How Insurance Works

Congressional Briefing—The Decline in Research: Should We Worry?


Arti Rai: Please pick that material up. I’m a professor at Duke Law School, and we have a group of
folks from Duke Law School, and also other institutions here today, but before we begin all that, we have a couple of
logistical announcements, and then I would like to
introduce our honored speaker, Senator Chris Coons. There are two logistical
announcements in particular. Lunch is back there. You should feel free to get it at any point during our event. In addition, the speakers for our event will be coming up sequentially
to give short talks. After every speaker has spoken, the panel speakers that is, we will then all convene as a group to have a panel discussion together with engagement with the audience. So, those are the two
logistical announcements. And now I am delighted to
introduce Senator Coons, probably needs no introduction, but as you know he is
a senator from Delaware who was elected in 2010, following terms as the New Castle County council president and New
Castle County Executive. In the Senate, he sits
on numerous committees, numerous important committees, including Appropriations,
Judiciary, Foreign Relations, Small Business,
Entrepreneurship, and Ethics. Senator Coons has made
manufacturing and innovation, the subjects of our discussion today, top legislative priorities in the Senate. He co-founded the bipartisan Senate Competitiveness Caucus, and co-chairs the Manufacturing
Jobs for America initiative, two efforts that have
attracted participation from dozens of senators of both parties. Senator Coons has authored numerous pro innovation bills that
have been signed into law, including a permanent extension of the R&D tax credits for startups and small businesses, as well as the Defend Trade Secrets Act, which creates new federal protections for intellectual property. Before entering government, Senator Coons spent eight years working for an advanced materials
manufacturer in Delaware. So, without further ado, Senator Coons. (applause) Senator Chris Coons: Thank you. Well, thank you, Arti, and thank you everyone for taking time out of your busy schedules to
serve on a fantastic panel, to listen to their wisdom, and to contemplate our common topic today, The Decline in Research, Should We Worry? I know I will reveal no great
state secrets by saying yes, I think we’re done now, are we done? (laughter) To the panel that is about to present, let me just say my thanks to
Stephen Merrill’s leadership at the Center for Innovation Policy. You’ve put together a
terrific panel of experts with an important and timely topic, to Wesley Cohen, to
Ashish Arora, to Pian Shu, thank you for your work on
the economics of innovation, and to Steven Freilich, thank you both for being a good companion
on the train ride down from Wilmington this morning, and for all you’ve done
over the decades at DuPont. Congratulations on your new position as Director of Corporate Strategy at the University of
Delaware’s Energy Institute. So, I know you have a
far more substantive, thoughtful, and learned panel
to immediately follow me. I’m really to distract
you while you eat lunch, so feel free to keep eating, and I’ll try to be
distracting if not engaging. We are here to discuss
not just should we worry, but what should we do about it. It is clear the answer to both questions, unequivocally is yes. Yes, because a strong R&D investment is a key component of a
robust innovation ecosystem. Yes, because more and more US firms are performing R&D abroad. In fact, in 2015 over third
of all R&D for US firms was performed in Asia, a 30
percent increase from 2007. And yes, because the Trump budget proposes significant reductions, across the board cuts
really, in federal research, jeopardizing our
long-term competitiveness. Now why would we be cutting
research and development when it is so important to job creation and economic growth? No administration, no administration, Republican or Democrat, in 40 years, has been more hostile towards federal research and
development than this one. The Trump budget cuts non-defense R&D by more than $14 billion or nearly 20 percent in
their proposed FY 18 budget, It cuts the NIH, which
is essential to advances in all sorts of critical areas of science, and cures, and treatments, not least of which is a
search for cures for cancer, and it cuts the NIH budget by 18 percent. It cuts manufacturing USA, a vital set of institutes, aimed at boosting domestic manufacturing, and tying innovations in new cutting-edge areas of science to the development of new
capacities in manufacturing. It cuts that by more than 70
percent, eliminating all five of the Department of
Energy-run institutes. It cuts energy programs at
the Department of Energy by 60 percent, and basic
science research funding at the National Science
Foundation or NSF, by 11 percent. The good news is that both
Republicans and Democrats in the House and Senate have already said that this budget is dead on
arrival in this Congress, but the devil will be in the details, and the administration may
not achieve the dramatic cuts that they propose in federal R&D, but they may succeed in
significant sustained and across-the-board reductions. A reduction in the rate of growth, or a reduction across the board, either, would be a negative outcome
for the whole country. Budgets, as my predecessor
Joe Biden long said, is an expression of your values, of your values and
priorities as a country, and Congress doesn’t share this administration’s
values or priorities. When it comes to investing in basic and applied science to
try and find solutions to the very real problems that affect
Americans day in and day out, whether it’s in health care and diseases, or it’s in energy and transportation. That doesn’t mean we
shouldn’t be deeply worried about the impact of these
massive proposed cuts, and more importantly long-term uncertainty when it comes to federal
investments in R&D. Investments in R&D
are vital to our economy. One study of 15 leading economies globally showed that a $1 investment in R&D, generated more than $20
in economic activity. Even taking into account
private sector investment, as a country we still significantly lag behind our competitors around the world in R&D investment. One recent ranking of
total R&D spending as a percentage of GDP,
places the United States, not first, not second, not
third, but today in 10th place, even though our economy
is nearly twice as large as all of the other countries
that outrank us combined. We understand the importance of federal support for R&D, and what these investments mean for our economy and for our future, but many others particularly
here in Congress may not. It is on us to help the federal government advocate for R&D and
to help the American people appreciate the strong link between federal investment in R&D and economic growth, job
creation, and quality of life. Let me make one other side point, that’s small but significant. I was a major in chemistry in college, and I still remember as my classmates were making decisions
going into graduate school, whether to pursue a career in physics, or in chemistry, or in engineering, and the signal at the time, this was the Reagan Administration, was that there were dramatic increases in investment in defense-related R&D. So, a number of folks who had
a choice went into physics knowing that it would be
funded for the long term, and chose not to go into
chemistry or biochemistry, believing it would not be
funded for the long term. An entire generation of graduate students and researchers, of professors, and those who will lead innovation
for the next generation, are making career decisions today, and if all you see is a March
for Science on the Mall, out of desperation, and alarming news that there are dramatic cuts coming in federal R&D investment, how many of the most promising researchers for this next generation will take
the off ramp, and instead choose some completely
different path or career. These cuts have long term consequences. So, there’s a case we have to make to the American people, a case that investment in R&D makes us more competitive,
will grow our economy, and will give a higher quality
of life to future Americans. Why am I at all optimistic this is the debate we can win? It’s not just because the
Trump budget has rallied both Democrats and Republicans, to stand up for key US investments at home and around the world. I’m also optimistic because
despite the very partisan, very difficult, very divided
political environment here in Congress, I’ve gotten some real legislative
