Kevin Drum weighs in here:
Via Jay Mathews via Bob Somerby, today’s chart of the day1 demonstrates vividly the long, slow decline of American education. Except wait. It doesn’t show that at all. These figures come from Brookings Institution scholar Tom Loveless, and they show how American kids have done on international math tests compared to kids from eleven other advanced countries.
. . .
In 1964, we were 0.35 standard deviations below the mean. In the most recent tests, we were only 0.06 and 0.18 standard deviations below the mean. In other words, our performance had improved.
. . .
One thing that’s pretty clear, though, is that America does a terrible job of educating low-income students. But even there, the comparative data is unclear (or at least, I’ve never seen clear data). Do our low-income kids score worse than other countries’ low-income kids? Or do we simply have more low-income kids? Since income figures aren’t routinely gathered for these tests, and international comparisons of income are problematic anyway, this isn’t an easy question to answer.
As I’ve argued before, international comparisons — on education, on gun violence, on whatever — are absolutely useless without demographic controls, because the demographics with the most troubling statistics in the U.S. are a much lower proportion of the population in the rest of the developed world.