Comparing international test scores and drawing ominous conclusions is quite the rage. Also, as GEMS Educational Solutions found, it is a great way to garner credulous coverage from The Economist and the BBC. All that’s needed is a pile of data and a mathematical model, and one can do creative things like rank countries’ educational systems based on their “efficiency.” It apparently matters not how much sense it all makes, as long as it can be puffed up with something that sounds sufficiently intimidating, such as a stochastic frontier analysis, to lend an air of gravitas to an inherently silly idea. So armed, GEMS set out in The Efficiency Index to rank 30 countries. When the data emerged from the GEMS statistical grinder, researchers concluded from the resulting mince meat that to get a 5% increase in PISA scores, teacher wages would (on average) have to go up by 14% or class sizes would (on average) have to go down by 13 students per class.
The most entertaining part of this report is the efficiency index itself, which purports to list optimal wage levels and class sizes for each country. For four countries, the optimal class size is estimated at fewer than two students per teacher. The teacher salary part of the report is almost as risible. The Swiss, we discover, should cut teachers’ wages nearly in half to achieve that nation’s “optimal” teacher salary. Indonesian teachers meanwhile would see their wages triple. As explained by our reviewer, City University of New York economist Clive Belfield, such anomalies expose the weaknesses in each of the study’s three key elements: “the output measure is questionable, the input measures are unclear, and the econometric method by which they are correlated does not have a straightforward economic interpretation.”
Meanwhile, even those of us in the research community who have long pointed to the benefits of (smart) class-size reduction can savor the irony of an “efficiency” report suggesting that two students is an optimal class size.