Thanks to Jie Tang for sending this.
Economists Roland Fryer and Steven Levitt have a 2009 study, "An Empirical Analysis of the Gender Gap in Mathematics," in which they analyze the gender gap in the US, explore causes, and discuss cross-country data which shows that countries with same-sex schooling don't have this gap. Paper here; abstract excerpt below:
We document and analyze the emergence of a substantial gender gap in mathematics in the early years of schooling... in the United States. There are no mean differences between boys and girls upon entry to school, but girls lose more than two-tenths of a standard deviation relative to boys over the first six years of school... We explore a wide range of possible explanations in the U.S. data, including less investment by girls in math, low parental expectations, and biased tests, but find little support for any of these theories... The cross-country data reveal that girls do not lag boys in math in countries with same-sex schooling, raising an intriguing question as to whether this relationship is causal.
I am not surprised. Since traditional gender roles are so ingrained in our consciousness, it is natural that young men and women would look to them in determining acceptable behavior*. Having young women hold back in math causes a feedback loop that makes it increasingly difficult for young women to develop their mathematical ability: as fewer women become good at math, 1) people become less accustomed to seeing women who are good at math, making it difficult for women to proves themselves w.r.t. math and 2) women with mathematical talent are less likely to develop it, since they are not expecting to be good.
These findings corroborate the hypothesis that the US gender gap in math is largely due to cultural factors. (See my other link.) They also fit with my personal experiences: one of the biggest social challenges I faced in leaving my all-girls high school for a coed college environment was having people treat me like I shouldn't know what I'm doing when it comes to math. This was not only frustrating but also harmful to my self-confidence--it probably caused some amount of deadweight loss in my mathematical development.
Shifting current gender roles could go a long way in bridging the gender gap.
* I am curious to see studies analyzing how other traits associated with masculinity (physical strength, debate ability) fare under an analogous analysis.