Comment by programjames

> There are systems that spend more per student than the US does, more % of GDP than the US does, they have lower classroom sizes than the US, they have better systems for gifted students, their students spend more time in class, they pay teachers professional salaries, they have more state of the art instructional programs, etc. etc. etc. None are trending towards the outcomes you're suggesting; in particular the EU isn't producing more billionaires

Look, everyone acknowledges Europe has a stagnation problem, America is better for startups, and the USD being the principal reserve currency makes Americans richer. You cannot directly compare countries like that. I know you're smart enough to recognize that, so I'm astonished you wrote this down.

Ideally, you could just randomly assign two policies, and see which works better. But in the real world, pretty much all studies in education are surveys, and it's hard to account for differences in space (countries/culture/socioeconomic status). Even differences in time (when new policies are introduced) have confounders, but less so. If you look at those studies, you'll find that new policies that throw money at smaller class sizes or higher-quality teachers lead to better educational outcomes. It's uncontroversial to say that better educational outcomes lead to better salaries, and higher wealth generation. (BTW, I'm not using wealth to mean USD, I mean quality-of-life. The USD is just a convenient proxy. Not sure if that was clear earlier.)

Tracking is more controversial. However, the one study that did just randomly assign tracking to 120 first grade classes found it benefited everyone:

https://www.educationnext.org/tracking-improve-learning/

Most other studies use standardized exams (e.g. the PISA), and let me remind you, those are not difficult enough to see improvement for the upper percentiles! You flippantly dismissed it, but it's a huge deal. If your national assessment were the AMC 10/12 the bottom 50% would all score zero points (really 37.5, but that's irrelevant), and tracking would look like a resounding success if the top few thousand showed improvement.