The CIS rejects economics
One of the odd features of the free-market thinktanks in Australia is their insistence that standard economic theory does not apply to education. The Institute of Public Affairs spent several years running the line that class sizes are irrelevant to educational outcomes. Now Jennifer Buckingham at the CIS has picked up the ball, taking almost the identical line.
The basis for the argument is the fact that it’s extremely difficult to detect effects of class sizes in statistical studies. This is not exactly surprising. It’s very hard to measure anything here. Ideally, we would like to be able to measure how ‘educated’ students are when they start a school year, how ‘educated’ they are when they finish and then take the difference as the contribution of the school. The best available measure, in most cases, is that provided by US standardised tests, which are notoriously unsatisfactory, being designed for cheap mass administration rather than careful assessment. Then there are the many other factors like teacher quality, some of which are even harder to measure.
American economist Rick Hanushek has tried to overcome these problems by collecting lots of studies and counting the results. He comes up with no effect, and is pretty strongly committed to the view that this reflects reality. But experts in meta-analysis have pointed out that Hanushek’s method of ‘counting studies’ is amateurish and unsatisfactory, and have proposed alternative methods that do find positive results.
As with lots of controversies involving statistics, the end result is inconclusive, and most people will go on believing whatever seems most reasonable to them. (Disputes over guns and crime provide a pretty good parallel.) With that warning in mind, you can read my summary of the evidence here.
For an economist, the answer is straightforward. Other things equal, more inputs mean more outputs. This is true, even in the presence of ‘distortions’, such as overstaffing or union restrictions on technology. Hence, unless education is a special activity not governed by the normal laws of production economics, we would expect to see an inverse correlation between class sizes and outcomes.
But wait, there’s more. The CIS (or at least, Jennifer Buckingham) is discarding consumer sovereignty as well. Thanks to the existence of private schools, we have market evidence (in the absence of tight zoning, we get similar evidence from the preferences of parents when they choose between state schools). If small sizes are valueless, we’d expect parents to prefer private schools to spend their money on other things, such as computers or study tours. But in fact there is very little difference between the proportions of resources allocated to teaching staff and other resources for private and public schools. The more money the school has, the lower the class sizes.
To rationalise this practice you have to assume that parents (all former students themselves, of course) not only have mistaken beliefs when they initially choose to send kids to schools with low class sizes, but are incapable of telling whether the education their kids are getting is good or bad once it’s underway. If this weren’t true, schools could gradually shift resources from teaching staff to other uses, and parents would see the benefits and keep their kids in school.
Of course, once you accept this, the whole case for any sort of choice (individual or collective) goes out the window. If as Buckingham explicitly asserts, teachers, parents and governments are all fundamentally mistaken about what makes for good education, the only reasonable policy option is to appoint people like Buckingham and Hanushek to make our choices for us.