actions enacted into law. As you heard in the introduction by Arti, an expanded, permanent R&D tax credit for startups and small businesses, a new federal right of action
for defending trade secrets, an improvement for the Manufacturing
Extension Partnership Program, and a stronger iCore
program within the NSF. I’m optimistic because
in each of these cases I have found a Republican
willing to work together to legislate and they’ve
been signed into law. Each of the laws I just named had Republican co-sponsors, was passed by a Republican house, and signed into law by
a Democratic president. In fact literally today, I am introducing the Invent and Manufacture
in America Act, with Conservative Republican Senator Pat Roberts of Kansas, which increases the R and
D tax credit for companies that don’t just design, or
research, or develop a product, but they also manufacture
it in the United States. As I’ve often said in my
home state of Delaware, how much of the guts of this,
was designed, researched, and developed in the United States? How many of these are manufactured in the United States? If we can have more of
what is invented here manufactured here, we will have a stronger
future as a country. So it can be done, but
these types of victories aren’t going to materialize
magically in Congress without your support. You have to work with us to
convince this administration, and many in this Congress, that investments in science,
in R&D, and in innovation, deeply matter to our security, our prosperity, and our future. So, frankly, be persistent and loud. Assume nothing. Explain what R&D means
for the American people, and lean on your Republican members, those of you who work in Congress, your Democratic members, and your allies in the administration. With your advocacy and engagement, I believe it is possible
to not just sustain, but grow federal investment R&D, and to improve the policy framework for our future competitiveness
and innovation. Thank you very much. (applause) (speakers muffled off mic) Have a great conference, and thank you again to the panel. Arti Rai: Thank you, Senator Coons. Thank you again, Senator Coons. Well that was a rousing
start to what I think will be a very exciting panel. We will be moderated,
if we can be moderated, five professors and one ex R&D leader, by Wesley Cohen who is himself a professor at the Fuqua School of Business, and he is one of the
world’s leading authorities on innovation and entrepreneurship. His research on the economics
of technological change and R&D has really
literally changed the course of many debates in the area. The first speaker, the first panel speaker will be Ashish Arora, his colleague at Duke’s Fuqua Business School. Professor Arora is the Rex Adams Professor of Business Administration, and I think it’s fair to say that, he too, his work as well has changed the debate on questions of technological change. He also, Professor
Arora is also co-editor, or has recently stepped down, as co-editor of the leading journal in
this area, Research Policy. So, I will invite both Professor Cohen and Professor Arora up here, and Professor Arora will start the panel. Wesley Cohen: Very good. Just a couple of introductory points here, points of order. We will proceed sequentially
through the speakers, who will each come up in turn, and each speaker is allotted 10 minutes. I am the official timekeeper, and that is to allow for plenty of time, for engagement, and
interaction, Q and A, and so on. So, that’s really our objective. We really look forward
to hearing from you all, once the speakers have
finished their remarks. So, with that, the very
eminent Professor Aurora, please take it away. Ashish Arora: Thank you, good morning. The title of the what I
want to say is right there, which is I’m gonna talk about, we talk about R&D all the time, as if they were one word. I want to talk, break those
apart and talk about research in research and development, and specifically what the
corporate sector is doing. So, these data are from the
National Science Foundation. It’s a busy chart. This bar, these gray bars is real, that is to say, adjusted for inflation, basic and applied research,
funded and performed by the business sector
in the United States. You can see that it’s
been roughly constant, flat over the last 10, 15 years. So, this is in absolute terms, even while the economy is growing by about 60 percent or more, real spending on research by
the business sector is flat. Those two lines tell you the
same story as percentages. This is the share of
business in total research. Sorry, make sure I’m saying this right. Yes, this is the share of
business in total research, and you can see that’s fallen. And you can, and this
is the share of research in business R&D. So half of every dollar that
business spends in R&D, what fraction goes to research, and that’s steadily fallen. So, the takeaway here, and
you’ll see this repeatedly, is less R, more D, or less R in R&D, if you want to be precise. This is the same story,
different kinds of data. I showed you NSF statistics. These are data that Sharon
Belenson and I put together, where we matched corporate publications, scientific output, and they
tell you the same story. Even while patterns have
increase, that’s the blue line, publication output has fallen it, when you normalize it for
firm, or normalize it for size, and this is the same story in terms of publication
output by different fields. And the point of this picture is, the decline that the NSF number shows, you can see using a different measure, namely scientific publications,
you see the same results, if you do the appropriate controls, adjusting for the fact for example, that some companies you know, may have declined in business. So, if a company like, for example, Kodak, does much less manufacturing
that it used to, and is much smaller than it used to be, so you’d expect that
decline to take place. Once you adjust for all of that, you still get the same result. Alright that’s, so it’s
just different ways of making the same point, less R, more D. This is NSF data, once
again looking across sector. So, the point here is,
these trends of broad-based. They’re not driven by particular sector. So, we all have our best stories about how this happened in IT, or how this happened in
semiconductors or computers. Across a range of sectors, you still see the same effects. Here’s something that we could do, because we were able to match publication output at the firm level. What this tells you is,
so let’s start with this. This is the value, the
stock market premium companies used to get from
1980 up until the mid ’90s, so the premium they would get for being engaged in research. Formally it’s the premium
for having publication, that’s the way we measured it. And this is the same, the premium for having
patents was basically zero. The next half of the sample starting 1996, you see that the premium
on research disappears. A similar story happens if
you look at the premium paid in M&A, when you have acquisitions. There used to be a premium
if you were being acquired for doing research, now there is none. Now the premium is mostly you know, the patent premium remains,
research premium is gone. What’s the point? The point here is the decline that
you saw in research output, is consistent with a decline in the private value of science, or the private value of research. Managers, investors
don’t put the same value on the company performing research. That’s what the facts say. Why that is so is a separate inquiry. You’ll hear more about it, but
here is, that’s what the facts are. A quick side note, there’s been a great
deal of hope expressed that startups will pick
up some of the slack in terms of large firms
withdrawing from research, and the story would be
small farms that’ll do it. The short answer is no. If you look at this line, that’s patents, lots of patenting from
startups, much less research. This is per firm, that’s total. Sorry this is per firm, that’s total. Whichever way you look at it, small firms, startups, they’re not going to be
able to pick up the slack. In fact if anything from
what I know about VCs, we’ll see less research
from startups in the future. VCs don’t want to be
investing in research. Okay. So, one, inference you could draw, is well maybe we’re investing less in research, because research is less useful
for corporate inventions. So, one reason it could be that that’s why they’re cutting back, and the answer here is no. How do we measure usefulness? What we’re using is
citations made by patents, corporate patents to the
scientific literature. These are called non-patent references. We’ve done it more precisely than that. We’ve literally matched it to particular, to actual publications
