Dr. Subra Suresh
National Science Foundation
Engineering Deans Council
Public Policy Colloquium
February 8, 2011
Photo by Sandy Schaeffer
See also slide presentation.
If you're interested in reproducing any of the slides,
please contact the Office of Legislative and Public Affairs: (703)
[Slide 1: Title Slide]
Use "Back" to return to speech.
Thank you, very much, Ade [Ilesanmi Adesida]. It's a great pleasure to be back with this group, among friends. Having this many friends in Washington in one place is a rare event. [audience laughs]. I am delighted to be back here.
I want to tell you right at the outset that about three-and-a-half-years ago when I was appointed dean of engineering at MIT, many of my colleagues sent me emails both from within and outside MIT, chastising me for choosing to go the "dark side." But I've gone even to a much darker side, [audience laughs] even though I am in the comfort of and presence of friends who have chosen to go to the "dark side" as well. [audience laughs]
Had I had the opportunity to give this talk next week, it would have been a very different talk, because I would have had the benefit of talking about the budget that will be sent to Congress by President Obama. We have been working on this, but unfortunately I will not be able to talk about it until next Monday.
I would like to use this opportunity to talk about some of the broad themes we've been working on at NSF. None of this is new to you. Many of these have budget implications. More importantly, they all have national implications, and even in some cases, international implications.
This is more of a discussion for dialogue than any pronouncements, because of the timing regarding the budget, but these are long-term issues, which mirror many of the things that will be reflected in the budget for 2012 and in the years ahead.
So, I very much look forward to your input.
Just to tell you a little bit about where these topics came from, none of this is a surprise. Since my official nomination, I had the benefit of receiving a large amount of input including from many of you. I've talked with many of you personally. I've also held a number of retreats of the senior leadership team of NSF. And I have also met with every office and directorate in NSF. I can claim that I have pretty much met with every employee at NSF since my arrival in the middle of October.
Based on this, we started discussing a number of items that we would start to address. And I want to mention that we have had a lot of discussions with the National Science Board. And, Esin Gulari is here; she is Vice Chair of the National Science Board. The Board members had very useful input. Most of the issues I am going to talk about have evolved based on the conversations with the community at large, including thousands of items of input from rank-and-file faculty, university leadership, and within NSF.
Despite the economic climate we are in, it's really an honor and a privilege to have the opportunity to lead an agency like NSF at this point in time.
Most of you know this, but in case someone is not aware of the details, more than 20 percent of all university-based research is funded by NSF in the United States. More than 50 percent of all non-medical research in the United States at universities and other academic institutions is funded by NSF.
In some areas like computer science, NSF supports 82 percent of university research in academic institutions. And in mathematics, more than 80 percent of university research is funded by NSF. By its charter, NSF, unlike mission agencies, funds all branches of science and engineering, in the broadest sense. There are many areas where people ask: If NSF does not fund these areas, then who will? And the answer to that is no one, in many areas. So, it's an enormous opportunity, but it's also an enormous responsibility.
It's in this context that we examine the things that NSF does. Let me start, in the limited time that we have, with four broad areas that I want to briefly talk about. Unlike pretty much every other agency, NSF has the mission of a long-term perspective in supporting fundamental and basic research for the country.
The first idea I want to touch upon is how we maintain this core value, this core mission of NSF to fund the generation of ideas at the individual-investigator level and also think about broader infrastructure that no single university, no one institution, or even group of institutions can do because of the cost.
The second thing is especially relevant to engineering and computer science, and other areas such as mathematics and physical sciences. Without diverting precious research dollars, there are many examples and opportunities for nudging fundamental discoveries--that come from basic research--much closer to the market place. There are many examples. So NSF already does a lot; we could do more. This is one of the things we can do, not just by ourselves but in concert with other federal agencies.
Underlying all of this is not only an opportunity but a responsibility for us to address the STEM education pipeline. NSF funds not only research but also education and best practices in education. And NSF works with the Department of Education and other federal agencies. We have an opportunity to influence STEM education in a significant way.
So I am going to focus not broadly on STEM education but on one aspect of it, and that's the STEM workforce for the country. I'll provide some examples from data. And related to that, of course, is where we stand with respect to our international competition, and this is on the minds of us all, especially as we engage more and more on a global scale.
