Skip to main content
economy lab

University of Calgary Geoscience Professor Leslie Reid visits students in her class last month.Chris Bolin

Frances Woolley is a professor of economics at Carleton University, where she teaches public finance. Her recent Economy Lab posts can be found here.

In a recent column, Jeffrey Simpson argued that governments should be pressuring universities to reduce class sizes. This raises the question: do students actually learn more in smaller classes?

Teachers know that, when it comes to class size, certain thresholds matter. A class with fewer than 10 students lacks energy, and is hardly worth teaching. Seminars -- where each student has a chance to participate, share their ideas and thoughts with the rest of the class -- work well with up to 15 or 20 students. Above 20 students, it becomes necessary to lecture, but it is still possible to have some group work, and get to know each student individually. With more than 40 students, it is impossible to do anything but lecture.

In lecture-style classes, student numbers are relatively unimportant. A class of 60 is little different from a class of 120. Large classes -- 300, 400, 1000 students -- can be impersonal, and students tend to sit in the back and whisper. But given a choice between being one of 300 students listening to an engaging and entertaining speaker, or one of 100 students listening to a dull and pedantic one, most students will pick the good talker every time.

This has been known for centuries. Maimonides formulated a rule almost a thousand years ago: when a class has 40 students, it is necessary to hire a second teacher. (Universities today use a similar rule, hiring teaching assistants for classes with much more than 40 students.)

Now teachers' intuition is backed up by serious research. A study published last year in the Economic Journal (preliminary version here), estimated the impact of class size on university students' performance on standard, university-wide examinations. When the researchers focused on small classes -- those with 33 students or less -- they found that smaller classes promote learning. Students learn more in a class of 15 students than in a class of 30. When they looked at mid-size classes -- those with 33 to 104 students -- they found that class size did not matter. Students learn just as much in a class of 80 students as they do in a class of 45. In the very large classes, however, there was again a negative effect of class size on learning.

Similar results have been found in other studies. Spanish researchers, again focusing on mid-sized classes, found "no evidence that class size affects the probability of sitting an exam, the probability of passing an exam or the exam grade." A Kenyan experiment in class size reduction found that reducing average class sizes from 80 to 46, without changing teacher incentives, did not significantly improve students' test scores.

The bottom line is: if universities want to improve the student experience by reducing class sizes, they will need to make massive reductions in class sizes – from 400 to, say, 20. The cost of such a change is mind-boggling. To begin with, university lecture theatres are built to hold dozens -- or hundreds -- of students. Reducing class sizes would necessitate a huge investment in infrastructure. The personnel costs are huge too -- going from a single class of 120 to 6 classes of 20 students requires six times the faculty hours.

Because smaller classes require so many more teachers, efforts to reduce class sizes can have unintended consequences. When California spent billions of dollars on a program to reduce class sizes, schools -- especially those in less affluent neighbourhoods -- turned to inexperienced or unqualified teachers to fill positions. An analysis of the California experiment found that the hiring of lower quality teaching staff offset some of gains of reduced class sizes.

The only cost-effective way to give university students the same size classes as high school students enjoy is to staff those classes the way that high schools are staffed -- with full-time teachers who are not expected to do any research or develop their own curriculum. University is enough like an extension of high school as it is. There's no need to make it more so.

Follow Economy Lab on twitter