skip to Main Content

The optimal class size is one of the most disputed topics in education. There is considerable divergence of views amongst researchers and academics, though most teachers and parents tend to anecdotally agree that smaller is better. On the other hand, politicians and policy makers tend to take the opposing view, citing the costs of smaller class size amongst other factors. In the following paragraphs an attempt has been made to summarise some of the research and views regarding class sizes, and touch upon the views of others such as teachers, politicians and policy makers.

One of the most influential reports in recent years on class sizes and oft quoted by politicians and policy makers is a report by The Grattan Institute, authored by Ben Jensen (2010).

In that report Jensen states: One of the most enduring policy prescriptions in education is to reduce class sizes. Smaller classes are intuitively appealing. It is easy to imagine that they result in more one-on-one interaction with students, more effective teaching and learning time for each student, and a reduction in the burden of dealing with negative behaviour. Unfortunately, the evidence does not support these assertions. In fact, most studies find that despite spending significant resources on reducing class sizes, the effect on student performance is either negligible or there is no effect at all.

He goes on to cite academic research studies by Hoxby, 2000; Bohrnstedt and Stecher, 2002; Jepsen and Rivkin, 2009 and Chingos, 2010. For example, Bohrnstedt & Stecher concluded from their study of class size reduction (CSR) in California that their analyses of the relationship of CSR to student achievement was inconclusive. In this initiative, Californian class sizes in the early years of schooling were reduced.

Another example of the lack of a strong link between student attainment and class size is a meta-analysis of nearly 60 studies. In this meta-analysis, Hanushek found that a little less than 15% reported a positive and significant effect of reducing class sizes (1997; 2003).

This meta-analysis has been disputed on technical grounds. Krueger (1999) challenged the methodology of Hanushek’s meta-analysis. That said, Krueger’s revised version did not offer significantly greater support for the benefit of CSR. Krueger’s analysis showed higher levels of successful improvement of student attainment through CSR but overall only 33% of studies in Krueger’s revised meta-analysis reported significant positive results (2002; 2003).

Mishel and Rothstein, (2002) summarised both Krueger and Hanushek’s analyses citing that there was clear evidence that the majority of studies (between 66% – 85%) in both meta-analyses of the research literature showed that class size reductions had no impact, or no significant impact on student attainment.

Jensen’s (2010) conclusions state that the majority of studies examining CSR find little or no significant positive impact on students, even in the early years of schooling. Of the minority (15%-33%) of studies that find a positive impact on students, a greater proportion of these focus on the early years. However, Jensen concludes that the positive impact shows students making marginal improvement and goes on to say. <em>The evidence is clear: Class size reductions, even in the early years, are very expensive and have a negligible impact on student outcomes. …Even if there were positive outcomes, the question remains whether reducing class sizes offers good value for money.</em>

Politicians tend to cite Jensen’s report on the research of others suggesting that reducing class sizes has a significant impact on school budgets as more teachers are required to teach the greater number of smaller classes and that the larger number of teachers across a large number of school tends to reduce the overall quality of the group of teachers.

Teacher salaries comprise around 67% of all school education expenditure in Australia (OECD, 2010). For individual schools, especially very small schools, the proportion of a school’s total budget dedicated to teacher salaries can be higher. (It is worth noting that approximately 30% of schools in Australia have four or less teachers.)

Yet there is a countervailing view within academia in Australia led in part by Dr David Zyngier of Monash University. A summary of his findings follow.
Zyngier has completed a review of 112 research papers on class sizes written by researchers in Australia England, Canada, New Zealand and non-English speaking countries of Europe. These papers were written between 1979 and 2014.

He reports that reducing class size in the first four years of school can have an important and lasting effect on student achievement. He goes on to say that the longer students spend in small classes during years prep to year 3, the longer the benefits for achievement last during years 4-8.

Zyngier claims this range of more recent studies on the effects of small classes have emerged and they suggest a different picture to the report authored by Jensen (2010). Zyngier reports that of the 112 papers reviewed, only three authors supported the notion that smaller class sizes did not produce better outcomes to justify the expenditure. He claims that reducing class size to increase student achievement is an approach that has been tried, debated, and analysed for many decades. The premise seems logical: with fewer students to teach, teachers should achieve better academic outcomes for all students. (2014a)

He suggests that politicians and media commentators both point to high performing systems such as Shanghai, Hong Kong, South Korea, Taiwan and Singapore, where large class sizes are the norm, suggesting that reducing class sizes is unnecessary. Zyngier cites research indicating that a Confucian heritage socialises students in ways that makes them productive in large classes. The same Confucian culture respects teachers highly making classroom management problems minimal. Therefore teachers are able to focus on whole-class learning strategies.

