Gyula Seres

PISA just dropped new test results. The headlines are unequivocal: "lowest OECD scores ever recorded"; "US at its worst mark in history." A year-plus of learning "lost." Then comes the usual hand-wringing about American kids falling behind other countries we presume we should beat. I want to call a pause to this for just one moment and ask: what exactly does PISA measure? A few years back we published a paper that puts things into perspective: ideas.repec.org/p/feb/framed… Surprise students with cash for right answers just before a PISA-style test begins and US scores jump substantially; Shanghai scores don't move. Same kids, same knowledge, different effort, different PISA percentile. In this way, the paper shows that PISA scores bundle two things: what kids know and whether they feel like showing it on a test that counts for nothing in their lives. I suspect the same happened this year, and the fact that over 40% of US test-takers didn't even bother finishing the accompanying questionnaire is consistent with that hypothesis. That number was so high that the OECD flagged the US results for potential bias. I am not saying that US education is fine. Indeed, quite the opposite, as we should be doing much, much more, beginning prenatally. I am just calling into question the received wisdom: "American kids know a year less than they did" and "American kids try less on tests that don't matter" are different diagnoses with very different treatments.