 20 students. They show essentially nothing when classes drop from 28 to 23 or 25 to 22. If your school is considering a class size initiative, push for specificity: which grades, and what target sizes?

Resources spent reducing third-grade classes from 24 to 22 students produce negligible academic returns. Those same resources spent on high-quality K-1 classes under 18 students, or on targeted tutoring for struggling readers, produce measurable returns.

This isn’t a comfortable argument to make at a school board meeting when the ask is “smaller classes for everyone.” But the evidence supports concentrating resources on early grades and meaningful size reductions rather than spreading modest reductions across all grades.

Prioritize teacher quality over class size when trade-offs arise

Rivkin, Hanushek, and Kain’s finding that teacher quality effects are three to four times larger than class size effects has direct implications for how parents should think about school quality. A highly skilled teacher in a class of 26 is almost certainly producing better outcomes than an inexperienced or ineffective teacher in a class of 18.

This is not a reason to accept large classes uncritically. It’s a reason to pay as much attention to teacher quality indicators as to class size numbers. What should parents look for?

• Teacher experience: the steepest learning curve in teaching is years 1-3; first-year teachers are, on average, meaningfully less effective than teachers with four or more years of experience
• Teacher subject knowledge, particularly in reading instruction (whether the teacher is trained in Science of Reading approaches) and math
• Instructional coherence: does the classroom have clear routines, does the teacher give explicit instruction, is student work visible and meaningful?

These are harder to observe from the outside than class size numbers, but they’re more predictive of your child’s outcome.

Understand what class size does and doesn’t explain about your child’s experience

A teacher with 28 students has less time per student than a teacher with 18 students. That’s real, and it affects both academic support and the ability to catch early learning difficulties. If your child is struggling, class size contributes to the constraint on how quickly a teacher can identify and respond.

But the solution for a struggling individual child is not primarily smaller classes — it’s targeted assessment and targeted support. Children struggling with reading, executive function, math, or attention need more than the incremental benefit of two fewer classmates. They need specific interventions. Pushing for your child to receive those interventions is a more effective use of parental advocacy energy than pushing for class size reduction for the whole school.

Ask about student-to-adult ratios, not just class size

Class size is one dimension of the adult-student resource ratio. It’s not the only one. Many schools have instructional aides, reading specialists, special education co-teachers, and paraprofessionals who are present in classrooms and can substantially affect how much individualized attention struggling students receive.

A class of 26 with a full-time aide for targeted reading support may provide more effective individualization for at-risk readers than a class of 20 with no additional adult support. When evaluating your child’s school, ask: what additional adult support is available in the classroom, particularly for students who are behind grade level?

Push for tutoring programs as a cost-effective alternative

The research on high-dosage tutoring — one-on-one or small group (one adult to two or three students) — shows effect sizes in the 0.3-0.5 SD range, which is competitive with small class size effects for a fraction of the cost of whole-class reduction. Several rigorous studies of tutoring programs, including a large-scale RCT by University of Chicago researchers, have shown that trained tutors (including nonprofessionals) working with students three to five times per week produce substantial achievement gains.

This is particularly relevant for students who are behind grade level. The marginal benefit of small class size for a student reading at grade level is smaller than the marginal benefit for a struggling student — and for that struggling student, targeted tutoring is more powerful than class size reduction.

Know when to prioritize class size advocacy regardless

The evidence doesn’t support dismissing class size as irrelevant. For K-3, for low-income students, and for students in high-poverty schools where classroom disruption and behavior management demands are higher, the case for meaningful class size reduction (to 15-18 students) is solid. The STAR study found the largest effect sizes exactly in this population. Chetty’s follow-up showed these effects persist into adult outcomes. For schools serving high proportions of low-income students in early grades, class size advocacy is well-grounded.

The policy argument isn’t that class size doesn’t matter. It’s that the effect is specific to early grades and meaningful size reductions, and that teacher quality deserves at least as much attention and investment.

What to Watch for Over the Next 3 Months

If you have a child in elementary school and are trying to understand what class size means for their specific situation, here’s a useful focus for the next quarter:

Track your child’s individual attention from the teacher. The question isn’t whether the class is large — it’s whether your child is seen. Are you receiving regular, specific feedback about your child’s progress? Does the teacher know your child’s specific strengths and challenges? A teacher with strong relationships and good classroom management in a class of 26 may be reaching your child more effectively than a struggling first-year teacher in a class of 20.