in the web of science. Whichever way you do it, you look at it, this line says the average age. So, this line says you’re
citing more science, not less. And moreover the science that
you’re citing is not older. It’s got the same recency. Alright, so if you could take a
patent that’s issued, a year from now, the average age of the science
cited will be a year less. What’s the take away? Science continues to be useful. Research continues to be
useful for corporate invention. Alright, these are corporate
patents that are doing it. So, corporation are using
more science not less and they’re producing less of it. They’re investing less of it. They’re producing less of it. And so, where is this science that they’re using coming from, it’s coming from the outside, mostly publicly funded research, research by other companies that could be a part of the story. So, that’s if you like
that’s that’s the point here. So, I can now just summarize
what I think we’ve found. This has been mostly just the facts. I’ve not put any interpretation,
or much interpretation. So, what are the facts? The facts are companies
are doing less research, particularly clearly as a
fraction of the R&D, and even in absolute amounts, they’re not doing more
over the last 10, 15 years, whereas the economy has
expanded by considerable point. This gap and clean up, private research is still
a big chunk of research. It’s over 30 percent. So, there is a gap. This gap will not be filled by startups. VCs are not going to
come in to save the day. Oops. Well, that wasn’t good. Research continues to be very valuable, and companies continue to use science in their inventions and they are
relying on outside science. Now, interpretation this is, this is as yet not fully
backed up by evidence, so you know, there are many
different interpretations, you’ll hear today. Here’s the one Sharon and I subscribe to, which is what we’re seeing is
a growing division of labor, between the university sector
and the corporate sector, where increasingly
corporations are going to rely on universities, government labs, publicly funded research
to do the basic science to do that’s needed and will use it, and some of it may be
mediated by startups. Okay. The policy implications
I think are quite clear. Companies are not gonna do it. Startups are not gonna do it. The only real source for this is public support for research. If we don’t have this, this proposed division of
labor just will not work. So this, these budget cuts
that were discussed, I think are come at a
particularly bad time in terms of these data. Thank you. Wesley Cohen: Very good, Thank you very much, Ashish.
(applause) Very good, we will now
hear from Stephen Merrill, who was the longtime head
of the National Academies’ Science, Technology and
Economic Policy board. We were very fortunate
to have Steve join Duke, as the Executive Director of the Center for Innovation Policy. Stephen has worked obviously for well, for decades on innovation policy, and has been really an instrumental player in any number of policies including notably the America Invents Act. With that I introduce Steve Merrill. Please, Steve. Stephen Merrill: Thank you very much, Wes. So, it’s my job to job to talk about the public investment
picture in 10 minutes, and it won’t surprise any of you that federal government spending on R&D is flat or declining. In fact as a share or ratio
of federal R&D to GDP, with the exception of one year, it’s at the lowest it’s been
since we began keeping track in the early 1950s. And with the proposed budget cuts that the senator described
of 11 percent, 13 percent, 17 percent, 22 percent, my back of the envelope
calculation is that the ratio would drop
from about 0.75 to 0.6 in fiscal year 2018. So, all the concern about
the overall science budget, and the proposed cuts
to individual agencies is entirely appropriate. Well, I want to focus my remarks on one important variant of this issue that I think may be lost or may be subordinated in this debate. So, economists have shown
that investments in ICT, information and communications technology, were a major contributor
to the economic growth and productivity spurts that occurred in the 1990s and 2000s. These investments were in turn dependent on physical science and
engineering research investments in the 1960s and 1970s that enabled the spending
on hardware and software, but since 1980 public
support of physical science and engineering research has stagnated or actually declined. In contrast to admittedly irregular, but substantial growth
in the life sciences, and I’m using entirely NSF data, and like Ashish, I’m
talking about research, basic and applied research
as distinct from development. So, if you look at the broad fields over this 35 year period, physical science and engineering research used to represent more than 40 percent of the federal portfolio, but in 2015 that was down to
a little more than a quarter. And here you see what has happened to particular subfields of research. I didn’t choose this set of six fields, biology, medical science,
computer science, materials, electrical
engineering, and chemistry. I didn’t choose that
subset of about 35 fields, purely arbitrarily or to prove a point. Instead I relied on work
that Wes has done earlier, by surveying R&D managers
and manufacturing firms, and asking them what fields of science did they depend on for the productivity of
their work in industry. And these were the six fields that were identified by those respondents. So, you can see what happened
to biology and medicine over this 35 year period,
they tripled in real terms. Computer science and material science have also increased, but
from a very small place, but chemistry and electrical engineering have actually declined in real terms. Chemistry by 11 percent, and electrical engineering
by 4 percent. Imagine, we’re spending less in 2015 on electrical engineering, than we were in real terms in 1980. Now of course, university research, is an important component of this. And, we know that university research derives support from other sources than the federal government. Principally, three sources. One, and the most rapidly growing, is the funds of the institution
performing the research. Those funds, drawn from
a variety of sources, from tuition, from grants, from individual donors and so forth. That represents almost 25 percent of university funding of R&D. Now, in this case, I’m
talking about R&D, but 90 percent of that is
basic and applied research. The other sources are much smaller. Nonprofit institutions,
principally foundations, and business investment
in university research. So, the question arises, have they helped rebalance funding. No. They’ve been even more heavily weighted toward the life sciences. One recent informal
survey of universities, by a consortium of foundations asked what their individual donor and
foundation support went to and they reported that
last year 84 percent of it went to the life sciences. So, I call this an imbalance. What do I mean, what do I
mean literally by imbalance? That there has been negligible growth in physical science and
engineering support, while life sciences support
has generally been keeping up with the growth of the economy. Now this is not a new problem, has not suddenly been
recognized as a problem. In fact there’s been a decade of effort to rebalance the federal portfolio, beginning with the Competes Act of 2007, and its reauthorizations in 2010 and 2017. The original act proposed
to double the support of physical sciences and engineering through the National Science Foundation, Department of Energy, Office of Science, and the National Institute
of Standards and Technology. The 2010 reauthorization
reduced those figures somewhat and the 2017 reauthorization abandoned the effort to double
those fields’ support. So, this is the depiction of the Congressional Research Service, of how this effort has largely dissipated, largely been a failure. First of all, because it focused on authorizations, and the similar effort
was not applied to the, in the Appropriations process, but now even the authorization route has been largely abandoned. Now, there has been a succession of White House S&T initiatives during the period of the 2000s. Not all those I emphasize have been focused on research. A lot of them on things
like STEM education, commercialization of results of research, and other aspects of S&T, and a number of them have
been in the physical sciences and engineering, a couple
in the life sciences, but I think this slide illustrates why it’s far more complicated to design the growth of physical
