Class sizes and the dead hand of history

One of the deepest and most heavily defended myths of education talk, research and policy is that small classes are good, and therefore smaller classes are better still. The argument is as simple, appealing and popular as it is facile, wrong and damaging. It has acquired current political significance from Gonski's call for a substantial increase in spending on schools, and from the ham-fisted efforts of Coalition shadow education minister Christopher Pyne to incorporate criticism of the class size reduction campaign into his platform.

One of tens of thousands of articulations of the "small classes" case (and one of many illustrations of its unsustainability) appeared in the (excellent) online outlet The Conversation and was retailed by The Monthly magazine's (excellent) daily news digest. "Class Sizes, Gonski and Schools Funding: The Facts," by Monash education academic David Zyngier, is (in fact) a convenient selection from the available facts. More important: it misses the central point entirely.

Smaller classes can and often do improve student learning and welfare, as Zyngier asserts, and they can make teachers' work more doable and satisfying. Of course they can! But the point, entirely invisible to Zyngier, is this: could other ways of spending the same amount be more effective? For students and/or teachers?

In Australia the answer is almost certainly yes, they could, partly because each class-size reduction is very expensive, partly because each increment doesn't make much difference, and partly because there are more cost-effective strategies available or on the way. Some examples.

• A quasi-experimental study in the United States compared five ways of improving literacy and numeracy: CAI (computer-aided instruction); increasing the length of the school day; class size reduction; and peer and cross-age tutoring. Of these five strategies class-size reduction (in various "steps," from thirty-five to thirty to twenty-five to twenty, and one big step, from thirty-five to twenty) vied with increasing the length of the school day for the worst cost-effectiveness ratios. CAI was twice as cost-effective as either, remarkable when we learn that the study was conducted in the mid-1980s. Various forms of tutoring came out on top, with peer tutoring more than three times as cost-effective as class-size reductions. It is perhaps worth noting that the two senior authors of the study, Henry M. Levin and Gene V. Glass, are among the most eminent of US education researchers.

• An example of the same approach applied at the macro level: a recent US calculation estimated that just five more students in each US classroom would fund an across-the-board salary increase for every teacher of 34 per cent. That would in all probability improve student's working lives as well as teachers'. There is good evidence to suggest that higher salaries attract more able people into the profession, and that they become more effective classroom teachers.

• A third example: Australian teachers are the victims of something approaching world's worst practice in teacher education, both before they go into the classroom and after. There is good reason and evidence – ably summarised by Zyngier's bête noire, the Grattan Institute – to suggest that teacher effectiveness and professional satisfaction can be substantially improved by well-organised, workplace-based, collaborative appraisal and professional development. The problem is that this takes workplace-based time. Perhaps, Grattan suggests, teachers and their students would both gain if their schools traded off the number of classes per week and/or the number of students per class to make the time?

• A fourth: virtual high schools in the US have substantially lower per-student costs than conventional schools. And so, proportionately, do "hybrid" schools, which find various ways to mix virtual with mainstream instruction. But are these new schools as effective as the old? It is too soon to tell, but so far there is no evidence to suggest that they won't be, and some reason to believe that if they were to spend at the same per-pupil rate as conventional schools they would be rather more effective.

To repeat: to argue that smaller classes are better than big ones is to state the obvious and ignore the important. In doing so, and by being almost silent on costs and completely blind and deaf to opportunity costs, cost-effectiveness and productivity, Zyngier is representative of his profession.

Education research is dominated by the so-called "effectiveness" paradigm, the findings of which are so ably summarised by John Hattie's Visible Learning. Hattie's tour de force digests more than 800 meta-studies, themselves digests of a total of tens of thousands of studies of "effectiveness."

Had Hattie digested cost-effectiveness research he would have published a very slim volume indeed. Education researchers who, like Henry M. Levin (and the Grattan Institute's Ben Jensen), are trained in economics and therefore able to see and understand the relationship between both sides of the equation, are rare. Their research is rarely noticed or used.

These facts reflect the circumstances in which most education academics have conducted their careers and their research: in schools and school systems in which just about every dollar is spoken for well before it passes the parliamentary vote, committed to salaries of tenured teachers (between 60 and 70 per cent of most school system budgets), on fixed rates of pay, working in numbers and combinations determined by fixed class size maxima, all underwritten by industrial agreements and regulations. Resource use is fixed, familiar, and uniform, and therefore invisible. It is not accidental that Levin et al had to set up an experiment; the comparison was not there to be made in the status quo.