Class Sizes Revisited: Denny and Oppedisano

While we are on the subject of what really matters for human welfare, many people expressed approval of the government’s decision to not allow class sizes to increase further. Kevin Denny has been talking about this issue for several years and basically arguing that the evidence on class sizes is mixed at best and that it is more of a teacher workload issue than a pupil welfare and performance issue. He has just put out a paper with colleague Veruska Oppedisano testing the effect of classsizes on pupil performance. According to his results, bigger class sizes are associated with better performance in the PISA data, even controlling for a wide range of controls and using different estimation techniques. I think Kevin would be the first to say that you should be careful about overinterpreting any individual data analysis but it certainly points away from any simplistic assertions about the effects of class-sizes.

19 replies on “Class Sizes Revisited: Denny and Oppedisano”

Well, there is arguably a fixed amount of work involved in preparing each class, regardless if it is of size 4 or 40.

Might be for preparation but there is more to teaching than preparation. For example, what about grading and dealing with individual problems pupils have (the more pupils the more absolute problems)?

@ Liam and Kevin
“This paper estimates the marginal effect of class size on educational attainment of high school students. We control for the potential endogeneity of class size in two ways using a conventional instrumental variable approach, based on changes in cohort size, and an alternative method where identification is based on restriction on higher moments. The data is drawn from the Program for International Student Assessment (PISA) collected in 2003 for the United States and the United Kingdom. Using either method or the two in conjunction leads to the conclusion that increases in class size lead to improvements in student’s mathematics scores. Only the results for the United Kingdom are statistically significant.”

Please comment on the reasons for not being specific on what this paper actually covers
1) only high school students;
2) “Only the results for the United Kingdom are statistically significant.”

From a quick read of the results it appears that large differences in class size do make a difference in terms of student results. I would be fairly confident that smaller class sizes and better staff-student ratios would also have beneficial broader effects e.g. allowing a greater range of non-exam based activities in a school, allowing exxperimentation with teaching methods, greater use of project work, etc. But I’d be interested to hear of research on these kinds of questions.

Of course to get to these large reductions in class size you would have to go through a series of small reductions in class size – and for these smaller reducations the benefits won’t be nearly as obvious. Or indeed the immediate effects may not be that significant at all. But what would be significant is that they would be part of a move towards a more substantial reduction – which would have significant benefits.

Just a few comments.

1. Class size does matter. Teacher is only capable of dealing with so much at any one time.

2. Reading skills. All children have a good command of their mother tongue by age 3. Learning to read is very tricky and needs great care and attention. All primary pupils must be able to read fluently (sentences containing about 10 words) by age 5 – 6. If not, then they are at a definite disadvantage as they progress. Please do not underestimate the difficulty of this task – this is where you do need those small class sizes and sophisticated reading materials.

3. Numerical skills are not same as reading skills. Think about this!

4. Primary curriculum is selected and designed for a bye-gone era. My grandkids use electronic devices where I used to use pencil and paper (still do!). Need a big re-think here, but not by the ‘usual suspects’. This is a vexacious political matter.

5. Most parents would be better off (provided they possess the relevant knowledge and skill) to teach their young children at home (up to age 7 – 9). They would have to compensate somewhere for social contact – just as valuable as classroom learning.

Our young generation are not dumb: they’re just different!


Donal: Liam kindly posted the Abstract. An abstract is a short piece summarizing the paper in a non-technical way. To find out more you have to read the paper.
Ernie: I don’t know of any research that is informative about class sizes of 100 or even 60 or 70.
Sean,Brian: Interesting, but what’s your evidence? Remember, without evidence you’re not just another person with an opinion. Our evidence for the UK points to larger classes being better.

I’ll back Kevin up here, with data from another domain: 700k grades from an unnamed Uni in the (Irish) mid-west. Analysis suggests that module size has no, or a truly negligible, effect on measured achievement.

There are caveats: even in large modules, a proportion of the teaching occurs in small tutorial or lab groups. It is also possible that in some cases students in large modules are marked more easily, so that actual achievement is less than measured achievement. However the effect size (about 0.004 on a 1-4 GPA scale for a 100-student difference in size) is utterly tiny.

