The survey doesn't consider serial entrepreneurs. They just ask the age of the founder during the current company's incorporation. (For example, Andreessen wou

The survey doesn't consider serial entrepreneurs. They just ask the age of the founder during the current company's incorporation. (For example, Andreessen would contribute an older data point with Ning even though he was a prototypical young guy success story.) Also, I think the attention paid to the Ivy-League is unwarranted given other possible groupings in the data. If you look at the data in the paper, it looks like you can create a non-Ivy "Entrepreneurship" league comprised of the following schools: MIT, Penn St, Stanford, Berkeley, Missouri, USC, Texas, and UVa. By my mental math, this 8 school league graduated just under 14% of the founders compared to the 8% for the 8 school Ivy-League. Regional effects were probably even more interesting, and you wonder what a SF area (Berkeley/Stanford) + Boston area (MIT/Harvard) grouping would look like compared to other regions.