<p>As to the PhD productivity, cghen, I agree with Graham, and 25% is nowhere near what we have. I quote from Wikipedia:</p>
<p>“about 40 percent of graduates go on to earn a Ph.D.the highest rate of any college or university in the nation…”</p>
<p>Futhermore, there is a study done by Change magazine on Ph.D. rates over a 30-year period. Look at the rankings:</p>
<p><a href=“http://www.amatecon.com/etext/cac/cac-ch03.html[/url]”>http://www.amatecon.com/etext/cac/cac-ch03.html</a></p>
<p>As for the NSF fellowships, I found this link: <a href=“https://www.fastlane.nsf.gov/grfp/AwardeeList.do[/url]”>https://www.fastlane.nsf.gov/grfp/AwardeeList.do</a></p>
<p>Just looking at the year 2006, I counted 12 NSF fellowships from Caltech undergrads and 10 from Mudd undergrads. So there is a 6:5 ratio while the total undergrad population ratio is ~9:7. So there you have it.</p>