Very interesting article. Thank you for posting it Jessica. I have read it and also the full paper and appendix. For anyone interested you my find the published study here:
http://memp.pratt.duke.edu/outsourcing/
Having read this paper and looked at some of the data in it I feel that the paper misses a very important point as do most of the other articles and published studies I have seen on the subject. Whether you look at the raw numbers of graduates in the U.S. vs India and China or the population normalized values this study presents these studies all make the assumption that the graduates from U.S. institutions contribute to the U.S. work force and the graduates of India's institutions contribute to the Indian workforce and the same goes for China. I am not so sure this is a valid assumption. How many U.S. citizens per year go to India or China and seek out an engineering, CS, or IT degree and then return to the U.S. to work for a U.S. company? Not having ever been to India or China myself and not having ever seen any statistics on this I don't know for sure but I suspect not many. On the other hand, look at the number of foreign nationals that you see at American universities campuses. I would suspect that a large percentage of the graduates from U.S. institutions getting 4 year degrees in engineering, CS, & IT do not, in fact, remain in the U.S. and contribute to the U.S. workforce. Even when I was in school (way back) we had a very significant percentage on foreign nationals in my classes. In discussions with them most were studying in the U.S. because our schools were better and because their tuition was paid by either their government or a company in their country with the understanding that they would return to work there. This troubled me even then but more so now. The numbers in these studies do not alarm me nearly as much as my observations as to the number of student the U.S. is graduating but exporting. This may simply be my perception but I have not seen any statistics on the subject. I would like to see a follow up study to the Duke research that accounts for these percentages.
Also, something interesting I noted in the appendix, under USA Data, page 1 of 4 (3rd page of the PDF), under the field of study "Computer..." the U.S. apparently graduated 0 (zip, nada, not one) person under the "Artificial intellegence and robotics" heading. Is something wrong with the data or is FIRST failing somewhere? I really can't believe 0.