sciences and engineering than it is to boost the budget of the National Institutes of Health, quite apart from the fact
that biomedical research enjoys popular and political
support to a greater extent. There are more agencies involved with more diverse sources of support and of operation I should say. In-house laboratories, competitive grants, external contracts, FFRDCs, and importantly, these programs
are subject to at least three appropriations
subcommittees in each house, rather than the single
appropriations subcommittee that handles the NIH budget. Now today, in addition to addressing the challenges that the senator described, I think we should address
the question of the health of physical science and
engineering research. We could do this in part
through the rebuilding of the military and the
nation’s infrastructure. Those could be opportunities,
could be vehicles for rebalancing the
federal research portfolio, but that depends solely on
the agencies responsible. Unlike NIH, the physical
science supporting agencies generally don’t see as their mission the health of the fields
of research they support. So, it depends on a functioning Office of Science and Technology
in the White House, and a technologically sophisticated National Economic Council. Unfortunately we don’t
appear to be on the road toward either of those. Wesley Cohen: Great, thank you very much, Steve. (applause) Very good. We will now hear from Steven Freilich. Steve has a PhD in chemistry from Harvard, and is currently the Director
of Corporate Strategy for the University of
Delaware’s Energy Institute. He recently retired from
DuPont after a 33-year career in both research and business leadership, and for the last 12 years
of his tenure there, served as Director of Materials Science in the DuPont Central Research and Development Center there. And this is, so clearly Steven has a view of the issues that we’re talking about from the viewpoint of industry, and a terribly important domain, particularly materials science. With that, Steven, please. Steven Freilich: Thank you. I’m gonna give you a little
microcosm view of this. It’s not going to be so much data rich as what you’ve seen, but I’m gonna tell you what I’ve observed in the materials businesses, an industry that’s actually
incredibly important to the US economy and amounts to about three percent of the value add to the GDP and between direct and indirect jobs, employs about nine million people. This is a very vital industry that’s been extraordinarily successful
over the last 100 years, because what it does is
differentiate the products through innovation led by science. Now there are three things that a materials company needs to do in order to have science
be an important part of its business model. The first of these is that
it has to be comfortable with managing uncertainty. And it does this by having
a balanced portfolio that is balanced in
risk, time, and reward, so that you can absorb
both the psychological and financial implications
of the inevitable failures that that portfolio will exhibit. Second of all, the
industry has to understand that innovation happens
at the intersection of what’s known and what’s not known, and it’s bringing those things together that gives you something
new that you can create, but it’s very important that you can at the very front end of
your knowledge process, be able to sense what’s going on, both new inside your company, and what’s going on in
the outside environment. And then to be able to integrate that across your business units as appropriate, but really across that entire chain of knowledge development, from research to invention, ultimately to manufacturing and sales. But the markets have not
been particularly kind to the materials and
chemistry industry of late. If we look at data from
Mackenzie for 2004 to 2014, what we see very clearly is that the margins have been incredibly squeezed as products become more
rapidly commoditized. Now that drives corporate leadership to make a couple of decisions, one of which is to begin
to think about research less as the investment that it really is and more to think about it as a cost. And the other thing that it does is drive a very short-term
vision towards growth and we get these financial gimmicks of things like share buybacks to give the appearance of growth, when real growth is
not actually happening. Now this idea of short-term
view is not without its merit. If you look at what it
takes to do true innovation, where you’re dealing with new technology that is being implemented
in your product line, you can see from these data that involve polling
of a tremendous number of materials and chemical companies, you can see that the
success rates are low, as low as 15 percent, and the times to commercialization are as long as 20 years. And that’s gonna challenge the patience of most boardrooms. So what do they do? Well they become more focused on their existing products
within their existing businesses, and spend less of their research dollars thinking about how they
can create something new, because that’s significantly more risky. So, while the R&D budget may be cut by what will be considered to be a modest 10, 20 maybe 30 percent, what’s more important is
that, as you heard earlier, the shift is away from discovering the new and more towards application development, incremental changes in existing products, and those are very low risk, but they’re also low reward, and the value of them
disappears relatively quickly. But when you do that, there’s a great advantage that you have. You can eliminate the front end of this knowledge value chain. The implications of that are
pretty significant though, because it violates the fundamentals that makes science a
part of your business. You now find it much
harder to go from discovery to innovation because there’s
nothing at the beginning to either sense what’s going on or to incorporate it and
prototype it into manufacturing. You also reduce the ability
to attract new talent, and their new ideas. You heard Senator Coons
talk about the decisions that he made when he got out of college about where he wanted to spend his time. The same thing is happening
today in the materials world. People look around they see DuPont, Kodak, General Electric, Dow,
reducing their basic R&D. Where are they gonna go? They’re gonna choose different careers. And finally as I’ve mentioned before, this ability to sense what’s
going on in the early stages of development outside your
own walls begins to dissipate when there’s nobody
there watching anymore. Now, as you heard earlier, one of the potential solutions
that has been brought up is venture capital, and
people have said look, startups, university spin offs, they can do the invention. We’ll do the innovation by absorbing it. Yet again without that front end, you can’t find it as
quickly as you need to, and that actually raises
your risk profile. Secondarily, venture capital
isn’t a solution either, because venture capital
likes to get in and out, on an average of a three
to five year period for their investment. Remember the time periods
we’re talking about for developing products
with a material science. So, if you look at where venture capital has put their money
over the last 20 years, it’s gone the way of biotech and software, where you can either make
sexy hits or quick hits, and not so much in the
industrial and engineering space. And by the way these numbers are coming from the National Venture
Capital Association, and when I’m talking about
industrial and energy, this includes things like
mining, oil, things like that, well outside the material space as well. For materials alone it’s
significantly worse than this. So, no one is there to bridge
the so-called Valley of Death that exists between the invention and the commercialization
of a new product. Now, I’ve given you a rather bleak picture of the ghost of things past and
the ghost of things present. Let me try to uplift now,
assuming that isn’t a fire alarm, let me uplift now with
the way that it does, the way it can be, because what I’ve just outlined does not have to be the future, and I’ll give you an example of something in the solar industry. We all know that
manufacturing of solar modules has evolved away from the
US where it was invented and gone almost exclusively to
Asia, particularly in China. And some brilliant work
that was done by Goodrich at National Renewable Energy Lab has investigated why did this happen. And it’s not for those things that you hear in the political dialogue, it’s not about low-cost manufacturing, because of low labor costs, it’s not about the government interceding and cutting costs for