Last, but not least, disciplinary excellence is very important, but at the same time, "How do we sustain disciplinary excellence as we address complex challenges, global challenges and issues?" "How do we make disciplinary boundaries more porous?"
All of us face this in our own universities. But at NSF as well, it's a 60-year-old organization with a rich history. How do we create new opportunities for the community in multidisciplinary areas in more seamless ways than we have done?
These are four broad topics that I want to examine. They are all interconnected so I will not adhere to one particular order.
[Slide 2: Key Issues in U.S. Global Competitiveness in STEM]
We hear a lot about negative dependence on oil, but there is a good kind of "energy dependence": Our nation's long heralded innovation depends on the "fuel" that we call "basic research." Unquestionably, at least in the last 50 years, the U.S. has led the world in innovation, largely because we have had the opportunity of not only generating a lot of basic research and innovation but also having well-established institutions that interconnect in unique ways, yielding national and global leadership.
Last month, Foreign Policy magazine had an important op-ed. It was by Adam Segal, and the title was "Foreign Policy: How We Can Win the Invention Race."1 Adam is a senior fellow at the Council on Foreign Relations.
He argues that in order to out-innovate the rest of the world--a phrase that President Obama used his State of the Union Address a couple of weeks ago--we must appreciate the origin of the strength of our own system--the social, political, and cultural institutions that give physical form to ideas. So, our competitiveness derives from this strength.
And Adam Segal further discusses the growing competition, where significant resources at the central government level are given to educational institutions, especially research institutions, in large population centers of the world that are rapidly developing such as China and India. They are rapidly acquiring the "raw materials for innovation" at a pace perhaps much faster than we are providing at this point in time, especially for STEM education, science, and technology--generating a greater number of patents and more research papers. And the actual system for generating useful ideas in these economies has not yet caught up with where the U.S. is.
One can look at this in multiple ways. Segal's argument is that the U.S. still has a leadership role, and these countries are putting in significant new resources. Until they develop the kind of institutions in the innovation ecosystem that we have, we will continue to maintain leadership until we do something to jeopardize it.
One can also ask a slightly different question. As public and private universities in the U.S.--especially in the wake of the 2008 financial crisis--are seeking new models, one can ask: Without the baggage of a long history in education and research, can they more quickly learn from our mistakes and institute new organizations, institutions, and interconnected networks that leapfrog some of the educational issues we have inherited from the past? An example of this in the consumer domain is when citizens in countries like India typically had to wait eight years to obtain a land-line phone--because of the lack of a copper wire infrastructure--to the situation today in which fiber optic networks are prevalent, allowing that country to leapfrog over the cost of dismantling copper infrastructure.
These issues also talked about with PCAST [the President's Council of Advisors on Science and Technology] last week. Now what can agencies like NSF do? NSF started the SBIR [Small Business Innovation Research] program a long time ago. Now it's been adopted by nine different agencies. In the 1980s, we started the Engineering Research Centers. Later, the Science and Technology Centers. There are STTR [Small Business Technology Transfer] programs. There are IGERT [Integrative Graduate Education and Research Traineeship] grants. Individual-investigator grants that advance innovation closer to the market place. There are a lot more things that NSF can do, so we are looking into opportunities, and I would particularly welcome your input and suggestions.
For example, let's take institutions like those represented here--in Pasadena, Berkeley, North Carolina's Research Triangle Park, the Chicago area, Ann Arbor, Cambridge, Massachusetts, Princeton. Some of these institutions may not need an agency like NSF to nudge the innovation into the marketplace. They ordinarily do that very well, and some small incremental funding from an agency like NSF might help, but it's not going to completely change the equation.
But there are large numbers of institutions in the country where they don't have the ecosystem for innovation and for things like technology transfer, patent filing, technology licensing, and so on.
What can NSF do? There are careful studies that have been done by a number of entities such as the Deshpande Center for Technological Innovation at MIT to the Kaufmann Foundation, and a number of other entities.2 They have found examples of well-established innovation ecosystems in which for every "X" dollars in basic research, something as small as 0.1 to 0.5 percent of "X" dollars can nudge the innovation much closer to the marketplace. But, that's assuming the institution has the ecosystem.
