The gender gap in math is weak in Muslim countries

Moving to cross-country comparisons, we find earlier results linking the gender gap in math to measures of gender equality are sensitive to the inclusion of Muslim countries, where, in spite of women’s low status, there is little or no gender gap in math.

That is for students, not mathematicians, and it is from a new paper by Roland Fryer and Steve Levitt, hat tip goes to Chris Blattman.  Overall they conclude that the standard variables do not very well explain changes in the gender gap in math over time.

A non-gated version of the paper is here and it seems to be different.  Here is another version.


Muslim countries, where, in spite of women's low status, there is little or no gender gap in math.

But what are overall math scores/skills like in Muslim countries compared to other countries? Do they teach much math, and how much of that goes beyond basics?

And at what age groups? I'm pretty sure the math gap doesn't get strong until puberty and after.

From the first paper linked: " Note: the gap is eliminated in these countries because the girls achievement is higher, boy’s achievement is the same between countries with sex segregated classrooms and the rest of the sample."

Tyler: Also check out the papers by Andrew Penning at Irvine sociology. The bottom line is that math gender gaps vary greatly across countries. His case was East Europe, where female students often do better.

wtf. why is Liechtenstein considered a 'developing country' on NBER and not korea. what nonsense.

Do they eat more iron-rich foods? The gender gap starts when girls hit puberty and begin to have their periods, which causes iron deficiency anemia and low iron levels in some girls, and low iron is associated with poor math ability.

So, my personal theory is that the math part of the brain is particularly sensitive to decreased oxygen levels in the blood. If women of a particular culture tend to eat more iron rich foods, or if a particular ethnicity is less susceptible to iron deficiency, then that might make a difference.

Re: La Griffe du Lion -- a probability distribution is never really an explanation.

Let's say you are at an elite physics PhD program and are wondering why the gender ratio is so skewed from 50/50.

"Well, it all makes sense when you realize that the gender ratio at maximum-security prisons is also quite skewed. It's a law of nature, you see."

No, somehow that is not quite an adequate explanation. But throw around terms like Gaussian and variance and it becomes quite sensible.

Scott Hanselman has a podcast interviewing two women in Egypt who are computer scientists and finds that the split in technology companies are 50/50 (

They discuss the theory that in Muslim countries women are much less sexualized than in the western world, and so girls don't feel like they're worth is in their body.

Comments for this post are closed