everybody in the industry to make China look better. It actually is because they
built to scale of the market, and they made their supply
chain scale and price with the industry that
they hope to create. So, while that says there’s not much you can do about that if you want to compete in that technology. The way you saw this is in
fact through innovation. Again going back to this Goodrich article, what they did was then
model a much improved higher efficiency solar panel, and assume that in the US
you manufacture to scale, and you have the supply chain manufacture to meet up with that scale, the identical technology in China competes with the US almost
head-to-head at this point, and that’s before you take
into account the shipping cost, of shipping glass from China to the US. What this does, it shows an important role for the government here, not only in reducing the private sector risk of that innovation that
you’ve now required, but also in bringing the parties together across the supply chain similar
to what is done in DARPA, where it’s not just inventing something, but bringing all the pieces
together to make sure that the military can
actually have the product in the form that it wants to have. And this is done of course through public-private partnerships. So in conclusion, what I’d like to emphasize is the fact that the short term focus
has very serious implications on the impact of science as
a way of driving innovation. We’re separating innovation
from manufacture, never a good idea, and
it reduces the ability to have science impact innovation. You’ve crippled the front end through some of these short-term decisions about where to put your research dollars, and that increases the depth and the width of the Valley of Death. It also underscores
the very important role that the federal government
has played historically and needs to continue to play, despite all of the headwinds
that we’ve heard about with the government taking the position of reducing private sector risk. I want to emphasize this. This is not about the
government doing those things that the private sector can do better. It is not about the government
picking the winners. What this is, is the government saying, this is important to
the national interest, and we are going to invest in those things that the private sector
cannot and will not do to help the private sector
enable this country to compete. And finally part of that is not only driving the basic science that’s so important to forming
the foundation of innovation through, to organizations like
the NSF, DOE, and DARPA, but also in having thoughtful
public-private partnerships that drive innovation
across the supply chain to create US jobs. Thank you very much. (applause) Wesley Cohen: Steven, thank you very much. And next we will hear from Pian Shu. Pian received her PhD
in Economics from MIT. She is currently a Professor
of Business Administration at Harvard Business School, and is a visiting scholar this year at MIT Sloan School of Management, and with that we’d love to hear from you. Pian Shu: Great, yeah it’s a pleasure to be here. I’m gonna be talking about my research on the impact of foreign competition on domestic innovation in the US, and this is joint work
with a series of co-authors at MIT, Zurich, UCSD, and
Harvard Business School. So, I wanna start by providing
a motivating evidence why manufacturing is still
important to the US economy. So, despite the fact that
manufacturing only accounts for less than 10 percent
of the US employment, now it’s still a very strong
locus of US innovation. So, even in recent years
the US manufacturing firms still account for more than two-thirds of the US corporate R&D, and three quarters of
US corporate patents. So, innovation efforts in
manufacturing sector important, and it’s very important to
understand the potential forces that could influence firms’
incentives or capability to pursue innovations. And in the last, past couple decades, they were I’d say, two strong trends that really profoundly influenced the manufacturing sector. The first is the skill
by technological change, where a lot of routine manufacturing jobs are replaced by automation, and the second one is globalization, the fact that you can have
imports from countries with cheap labor such as China. Oops, that picture didn’t
show up correctly there, And the impact of globalization
on innovation in the US is the primary motivation of our study. So, in particular, we
wanna to look at the impact of the increasing import
competition from China. So, over the past from 1991 to 2011, you can see on this graph that the share of world
manufacturing exports from China grew from two percent in ’91 to over 14 percent in 2011. So, that’s like a seven-fold increase. And we as consumers can
pretty much feel that impact. Right, if we just go to Walmart, Costco, you know any of the supermarket, we can see a lot of goods, made both from China and made in China. So, we can feel the impact of this rise of Chinese
export as consumers, but I think more importantly
in this research, we wanna ask what is the impact of that rising import competition on domestic manufacturing firms in the US? And in theory that impact
could be ambiguous. So, on the one hand, a little bit of competition could be very good for innovation, because it propels, give
firms more incentive to innovate their way
out of the competition, right, to run ahead of the curve. But on the other hand you
have a lot of competition, it also squeezes profit margin, and decreases the incentives to innovate, because you’re not making
as much money any more in the same space. So, ultimately it’s an empirical question. So, in our study, we empirically matched the
trade data to firm data, and we look at large US
firms that are listed in North American stock markets. So these, our sample is large US firms, and specifically we wanna compare within the same large broad sectors such as chemicals, computing sectors. How does different industries which experienced different
degrees of import, changing import exposure,
how do they change in terms of their growth
in the patenting effort. So, that’s our key
outcome, which is patents. And we also look at other outcomes, such as R&D expenses,
self-employment, et cetera. So, our main finding that
within broad sectors, industries that faced
more import competition, showed a significant
decline in their patenting. And within industry, if you just compare within
the industry over time, this increase in import
competition from China could explain about 15 to
25 percent of the decline in patenting over the two period, which is ’91 to ’99 and ’99 to 2007. So, it’s a non-trivial fact. And in particular we also find, and more specifically we also find, that these trade-exposed US industries also experienced declining employment, sales, profit, et cetera. So, it’s a downscale across all margins, not just the innovation output, and this effect most negatively
affect the companies, you could call, that were
kind of lagging behind. So, the ones that were
initially less profitable, and more indebted at the beginning. So, going back to the
theory we basically find that the negative side
of the effect dominates. And why that’s the case? I want to go back and
sorta put the results in a broader context of the rise of China. So, I think there are a few features of the Chinese rising
exports that are important for our understanding the results. The first is that timing of China’s rise was quite unexpected. So, in the early ’80s,
President Deng Xioaping decided to open up the border, and start the transition
to a market economy. And whether you know that effort could be successful or not, I think it was viewed as
quite uncertain at the time. In fact Wall Street Journal
had a discussion in ’89. Wall Street Journal wrote an article about predicting the new growing economy in the new decades, and they ranked China
actually as the bottom of the countries they expect
to grow in the next 10 years. So, boy was they wrong about that. And the second part is
the degree of the rise was also quite expected,
and that’s because, because of the World War II,
and then the Great Famine, and the Cultural Revolution, that all happened within
about 40 years, 50 years span. China, China was really far
from its production frontier, which means when they decide to catch up, they have a lot of room to grow. So, and that could
partially seen on the graph, as you can see this dramatic increase. And the last is that,
because of its population, China has huge comparative
advantage in cheap labor, and manufacturing, and thus