There are other opportunities. The Secretary of Commerce Gary Locke has recently set up an innovation council. We have representation in that council. There are also other things that NSF does that in concert with what other agencies do, especially the Department of Commerce. We can set up virtual and physical opportunities for interaction, starting with the NSF grant community, so that we can implement some of these best practices on a national scale. We can also go further, based on the lessons learned, to an international scale, and this doesn't divert too much of precious basic research dollars into translational matters, which is not NSF's main mission.
This is a very good opportunity potentially, if done correctly. We have started a conversation not only internal to NSF but external to NSF, with other agencies, for greater interagency collaboration. This is something that many of you are engaged in, and I would particularly welcome your insights and input into this.
A week ago, the White House announced the launch of the Startup America Partnership, chaired by Steve Case. This partnership will receive launch funding from the Kauffman and Case Foundations. It will act as an independent private-sector alliance to increase the development, prevalence, and success of innovative, high-growth U.S. firms.
And there are a lot of things that we can do. Not completely, but as a nudge and a policy action, an agency like NSF can move us in the right direction.
Let me move on to the second major issue: That's the role of the U.S. at the hub of networks for global collaboration. And this is an area where we have started a lot of discussion internally. It's very difficult, and I know a lot of you resonate to this. In my previous job at MIT at the School of Engineering, 80 to 90 percent of all of MIT's international programs reported to my office, and it became very difficult to do bilateral agreements. Difficult because so many agreements really strain the system; they can strain your bandwidth; you can do only a few of them strategically. At NSF, it's even more difficult to do bilateral agreements because there are 195 countries that want to work with you. And there's not just the scientific issues, there are geographical and geopolitical advantages to consider. Even though this is not the main criterion, there is a science diplomacy aspect--in concert with the State Department--to consider.
So how do you balance all of these issues? One of the things we have been looking at--again, your experiences and input here would be very critical as we articulate our own policies as they affect the universities--is, if we have regional and multilateral alliances, what are the issues to consider? What are the strategic goals of different universities, especially in areas of science and engineering and how do we balance this? We have a number of ideas that are evolving internally. For example, designating the Middle East as an entity, South America as an entity, Asia as an entity. And, of course, within those regions there are large cultural differences, there are large internal differences and a lack of cohesion and perspectives. How do we bring these together? This is a topic that, again, is of a lot of interest, but as part of this, we are asking the question internally at NSF, "If we can do only three things internationally, because of our bandwidth and resource limits, what are the three things we can do that benefit the country?" And you can imagine that there is no unanimity of opinion on this. I look forward to hearing your perspectives on what that would be.
[Slide 3: Doctorates Awarded in Science and Engineering Fields, by Citizenship: 1989-2009]
Let me move on to one other point related to that. This is something that you will be very familiar with. This is the most recent data that we have on PhDs in the United States in science and engineering. This is for a 21-year period from 1989 until last year, the latest data. You can see a 10-percent, or so, uptick in the last years in U.S. citizens and permanent residents getting S&E PhDs. (The top blue curve shows U.S. citizens and permanent residents. And the bottom curve shows foreign students--people on temporary visas coming to the United States.) The entire uptick for the last year is because of women getting PhDs in science and engineering, in seven of the eight fields. This raises a very interesting issue. Let me review the latest data on this. In 2009, in the United States, 72 percent of all the valedictorians in high schools, public and private, were girls. And last year, the most recent year for which we have data, women comprised 20 percent more, than men did, of college graduates. Plus, the entire increase in the last decade in science and engineering PhDs for American citizens and permanent residents was because of women.
All of this is great news, for diversity and for a more representative workforce. But, then, in the most recent year for which we have data on women in the STEM workforce, 2006, only 26 percent of the workforce was female. To summarize, we have 72 percent of valedictorians being girls (many of whom are interested in science and engineering (highly motivated young girls); we have almost 40-50 percent entering college; and highly motivated young women getting PhDs in science and engineering in increasing numbers; but then it drops off.
There have been a lot of studies, but they reveal that it's a very complex issue. The most comprehensive studies were done by the Center for American Progress, and Berkeley Law.3, 4 And this, in conjunction with a number of parallel studies, shows there are many factors that contribute to this. But one of the most compelling factors for this drop off of women in the workforce--and that seriously affects our future as a nation in science and engineering research and education--are factors related to child-bearing and raising a family. So the Berkeley Law study actually provides recommendations for universities and for national agencies, like NSF, for things we can do, not just carrots we can provide, but ideas on how we can work together as partners.
