The graph shows doctoral degrees awarded by region in science and engineering (graph from the United States National Science Foundation Science and Engineering Indicators 2004 report). The data used to make the chart is included in this spreadsheet on the NSF site.
It seems to me the claims of the NY Times article discussed in our previous post are wrong. I would trust this NSF data to be fairly accurate. The full report includes a great deal of related data and is worth looking at.
The data from the NSF 2004 report (the data is from 2000 and 2001 [the most recent data they have access to]) show a total of 24,409 science and engineering doctoral degrees granted in all of Asia. How many in the USA? 25,509.
International Mobility of Doctoral Recipients from U.S. Universities by Jean M. Johnson, NSF, 2000, provides some good discussion of related issues. For example, the paper explores country of origin of the students as well as where the students go to work once they receive the degrees.
The percentage of foreign doctoral recipients planning to stay in the United States may
return to the lower 50 percent level that existed until 1992. The 60-70 percent stay rates of the 1993-99 period may have been driven by the expanding U.S. economy and employment opportunities.
In any discussion of the impact of the United States failing behind in science and engineering graduation, and the resulting economic decline, it is critical to understand where the graduates go to work. There are real changes going on:
For example, in the last 5 years, Chinese and Korean students earned more doctoral S&E degrees in their respective countries than in U.S. universities. And in 1999, Taiwanese students, for the first time, earned more doctoral S&E degrees within Taiwanese universities than from U.S. universities.
This is important information. It is also important to see that it was just 1998 when more doctoral degrees were granted in the US than in Taiwan to Taiwanese students.
It seems there are at least two critical issues that people are considering when quoting figures (or related statements about the decline of US science and engineering status). One is getting scientific and engineering workers working in the economy. Another is the actual education of students, which relates directly to the first issue and has many “spin-off” benefits.
One measure used to look at creating future science and engineering workers is the number of those earning degrees (undergraduate and graduate degrees). That is a sensible thing to look at, though it should be noted that such a measure provides a limited view (it is an input measure and not an outcome measure, which would be preferable).
I believe the graduate measure is used as a way to project into the future by many of the future health of the science and engineering success of countries. It seems a sensible measure to pay attention to: we cannot measure today the number of high wages scientists and engineers employed in specific countries 20 years from now (or the jobs those scientists and engineers create for others in the economy or the useful patents written, scientific discoveries made, engineering breakthroughs achieved…).
The number of graduates has some value in trying to predict that outcome years from now but it is only a proxy measure and not at all definitive. The United States has been remarkably effective at getting those who graduate with advanced science and engineering degrees in the United States to say (and even in getting those granted degrees elsewhere to move here during their careers and gaining tremendous benefits to the United States economy). Where students receive degrees (and where they grew up), I believe is correlated to where a person ends up working during their career, but that correlation is not perfect. And that correlation may change in the future – in fact I believe it will do so significantly.
I believe the correlation will decrease – movement will increase and much of this may not even make sense as work flows without much regard for national boundaries (while physical location is one factor if essentially workers in Singapore, India, Mexico and Germany all our working on the same project for a company based in Japan and owned 40% by Canadians… how all this is analyzed gets very confusing).
Looking at where they work immediately after graduation is a sensible thing to do, however we should also look at where they work 10 or 20 years in the later if we are interested in long term impact.
The actual education of the students is also seen as critical to many, and I agree. One reason this is important is you have many good jobs educating the students. But there are many other benefits. The students often do research which if they are in you country is much more likely to benefit your economy than if they are earning there degree elsewhere and supporting research elsewhere.
Also the leading educational hubs create a climate for technological innovation (proximity to the leading experts in the world often provides benefits in tapping that knowledge for purposes that often have economic advantages). If the students are educated elsewhere it is likely those hubs of technological innovation will move also (or at least the lure of the local hub will loose some to another hub that grows in importance). So measuring the number of graduate, post graduate and doctoral degrees granted in your country makes sense (again it is not a perfect measure but a valuable one).
While there is a great deal of worry about the importance of improving science and engineering education to capture economic benefit I think the understanding of the actual situation is lacking. I think we need to have a clearer idea of what the data actual shows. Then I think we can start looking at where we would like to improve. I am to explore related issues with this blog.