the manufacturing sector, was experienced most of the shock. So, how do we want to
interpret the results? I think these results highlight the complicated effects of globalization. And in particular, I think it speaks to two potential sources of
negative impact on competition. The first one, that when you get hit by a surge of foreign competition, especially to a large degree, it really could hurt
unprepared domestic firms, and it hurts them not only
in the innovation margin, but also all across the production margin, profit margin, all across. And the second which is
something I did not highlight in this presentation
but I do wanna bring up, because Senator Coons
also talked about it, that the ability to offshore to China and other developing countries, gives you some immediate
benefits of cost saving, but in the long term it could also hurt the innovation process, because the decoupling of
manufacturing and R&D could be detrimental to
the innovation process, especially in industries where
there’s a lot of the synergy between the two. We don’t suspect this to be the primary explanation of our results, but I think this is also
definitely at play here. I don’t think our results suggest that protectionism is an answer. It’s not that we want to
shut out these competition, ’cause you can’t, and more importantly, there are many benefits of globalization that we at this empirical
study did not focus. So, we specific look at the
impact of import competition, but we didn’t look at the
impact of for instance, as part of globalization, firms in the US also have market access
to Chinese consumers, and other countries and that
could be very beneficial, both for the firm’s profit, as well as their incentives to innovate. But I think the solar panel example from the last presentation was spot on, which really shows that the
way to counter competition is not through trying to shut them down, but it really speaks to
the ever more importance of effort to strengthen the innovation. And I think it really shows
how important innovation are, and R&D efforts are, to the core of US competitiveness. So, I wanna echo Ashish’s point, about the importance of policy effort to strengthen US competitiveness through R&D and innovation and private and public partnership. Thank you. (applause) Wesley Cohen: Thank you very much, Pian, and our last speaker will be Arti Rai. She introduced herself, but not sufficiently in my view. She is the Latty Professor of Law at Duke, and what she neglected to say is that she is truly an
internationally recognized expert in intellectual property
law, innovation policy, and administrative law. She’s taught not only at Duke, but Harvard, Yale, and Penn, and she has took time out
from her academic career to serve as the
Administrator of the Office of External Affairs at the US Patent and
Trademark Office, 2009 to ’10, and was instrumental in the analysis underpinning the patent reform legislation that became the America
Invents Act in 2011. With that, Arti, we are
pleased to hear you. And Arti, I should also say, will not only be
summarizing what we’ve said, but also really speak to the policy levers that might be considered when we think about what should be done. Thank you, please. Arti Rai: Thanks Wes. So, I will try to summarize very quickly. No need to go into great detail, because our speakers were so excellent in summarizing their own work, but I think it’s fair to say that our speakers made it clear, Ashish started out with this, that corporate R is declining and as Pian Shu’s work showed, some of this is probably due to the pressures of global competition. Steve Freilich talked about short-termism in key industries, such
as materials science. Important to note that startups are not picking up the slack. Ashish pointed that out very clearly, and as Steve Freilich pointed out, venture capital is also
short-term oriented. So, we don’t get a focus on R as a consequence of venture capital. However, and this is a
really important however, Ashish pointed this out
and all of us agree, research is still critical, and that’s just not simply our opinion, although it is our opinion, it’s the foundation for innovation. But also as Ashish’s research points out, corporate patents cite to this research. So, they too consider
this research essential. It’s just simply that they
don’t wanna do it themselves, but somebody’s gotta do
it, and that’s the problem, that we are not getting enough momentum in favor of doing it. So, what to do? So, again summarizing
what has been said so far. We need government support of research. I think that is unequivocally true. It is clearly the case that
current proposed budget cuts are gonna be a terrible idea
or would be a terrible idea, but even beyond the
current extreme situation, we need to think proactively about much better support of research, again particularly in the
physical sciences and engineering, where we have seen from
history that there are so many spillover benefits
to the economy as a whole. So, that’s one piece. Second piece, we need
public-private partnerships to bridge the Valley of Death. Steve Freilich has emphasized that point, and that point can’t be emphasized enough. Third, and this is
something that was implicit in light of the presentations, but I want to highlight it, because it’s something that Senator Coons has been heavily involved with. We often think of research as taking place in the context of efforts that are divorced
from manufacturing. Not true at all. Manufacturing innovation is critical, and manufacturing is a piece of the puzzle with respect to fundamental research. And in fact, in a lots
of different industries, manufacturing innovation
is absolutely necessary. So Manufacturing USA, which Senator Coons mentioned that he has been
instrumental in implementing, is a piece of the puzzle for solving manufacturing problems across different industries. One particular industry where
this problem is very salient, and I’ve written about this in the past, I’m just delighted that that an institute has been formed under the
Manufacturing USA umbrella to solve this problem, is this problem of
biologics manufacturing. We want faster, cheaper, better biologics and the University of Delaware has just received a $70 million grant to build and sustain innovation focused on the manufacture of biopharmaceuticals, obviously research and
manufacture are co-located, as we’ve all discussed
is really important. The partners in this case
include Amgen, Biogen, all the big companies one can think of. So, public-private is absolutely
essential in this context. Very quickly since one of
my areas of legal focus is intellectual property, I do wanna note a couple of things about how intellectual property reform may or may not play a role. So, historically, Congress
has often looked to IP in times of competitiveness
crises if you will. So, back in the late ’70s, we thought we were losing the
competition race to Japan, and we decided in that context, to pass legislation known as
the Bayh-Dole Act of 1980, to promote patenting of US
federally funded research. We also created something
called the Federal Circuit, which is an appeals court
that hears all patent cases at the intermediate appellate level. In recent years, we’ve
also seen lots of changes in the intellectual property system. So we’ve already talked
about Senator Coons’ Defend Trade Secrets Act, which I think is, on
balance, a very good thing. More recently though, there’s been concern that the America Invents Act of 2011, which, you know, full disclosure, I was involved with helping to push, maybe went too far in weakening patents. People have also been concerned that certain Supreme Court decisions on patentable subject
matter have gone too far. I would say on the IP front, we should proceed with caution. It’s obviously, IP is not the reason that we’ve had this bottoming out in terms of business research expenditure. That’s been occurring since the late ’80s, and if anything, people, most people would say
until about 2010 or so, patents tended to be low quality and tended to be asserted in ways that we’re not useful for innovation. So, I think we should proceed
with caution on those fronts, with respect to either changes that would roll back some of the work that was done in the America Invents Act, particularly the creation of the PTAB, the Patent Trial and Appeals Board, because I think on balance,
it’s working as intended, as a substitute for costly
District Court litigation on low-quality patents. And so, that it seems to me is not a place we should think of as
a legal reform arena. So, this is just some data showing how basically, the people who