It is claimed by Zyngier that only a few authors can be found to support the hypothesis that smaller class sizes did not produce better outcomes. He cites Hanushek and Hoxby, suggesting they seem to stand alone in their findings that class size reduction has little or no impact on student academic outcomes and also suggests they are disproportionately referred to for evidence in Australia. He further suggests that other researchers have refuted the work of Hanushek and Hoxby, pointing out that Hanushek does not examine class size directly, but rather a proxy measure intended to represent it (student-teacher ratio).

Professor John Hattie has made meta-analyses of research by others looking at what teaching strategies lead to the greatest improvement in student attainment. Hattie (2005) suggests that the lack of improved student attainment in smaller classes may lie in a combination of size and teacher strategies. He states: Teaching practices that are conducive to successful learning are more likely to occur in smaller rather than larger classes, and these practices do not actually occur more in smaller classes because teachers have been prepared to, and indeed do, work with larger classes using more transmission practices and therefore they are not so equipped to adopt the more effective practices when they are given smaller classes.

At the other end of the spectrum there is also a crucial question. Can a class be too small? The volume of research in this area is much smaller and much of it anecdotal and from popular rather than academic sources.

Gladwell (2013) polled a large number of teachers in the United States and Canada and asked them that question, Can a class be too small? Many teachers agreed that it can. Gladwell recounts that twelve is small enough to fit around a large dinner table and that is too intimate for many high schoolers to protect their autonomy on the days they need to, and too easily dominated by any one person. By the time the class shrinks to six students there is not enough diversity in thought and experience to add the richness that can come from a larger group.

The small class is as potentially difficult for a teacher to manage as is a large class. In a large class, the issue is the large number of students to know and understand, the large number of interactions and relationships to manage. In the case of a small class, it is the intensity of the potential interactions. Gladwell reports teachers commenting when a class gets too small, the students start acting like siblings in the backseat of a car. There is simply no way for the cantankerous kids to get away from one another.

Gladwell reports experienced teachers suggesting that if the class is too small you don’t have enough breadth of opinion for strong peer discussion and the teacher tends to dominate the intellectual scene. This is fine where transmission of knowledge is the goal but not as useful where deep understanding through discussion and interaction is required.

In a very small class it can also be difficult for personality clashes to be avoided and for differentiation tasks to be undertaken within one class by small groups.

Levin (2002) has conducted research work along these lines, looking at Dutch school children. He counted how many peers children had in their class, that is, students with a similar level of academic ability, and found that the number of peers had a correlation with academic performance, particularly for weaker students.

In other words, if you are a student, particularly a weaker student, it is better for you to have other people around you of similar level, asking similar questions and wrestling with the same issues.

Levin argues that the problem with really small classes, is that there are too few students in a room and thus the chances that a student has a critical mass of other students like them becomes low. Levin concludes that taken too far CSR removes the peers that a student learns with and from.

In conclusion, the research into very small class sizes is limited and mainly anecdotal. At the other end, that is making larger classes smaller, perhaps Whitehurst and Chingos (2011) are close to the truth in a vexed and unsettled field of debate suggesting that advocates for and against CSR have engaged in or been accused of engaging in cherry picking for as long as there has been research on this issue.

Addendum October 27 2014

Since the publication of my comments on class sizes in the Falcon and on this site, I have been asked how the class size research applies to The Knox School. This is a very reasonable question and I hope the following assists.

The current class size at The Knox School policy allows for class sizes of up to 24 through the school and up to 26 after negotiation with stakeholders (24 maximum is a common, albeit not universal maximum class size across many independent schools, though many classes at both TKS and other schools end up smaller than that.). For example, the majority of classes in our Junior School are less than 20 in size.

