Equally unfashionable is the idea that there are advantages to homogeneous societies and classrooms.
Education in the U.S. is currently deficient, and we should look at why and how and try to find out what works. Unfortunately, it's tricky to run experiments in education, because of political considerations and largely because the result has real effects on people's lives, but it's what needs to be done to improve.
Or perhaps we need to do things which it is intuitively obvious should work, but which there is currently not the political will to implement -- eg firing bad teachers and doing more to stop student truancy.
Fire bad teachers: good idea, but you need to train and attract good teachers at the same time. How do you do that? Teaching is a badly overeducated, underpaid, and under-respected profession in the U.S.
Doing more to stop student truancy: This is cultural. Put armed guards to stop the students from leaving, and they don't show up in the first place. We need to foster a culture such that school isn't simply a prison students are coerced into going to for 9 hours a day.
The path from here to there is difficult and there isn't currently any money in improving education, so there isn't any political will either. It's a lot more difficult problem than your simplistic world view admits.
That was my point, these are two examples of ways in which education which can be solved with political will alone rather than requiring complicated new research.
Teacher salaries aren't as low as often assumed, by the way; the average teacher's salary in the US in 2006 was a surprisingly respectable $51,009 (with all sorts of perks) which, while no path to riches is certainly a ticket to a good middle-class lifestyle, especially in a two-income household. I personally wonder if part of the low attractiveness of teaching as a profession can be attributed to the fact that teachers' unions spend so much time whining about how low teachers' salaries are.
You don't need to put armed guards to stop 'em leaving, you just need to make sure they're punished when they don't show up. I never once played truant from my school; not because it was such a fabulous caring-sharing environment that I never wanted to leave (this is impractical, children will get bored at school) but because I knew if I might get caught and if I got caught I'd be punished. Now for me personally, the threat of detention and having my parents find out was a scary enough punishment, but I realise that this won't work on every kid... which is why I strongly believe that every school needs a form of punishment which every kid is afraid of; and if suspension won't do it for some kids then we need to bring back the cane.
I have read that the number of American adults of working age who have teaching credentials is something like double the number of currently working teachers. But studies of teacher incompetence
http://www.ericdigests.org/pre-921/incompetent.htm
show that the least capable teachers stay in the occupation the longest, and sometimes their very presence drives out the most competent teachers.
I agree that making schools less like prisons would be better both for teachers and for learners.
What they should be doing is working on the baseline students, inner city schools, poor areas. But this requires professors to get out their comfort zones and they don't want to do that.
A lot of these research schools are very competitive. They'll basically ace any test you put in front of them just because they can.
You said it's "tricky to run experiments in education." I would modify that to say that it's tricky to run proper experiments. What they are doing is running experiments without using controls or metrics. How can they get away with this? I believe it is because of the unpopularity of the idea that differences exist and should be taught to differently.
If you take PISA scores to be a good indicator than there are homogeneous and non-homogeneous societies near the top and bottom.
Canada ranks right up there with Finland every year http://en.wikipedia.org/wiki/PISA sometimes topping Finland in a category. It is easy to find countries that are culturally homogeneous near the bottom of the rankings as well.
Does being officially bilingual, northern, and woodsy have an impact on PISA rankings?
Finns do not make up a large or even a moderate proportion of Canada's population.
If I had to guess: yes. I honestly do believe that winters and environmental controls (like pavement) affect your approach to a wide variety of problems, and are immersive enough to do so on a nationwide scale. Not so sure about language, but a lot of other people seem to be.
This might manifest in educational systems more prominently than other parts of culture, or it might not. The thread of discussion is mostly about why the US is feeling inferior to Finland in one particular aspect, and grandparent stated that Canada also tests better there, so I would figure there might be something that Canada and Finland have but the US (and UK?) don't. Like, say, some aspect of culture.
Although obviously bordering cultures tend to affect each other greatly, my first instinct is always to look at a nation's climate.
One of the top Canadian provinces educationally is British Columbia. The majority of the province lives in Vancouver and Victoria. Those two cities have an extremely mild climate that is similar to Seattle and bears no resemblance to Finland or the rest of Canada.
If climate was the underlying reason for education going well in Canada, then you would not expect British Columbia to do as well as it does.
(Disclaimer. I'm biased because I grew up in Victoria.)
Destroying this hypothesis entirely, though, we have the high educational achievement of Singapore, which is hot and humid as heck every single day of the year.