So, I have started a major discussion within NSF as an agency, and this afternoon I am meeting with a group that is congressionally mandated to look at what NSF can do in this particular area. I had a discussion with the leadership of the group last night.
So, NSF can take some of its existing policies and activities, not necessarily creating new programs--that not only help the country but definitely help the universities, especially this group. I would be very interested in your ideas. I have also started a conversation with the leadership of AAU [American Association of Universities]. I will be giving a presentation to them in a month and a half on a number of issues, including this one. So this is an area where agencies like NSF can partner with universities to work collectively.
I talked about the STEM workforce: Women entering the workforce but leaving early without being able to contribute to the future of the scientific and technological enterprise. So we have a good-news-bad-news situation. The second major issue is one in which there is almost no good news in either entry or retention, or, at least we are lagging way behind. I am talking about underrepresented minorities. And this is especially true in the natural sciences and engineering and especially in particular fields. So, given that by 2040 we will be a nation of a majority of minorities, who now are severely underrepresented, with respect to the general population, in the scientific workforce, we cannot wait another 10 to 20 years. Now is the time to act. So what are the things we need to do? I know this has very much been on the minds of everybody in this room.
Let me give you one data point to think about. There are no easy solutions. Take the Iraq War and Afghanistan War. Take the representation of African Americans in these wars. These are mostly young college graduates or not even college graduates who join the Army. They comprise 19 percent of the military; their representation in the U.S. general population is 13 percent. If we can think of a way to give them an opportunity for education beyond what is already done by DoD--especially if we can find a way to excite them about science and engineering--there is an avenue to tap into a group that is underrepresented in the scientific workforce. This is just one example. There are many other examples and avenues. NSF has a number of programs. AGEP [Alliances for Graduate Education and the Professoriate] is such a program and it's in a number of universities including the University of Michigan. There are a number of such programs that are working very well. Again, we have done quite a lot, but many would argue we have not done enough. What do we do to address this? It's important.
The third issue related to the workforce is the foreign talent that comes to the United States. Many of us in this room came to the U.S. to get a doctorate degree. Most of us chose to stay here. Will that start to change with new doctorates? And, if so, what will happen?
[Slide 4: Top 10 Countries/Economies of Foreign Citizenship for U.S. Doctorate Recipients: Total, 1999-2009]
Here is the most recent data; these are the top-10 countries/economies from which students have come [over the 10-year period, ending in 2009] to U.S. universities to earn PhDs in science engineering fields as well as non-science and engineering fields. By the way, here "science" includes social science. It's not surprising that China is number one and India is number two. It is surprising that South Korea is the same as India, with a significantly smaller population. And then you have Taiwan and Canada. But what is remarkable about this is that in the surveys that have been done in the last few years in the top five countries that are listed here, people indicated that they expect to return to their home country at a much greater rate than their predecessors did, even just 10 years ago. We don't know yet if it's a trend, but given that nearly one-third of the Nobel laureates in this country got their first degree abroad, and 43 percent of the faculty of the 375 at the School of Engineering at MIT got their first degree abroad, a change could affect the quality of the research that NSF funds. How will it reflect on the quality of the education for which we all collectively in this room are responsible? This is a question that repeatedly comes up. It's not a bad thing that these people go back. It's good for science broadly in different parts of the world, but how do we make up for it, especially for the pipeline that will be challenging us with respect to women and underrepresented minorities and native talent that doesn't make up for the potential drop off in the foreign talent that is likely to stay here?
[Slide 5: National R&D Expenditures and Share of World Total, by Region: 2007]
Let me move on to research funding on a global scale. This is a relatively easy graph to describe. In 2007, of the world total, a little more than one third is in North America; a little less than a third is in Europe; and just about one third is in Asia. Everywhere else it's very small by comparison. That's the current situation.
Given this, what will happen with this distribution as we move into the future? And, also, what will happen if other countries don't pick up the S&E infrastructure for which we are so well known, especially if there is a change in the talent pool that migrates in the traditional way as we have seen in the last 50 years or so? We need to think it through.