Watch for early academic difficulty and don’t wait. One genuine harm of large classes is that struggling students get missed longer. If you see signs that your child is not keeping up with grade-level expectations in reading or math, raise it explicitly and early. Don’t assume the teacher hasn’t noticed. Ask directly: “Where is my child relative to grade-level benchmarks in reading? In math? What are you seeing?” Large classes make early identification harder; parents being proactive compensates.

Watch district budget conversations. Class size is often a budget negotiation item. Understanding the evidence — that reducing third-grade classes from 26 to 22 produces no measurable benefit while hiring a reading specialist for early-grade targeted intervention does — puts you in a better position when those conversations happen.

Frequently Asked Questions

Does class size matter equally at all grades?

No. The strongest evidence for class size effects is in kindergarten through third grade. The STAR study and Chetty’s follow-up specifically show lasting effects from K-3 small class assignment. Evidence for middle and high school class size effects is weaker and less consistent. If you’re choosing where to push for smaller classes, the early grades are where the evidence is strongest.

My child’s class has 30 students. Should I be worried?

It depends on what’s happening in the classroom. A class of 30 with an experienced, skilled teacher using strong instructional practices and with access to additional adult support for struggling students may produce better outcomes than a class of 22 with a first-year teacher and no support structures. Class size is one variable among several. The more important questions are: Is my child progressing at or above grade level? Is the teacher giving me specific, useful feedback? Is my child engaged?

Why don’t schools just reduce class sizes everywhere?

Cost. Class size reduction requires more teachers, more classrooms, and more physical space. A district that reduces average class size from 25 to 20 needs 25 percent more teachers. The research on California’s 1996 statewide class size reduction initiative showed that the cost was enormous — and that because the reform required hiring so many additional teachers quickly, average teacher quality declined, partially offsetting the class size benefit.

What’s the difference between class size and student-teacher ratio?

Student-teacher ratio includes all teaching staff, including specialists, aides, and part-time teachers. Class size is the number of students in a given teacher’s classroom at a given time. A school can have a low student-teacher ratio (many adults per student in the building) but high class sizes (all students grouped in large classes during instruction). For academic outcomes, class size during direct instruction is the more relevant figure.

If teacher quality matters more, how do I evaluate teacher quality?

Teacher quality is harder to observe from the outside than class size. Useful indicators include: years of experience (effect sizes are substantially larger for teachers with four or more years of experience compared to first or second-year teachers); whether the teacher can clearly articulate what your child is working on and why; whether you’re getting specific, actionable feedback rather than general impressions; and whether the teacher’s instructional approach matches what research supports for the subject (for example, whether reading instruction includes systematic phonics).

Does any of this change if my child has a learning difference?

Yes. For children with identified learning differences — including dyslexia, ADHD, or other conditions — individual attention and targeted intervention matter more than class size per se. An individualized education program with appropriate supports and services can be more impactful than class size reduction. For these children, the question is less about how many students are in the room and more about what specific supports are in the plan.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

Sources

• Mosteller, F. (1995). The Tennessee study of class size in the early school grades. The Future of Children, 5(2), 113–127.
• Chetty, R., Friedman, J. N., Hilger, N., Saez, E., Schanzenbach, D. W., & Yagan, D. (2011). How does your kindergarten classroom affect your earnings? Evidence from Project STAR. The Quarterly Journal of Economics, 126(4), 1593–1660.
• Hanushek, E. A. (2003). The failure of input-based schooling policies. The Economic Journal, 113(485), F64–F98.
• Fredriksson, P., Öckert, B., & Oosterbeek, H. (2013). Long-term effects of class size. The Quarterly Journal of Economics, 128(1), 249–285.
• Rivkin, S. G., Hanushek, E. A., & Kain, J. F. (2005). Teachers, schools, and academic achievement. Econometrica, 73(2), 417–458.
• Bohrnstedt, G., & Stecher, B. (Eds.). (2002). What we have learned about class size reduction in California. California Department of Education.
• Nickow, A., Oreopoulos, P., & Quan, V. (2020). The impressive effects of tutoring on preK-12 learning: A systematic review and meta-analysis of the experimental evidence. NBER Working Paper 27476.