are petitioning at the PTAB are those who have been sued, often on low-quality patents. And so, I don’t think the PTAB changes are necessary at the legislative level. There’s also been quite a
bit of recent discussion about changing statutory language to address Supreme Court decisions on Section 101
of the patent statute, so-called patent eligible subject matter, and we can go into this
more in the Q&A if this is of interest. I do think perhaps this may be an issue for medical diagnostics, but in general, some of
the legislative proposals that are floating around the
Hill sweep far too broadly and would take us back to an era of low quality business
method-type patents of the sort that the American Invents Act, through the installation of the PTAB, was intended to address. So that’s a little bit
of a diversion into IP, but in case that is a reaction that some people have with respect to potential legal reforms, I don’t think that’s the
place we should be focusing. So, returning to the
key policy conclusions, we need to strengthen public funding of foundational and
translational research, and I would add manufacturing
research to that list, and including through the
very important mechanism of public-private partnerships. And so with that, I think we
will get the panel up here. Wesley Cohen: Right, so let me introduce, or let me invite the panelists, please to join us here at the, on the dais here, and while they’re coming up to the table, I just want to say, make a remark, that economists, both theorists and
empirical scholars alike, have understood now for a
great more than 50 years, that the invisible hand of
purely private incentives will not work their magic when it comes to the conduct
of research in the economy. And as we’ve seen from Ashish and supplemented by other
arguments this afternoon, as public corporations have withdrawn from the conduct of the R, the research, over the past 30 years, that understanding of
economists had become, has become really all that much more true, and that much more acute when we think about the
future of the economy, and economic growth, which depends on that underlying research. With that, why don’t we
open up the discussion to questions, I see a hand in the back. And why don’t we do this. As you ask questions, if you
could please identify yourself and any institutional affiliation
you might have, please. – [Josh] Josh Sarnoff,
with DePaul University. So I wonder what, if you feel that concern for Ashish and Arti. If the problem is short
term profit maximization and short term venture asset, and you should see that in the data in a number of ways as
(muffled speaking off mic). those include the time from research to product development, the externalizing of other people using that research for other products, changes in the IP portfolio management and in the later stage
vertical integration by acquisition or merger, right? So I hope that you’ve
already done that research. If not, I encourage you to do it. For Arti, if that’s really the problem, particularly in a time
of diminished support for progressive taxation, subsidies are really hard to maintain at significant levels we want. But what you can do is
differentiate taxation so you can do it on
two bases with regard to the capital gains tax, we’ve already got people
in short and long term. You can just make the
long term really long and you can tie it more explicitly to basic science research that’s been done with that money, and similarly in regard to
product income or product sales, you can allow deductions
for that kind of research. I think that is a much more powerful and much more appropriate way
to try to go into this. It will also directly tie the moors of the business community on profit to the value that can achieve. Wesley Cohen: Okay, that there are
I guess a couple of questions buried in there. Ashish and Arti would
you like to take a step? Ashish Arora: I think the question of short-termism, and when is it, when is, so short-termism is used as a dirty word. And sometimes it’s a good
thing to be short-term, and for certain types of companies, if your pension funds
were invested in those, you would want them to be short-term. So, this is a complicated issue, and I think this, it’s a
rich one for research. Wesley Cohen: Arti, do you have one? Arti Rai: So, I think the capital gains tax issue is an interesting one, and
I can’t say that I have thought through all of the possibilities of requiring a longer holding period. I know that candidate Clinton had suggested that possibility in one of her platform proposals. I don’t know whether it would necessarily be more politically palatable at this time than some of the other things that are not politically palatable, but putting aside political palatability, yeah it’s something that
is worth thinking about. I think the whole tax
structure is not something that we focused on in this panel, but you know tax credits and
the like are, it seems to me, a very important way of, of
thinking about the problem, and perhaps more politically salable than outright subsidies. Wesley Cohen: Right. Historically tax credits, part
of the rationale for those, are the kinds of, if
you will market failure with regard to research
that has been recognized for some time. Arti Rai: And actually, one point that
is worth, sorry to interrupt. In the orphan drug context, I think it’s pretty clear that the tax credit is one of the key, whether everyone thinks that what’s happened with respect to R&D and orphan drugs is socially welfare enhancing or not, is clearly worked to incentivize R&D, and the tax credit has been a really, I think, under-appreciated aspect. Most people have focused
on the market exclusivity, but it’s actually the
tax credit that’s been much more important.
Wesley Cohen: Great, good. Other other questions, yes. – [Sayeed] My name is Sayeed Johamir, I’m on the staff of the Manufacturing Caucus office [inaudible] I have a question for you, Dr. Shu. First of all, I read
many of the papers from [inaudible] and you showed the graph of Germany [inaudible], how Germany’s position
has been relatively stable with respect to manufacturing, whereas the US has dropped
and China has picked up. Do you know why? What did Germany do to
protect itself from China or what did they do to protect themselves and basically have no job
losses in manufacturing and no loss of their funding. That would be a very
interesting example for us. Wesley Cohen: Pian, please. Pian Shu: I’m not sure about the evidence of whether manufacturing
hasn’t declined in Germany. So, the graph I showed is
a share of world exports. That hasn’t changed, although, so I can’t speak to that
direct evidence unfortunately, ’cause I don’t recall them
off the top of my head, but there is some evidence on how the Chinese competition affected European firms
differently from the US firms. So, there is a fairly
recent paper that came out using European data, so not just Germany, but UK, et cetera, et cetera, and they found that in textile
and chemical industries that European firms that
faced more competition from China actually
increased their innovation, and IT spending, conditional and survival. So, so the competition wiped
out some of the you know, relatively poor performing firms, by conditional on survival, the surviving firms had
increased innovation. So, why don’t we see that in the US, and also not in Canada. So, the empirical another
new working paper, using Canadian data
found the same evidence as we found in the US, which is that increasing competition lead to declining patenting, specifically in process innovations. So, a part of the reason I think is that the nature of the shock
is a little bit different. So, in the US it’s a very
strong unexpected shock, whereas in Europe the
degree of the competition, the degree of the import competition shock is not as strong. And the second that it also
depends on the starting level of how much competition
there is in the market. So, in the US there is
the competitive stats suggest that the US industries are generally more competitive
than European industries. So, if you start within a place where it’s not very competitive, then a little bit of extra
competition can be very helpful for incentives to innovate, because if you’re not
in a competitive state, and you have a little bit of competition, it gives you incentive to innovate, and as a way to escape competition. Whereas if you already
have a lot of competition, and then it becomes a
lot more competitive, then it’s much harder to escape competition through innovation, and seems like the
option is to just drop out. So, the nature of the
competition is very different, and also related to competition,