[Slide 6: Gross Domestic Expenditures on R&D by United States, EU-27, OECD, and Selected Other Countries: 1981-2007]
There is one other cause for concern. We know that small countries like Finland heavily support basic research at a significant fraction of their GDP, one that is much greater than ours. So in this next graph--shown in constant 2000 PPP dollars from 1981 until 2007--these are the gross domestic expenditures on R&D in billions of dollars. So you can see at the top the OECD countries, and then the G7, and then the United States, then the EU-27, and Japan, and so forth.
This doesn’t quite give the picture. And the next two graphs will give the picture.
[Slide 7: U.S. R&D Share of Gross Domestic Product: 1953-2008
So if you look at the trends from 1953, three years after the National Science Foundation was created, you can see how the federal share and the non-federal share have changed with regard to the support of R&D as a share of GDP. A shift happened around 1978 or so. This is one issue that is a cause for concern.
[Slide 8: Nondefense R&D Gross Expenditures as Share of Gross Domestic Product, for Selected Countries: 1981-2007
Another issue is shown in this last graph. It shows non-defense R&D as a fraction of GDP in select countries. At the top is Japan, then South Korea, and then Germany, and the United States. Not shown is Finland, which is at 4 percent. Some of the other Scandinavian countries are also pretty high. Here is the take-away message: Japan, South Korea, and Germany surpassed us in 2000. We have been behind them for almost the last 11 years. This is something that is a major cause for concern.
So, this brings me to my last point, in the four bullets listed at the beginning. This is multidisciplinary research, which is on all our minds. Traditionally disciplines are very important in many areas, but at the same time, the problems are so complex that we need to bring down the barriers to research both organizationally and in terms of having the community do this. Organization is internal to NSF. Our directorates were created over the years; it's not as if we are going to dismantle all the Directorates and their current structure. The same is true of large departments and large schools of engineering.
But how do we create cross department, and, in our case, cross-Directorate and cross-office collaborations? This year, we have a number of activities that address this issue. Both in terms of organizational mechanisms internal to NSF but also in terms of funding mechanisms, how do we help the community do this? How do we make it easy for you to effect change in your own organizations? Again, this is not an easy thing; it's not going to happen in one year. It's a really long-term thing. Any input and advice you have for us would be most welcome, because this is something we are all struggling with especially in well-established universities.
Let me mention two major initiatives that NSF has articulated for the coming years, each of which will involve all of the Offices and Directorates in all of NSF. The first one we call SEES. It stands for "Science, Engineering, and Education for Sustainability." The second one is Cyberinfrastructure for the 21st Century, or "CIF21." Both of these initiatives are created so that they take existing activities in all the Offices and Directorates. And the CIF21 has been articulated publicly--and next week you will know the budget request from the White House and Congress for this--on how to use this as a vehicle for hitherto disparate disciplines, not only to create potential opportunities that may lie outside of traditional disciplinary boundaries, but as a community, we can create unique infrastructure to take on grand challenges and opportunities that we may not be able to do institution by institution but could do as a country.
With that, let me stop so that I can hear from you.
- Segal, Adam. 2011. "Foreign Policy: How We Can Win the Invention Race." Opinion article from Foreign Policy featured on National Public Radio website, (January 28). Retrieved 1/27/11 from http://www.npr.org/2011/01/28/133297705/foreign-policy-how-we-can-win-the-invention-race.
Return to speech.
- Stangler, Dane and Paul Kedrosky. 2010. "Neutralism and Entrepreneurship: The Structural Dynamics of Startups, Young Firms, and Job Creation," in Kauffman Foundation Research Series: Firm Formation and Economic Growth. Kansas City: September. http://www.kauffman.org/uploadedFiles/firm-formation-neutralism.pdf, retrieved 1/31/2011.
Return to speech.
- Center for American Progress. 2010. Creating 21st Century Jobs: Increasing Employment and Wages for American Workers in a Changing World. Washington, DC: Center for American Progress.
Return to speech.
- Berkeley Center on Health, Economic & Family Security and Georgetown Law's Workplace Flexibility 2010. (2010, December.) Family Security Insurance: A New Foundation for Economic Security. Retrieved online on 12/2/2010 at http://www.law.berkeley.edu/files/chefs/ family_security_insurance_2010_Final_web.pdf.
Return to speech.