@ KD: “…Interesting, but what’s your evidence?”

OK. I surrender! Anecdotal observation and experience with Montessori education. Cannot refute your findings.


I remember seeing a report a couple of years ago of a high school in the UK who lost one of the school’s two maths teachers [he died suddenly, IIRC]. As an emergency measure they merged the two maths classes and started teaching them in a very large room.

Instead of seeing a drop in achievement, they instead discovered that this worked really well and achievement levels in maths went up.

I can’t find a link to the story, so mark it down as unsupported anecdote.

Meantime, here are two interesting stories from US and UK press. (i’ll post them one at a time to escape the link police)

Those looking for an introduction to how economists consider the importance of class size in *primary* education could do worse than investigate this debate between Alan Krueger, Eric Hanushek and Jennifer King Rice:

“Alan Krueger maintains that smaller class sizes can improve students’ performance and future earnings prospects. He challenges Prof. Hanushek’s widely cited analysis of the class size literature, arguing that it gives disproportionate weight to single studies that include a large number of estimates… Jennifer King Rice brings a third-party perspective to the debate. She addresses each author’s arguments and focuses on the policy implications of the class size literature.”

Looking at the evidence on class size in *higher education*, one result is that smaller classes do not translate into gains in achievement (Martins and Walker, 2006). Looking at economics students only, Kennedy and Siegfried (1997) find the same result i.e. that class size does not affect student achievement. Other work by Gleason shows that the same holds for mathematics students.

The opposite result (that class size matters) is found in a study examining peer effects and class size in higher education; Machado and Vera-Hernandez (2010) find that class size negatively influences medium ability college students. Dillon and Kokkelenberg (2002) show that class size “has a negative logarithmic relationship to grades and that the effect on class size on grades differs across different category of student.”

In a recent Vox article, Bandiera, Larcinese and Rasul (2010) state that “the effect of increasing class size in higher is not yet well understood”. Drawing on their interesting article forthcoming in the Economic Journal, Bandiera, Larcinese and Rasul describe how they estimate the effects of class size on students’ exam performance by comparing the same student’s performance to her own performance in courses with small and large class sizes. “Going from the average class of 56 to a class size of 89 would decrease the mark by 9% of the observed variation in marks within a given student. The effect is almost four times larger for students in the top 10%.”

The evidence on class size in higher education, primary education, and secondary education (which Kevin’s paper deals with) is certainly mixed. Of course, some useful principles for a critical literature review would be to

(i) place greater relative weight on experimental and quasi-experimental methods (as used in Kevin’s paper)
(ii) consider underlying mechanisms in depth (which is done in Kevin’s paper)

@ Brendan H:

In the introduction to the Bandiera, Larcinese and Rasul (2010) paper, the following comment is offered about mechanisms that may underly the impact of class size in higher education:

“There are several mechanisms through which class size can affect the behavior of students and faculty. These behavioral changes can occur inside and outside the lecture theatre. For example, students may be less attentive in larger classes, or may compensate for larger classes by exerting more effort either in the library or with their peers. Faculty may be better able to identify the ability and interests of the median student in smaller classes, or be more able to answer students’ questions directly. Outside of the lecture theatre, faculty might devote more time preparing the delivery of lectures and organization of materials for larger classes, or there may be congestion effects if faculty have less time to devote per student during office hours.”

Finally, I will offer some motivation for current work that I am doing on the impact of lecture attendance in higher education. It may be the case that universities (and other institutes of higher education) can measure and manipulate student enrolment relatively easily. However, if only a fraction turn up for class, then it is a somewhat blunt tool for policy.

@Comrade Ball:
“I’ll believe that class size doesn’t matter when I see the researchers who make this claim willingly put their own kids in classrooms with 100 students in them.”

Year 1 university courses often seem to have larger numbers of consumers in them. Why is third-level productivity so much higher?


Third level has entry restricted to academically able people. Note that large classes are more a feature of universities rather than ITs. Also these students are adults more or less and can be expected to take responsibility for their own learning. Also there is a significant dropout rate in first year third level courses. But it might just be that people in universities teach better!

Comments are closed.