There's a great thread and discussion going on over at DrugMonkey's about the Ginther report and how minorities are being screwed over in the NIH grant game. If you haven't already, go read it.
It's clear that there's a strong bias against people of color (POC) somewhere in the review process. How else to explain a 50% higher rate of triage/not discussed? I don't have anything substantive to offer to the discussion right now beyond some more data.
The NSF send reports on their merit review process to the National Science Board every now and then.* The 2012 report, which covers fiscal year 2011, can be found here (pdf). There's a lot of really interesting stuff included within this report, including some demographic/funding success data (Appendix 3).
Here's Appendix 3:**
Interesting. Minorities still suffer, but perhaps not as much as at NIH. Why? What is it about the NSF that leads to slightly better (albeit still not good) agreement between the funding rates of minority PI's versus white PI's?***
Have a read and think about it. As noted by DrugMonkey in a conversation on the twits yesterday, we're talking death by a thousand cuts here. There won't be one simple explanation.
* Thanks to Bashir for his post here which prompted me to dig up this report.
** Also have a look at NSF reviewer (page 29) and Program Officers (page 32) of the report.
*** It may not be better really. We're not comparing the same sorts of numbers here.