There seems to be no set figure for the School at the other end of the spectrum, though there has been a few classes in recent years with very small numbers of students in secondary classes in Years 8-10. There is now a general indication in the course selection documentation that an elective class that does not attract at least 10 students may not run at years 8-10. The same stipulation is not written at years 11 and 12. The wording at VCE level is: <em>Finally, the school will make every effort to enable you to take a subject that you want to take. Sometimes, however, not enough students select that subject and the class cannot run. In these cases the school will work with you and your parents to help you to select a suitable replacement subject.</em>

These guidelines are not hard and fast rules but an honest indication to parents and students that a careful consideration of the educational and financial costs and benefits of very small classes needs to be made. There are a small number of factors that come into consideration, the educational ones are outlined in the article above. In addition to the educational factors, other major factors are firstly, to keep an array of curriculum pathways open for students through the secondary school to year 12 and secondly, the opportunity cost in running a very small class.

This year as student choices were collated it was decided to lower the number required for a class to run to eight to ensure that curriculum pathways remained open. It was decided that the opportunity cost of not using the resources in other places in the school was worth the benefit, even though eight students starts to arguably become less educationally productive as the commentary above points out.

All classes running from year 11 into 12 continue so as to preserve student academic pathways. Any class commencing in year 11, 2015 will continue through until the end of year 12 in 2016 to achieve the same result for the year 11 cohort of 2015. This commitment over the two year of Years 11 and 12 is a normal School commitment to the senior cohorts. In addition, the Director of Curriculum will investigate additional options where the School may be able to facilitate more personalised academic pathways for VCE students in the years to come, to complement the school culture where all students are well known and well cared for. I am not yet sure of the outcome of these investigations but remain quietly confident we will not only maintain a broad curriculum offering in a small to medium sized school but perhaps one even better than it is at the moment.


Bohrnstedt, G. & Stecher, BM. 2002. What We Have Learned About Class Size Reductions in California. Palo Alto, CA, CSR Research Consortium.

Chingos, M. 2010. The Impact of a Universal Class-Size Reduction Policy: Evidence from Florida’s Statewide Mandate. Program on Education Policy and Governance Working Papers Series, Harvard.

Gladwell, M. What is the “Good” Class Size? Horace Mann League Blog Accessed October 15 2014.

Gladwell, M. 2013. David and Goliath: Underdogs, Misfits, and the Art of Battling Giants. Little, Brown and Company.

Hanushek, EA. 1997. Assessing the Effects of School Resources on Student Performance: An Update. Education Evaluation and Policy Analysis 19(2): 141-164.

Hanushek, EA. 2003. The Failure of Input-based Schooling Policies. Economic Journal 113(485): F64-F98.

Hattie, J. 2005. The paradox of reducing class size and improving learning outcomes. International Journal of Educational Research, 43:387–425.

Hattie, J. 2008. Visible Learning: A Synthesis of Over 800 Meta-analyses Relating to Achievement. Routledge.

Hoxby, CM. 2000. The effects of class size on student achievement: New evidence from population variation. The Quarterly Journal of Economics, 115(4): 1239–1285.

Jensen, B. 2010. Investing in Our Teachers, Investing in Our Economy. Melbourne. Grattan Institute. Report No. 2010-6 NOV 2010 Pages 8-9

Jepsen, C. & Rivkin, S. 2009. Class Size reductions and Student Achievment: The Potential Tradeoff between Teacher Quality and Class Size. Journal of Human Resources 44(1): 223-250.

Krueger, AB. 1999. Experimental Estimates of Education Production Functions. Quarterly Journal of Economics 114(2): 497-532.

Krueger, AB. 2003. Economic Considerations and Class Size. Economic Journal. 113(485): F34-F63.

Levin, J. 2001. For whom the reductions count: A quantile regression analysis of class size and peer effects on scholastic achievement. Empirical Economics. March. Volume 26, No.1. 221-246

Mishel LR. & Rothstein, R. 2002. The Class Size Debate. Economic Policy Institute.

OECD (2010). Education at a Glance 2010. Paris.

Whitehurst, GJ, & Chingos, M 2011. Class Size: What Research Says and What it Means for State Policy. Brookings Institution. Washington, DC.

Zyngier, David. 2014a. Latest research shows class size DOES make a difference: pay attention Minister. EduResearch Matters: AARE blog: a voice for Australian educational researchers. May. Accessed October 17 2014.

Zyngier, David. 2014b. Class size and academic results, with a focus on children from culturally, linguistically and economically disenfranchised communities. Evidence Base. April. ANZSOG. Accessed October 17 2014.

Zyngier, David. 2014c. Class Size and Academic Results. Monash University. Accessed October 16 2014.

Back To Top