Didn't read the paper, but I am wondering: How can they possibly distinguish language from all the other cultural background elements (taxes in the country, education, culture...) and still get statistically significant results?
As in, I can see how we can verify that "people in China save more than people in the US", but not how "people who speak mandarin save more than people who speak English, *because* they speak Mandarin". I glanced at the paper, but would love if someone could dumb this down for me...
Edit: also wondering what that does for people who speak several languages :)