but the co-location aspect the way European, I think offshoring of European companies, is like they offshore to Asia, but also to eastern Europe. Right, and that distance is lot closer than China versus US. So, that could also
explain that, how you know the way globalizing and trade the shocks affect the European companies so different from US and
North American companies. Steven Freilich: May I add one point to that? I think another thing that has been done, more particularly in Germany is the public-private relationship between the Fraunhofer
Institutes and industry, the way that’s funded,
the way that’s used, and it’s designed around
specific industries, exactly as the professor said, where the competition
basis is still sound. And it helps to enhance it,
it helps to think ahead, because the inevitable competition can lead to innovation
or it can lead to death, and it’s a decision, and the better you plan through having the basic
foundations that you need, and the relationships to take
those science foundations into innovations and
ultimately into the market is gonna determine the success rate. Wesley Cohen: And, if I can add
to that Steve, mind you, the Fraunhofer Institutes
in Germany are organized on a very different basis
than the way tech transfer under the auspices of Bayh-Dole
and related legislation have led to the organization
of that interaction in the US. So, that it’s something
that’s very interesting and perhaps institutional
innovation that the US would be well-advised to
themselves adopt here. Steven Freilich: Not the least of
which is the co-location aspect. Wesley Cohen: Absolutely. There are other aspects having
to do with the incentives of the staff and within
the Fraunhofers as well, very interesting. Yes, please. – [Audience Member] Hi, I’m [inaudible] of the
National Academy, Board of Science Economic Policy. So, I’ve been hearing two models that have sort of been mixed in some
of the discussion here. One is sort of a linear
model of innovation, which is underlying a
lot of the statistics on inclusive research and so forth, and that picture says
that we should be worried. The other picture that I have seen is sort of a more ecosystem
model of innovation, and those talk about partnerships and other types of
multidimensional relationships. And first of all, do
you have some measures of how robust or how
growing this part of the, this aspect of the innovation picture is. Arti Rai: You gonna take that? – [Audience Member] and will that allow us to bypass [inaudible]. Wesley Cohen: Okay, Ashish, and then if you wanna even comment on some
work that we did before. Ashish Arora: Why don’t you. Wesley Cohen: Okay, well Ashish and
I actually did a study on what might be called the
division of innovative labor. We published this in just 2016. We surveyed, what we had response, we received responses from about six and a half thousand firms from the manufacturing sector. Our data was representative,
fairly representative. Bottom line, and the question is, to what extent do corporations when they, in the introductions of say
new products to the market, to what extent did the
underlying inventions come from outside those corporations. This speaks to the issue
of the division of labor. We were quite surprised to learn through this survey, 49 percent. So, this ecosystem that
you’ve talked about, is extremely important,
and what that means is that the innovative performance of the economy just doesn’t depend on
the existence of startups, or large firms, or whatever
kind of firm you want, but the relationships. It’s the systemic relation. It’s a systemic process
that’s very important, this ecosystem that you point to. Then let’s bring it back home
to what we just talked about in the panel today, which is, is this division of innovative labor, okay, this ecosystem, making up for the gap that Ashish with the
co-authors Sharon Belenzon and Andrea Patacconi have highlighted in the work that they presented. Ashish spoke to that a little bit, and the other speakers did as well. The point is no, okay? That certainly government contributions, research at which is the major support for university research, they, has that been active
particularly in life sciences? Sure, but has it made
up for this gap by no, in no way has that happened. And as Steve Merrill pointed out, there are even compositional issues there that should be paid attention to, and then you would think
ah, so it’s the startups, and it’s the VC’s that will save us, and we heard again from the panel today, that no that’s not the case. So, we have declining
private support for research in the nation, and public support
is not filling the breach. So, I think even from the
perspective of this ecosystem, of this feedback, this
nonlinear model, if you will, there’s something to
pay attention to here, and certainly in the
world of policy issues that should be seriously considered for the future economic
growth of the nation. Pian Shu: And can I just quickly
add a couple of our results that I didn’t discuss in the presentation. The first one is that we,
so that our main results look at the impact on US big firms, and we also have additional results looking at the impact on smaller US firms, as well as university patenting. So, those are not at the firm level, ’cause we couldn’t match the firms but I’ll skip the details. What would we find is that even if you include the
US corporate patents, regardless of large and small firms, you find the same results,
which is an industry that experienced more import competition, experienced decline in patenting. Interestingly, that’s not
true for university patents. So, non-corporate patents,
we don’t see that trend, which means that this
association is not driven by some kind of technological trends where you could argue that maybe some technological areas experience the declining opportunities, and therefore firms go out to, you know, looking for import competition, I mean looking for imports
and that’s not true. So, that’s not happening. And so that’s a, this
competition effect that we find is specific to corporate patenting. Wesley Cohen: Great, thank you. Are there other questions
please from the audience. Alright, yes please. – [Audience Member] I’m (mumbles) from the Congressional Budget Office. Question for Steve Merrill. You’ve identified a big
increase in life science funding and the building or follow up in computing and mechanical engineering. I was curious whether
some of that increase in life sciences could be
providing indirectly funding for say computer R&D, since so much research saves money for neuroscience [inaudible] rely
on the power of computing. Stephen Merrill: The first answer is
that computer sciences is sort of an outlier as
far as the physical sciences and engineering are concerned, so it’s done better but
from as I emphasized, from a relatively small base, yes NIH could and is
supporting developments in computer science and IT, but to a relatively small extent compared to other federal agencies. So, yes it’s, but these
figures are figures for research fields, not agencies. Wesley Cohen: Okay, good, thank you. Are there other questions? I had a question for Steve. Oh, there is a question, great. Oh, in the back go ahead.
– I should comment on the ecosystem that brings it back to (muffled speaking off mic) which is that trade secrecy increases the level of passive knowledge, and therefore, the search
costs to identify the labor that you wanna get is
increasingly high and similarly, the additional enforcement of noncompetes, make it really difficult
to transfer the knowledge. That’s why we can’t substitute. Wesley Cohen: Good, good, yes, Sharon Belenzon. – [Sharon] Just a quick question, Steve. If Ashish is right, and the right way to organize innovation is adequate investment in research and if you are right as well saying that if you do that, you’re gonna benefit from universities, to some hire scientists, to learn and (muffled speaking off mic) what is being done outside. How many scientists do
you think want to hire, (muffled speaking off mic) how many, it’s like 200 scientists, 20? Is that right? Steven Freilich: While I appreciate the question I think it’s an impossible one to answer, because it has to do with the strategy that you’re trying to implement. You can’t say if I have 600 scientists, they will be twice as
productive as 300 scientists. If they’re pursuing a
strategy of innovation, that’s a flawed one, And that I think is part of the challenge that corporate America faces today as they inherit an existing infrastructure of scientists ostensibly driving innovation, but it may not be properly
designed for the effort that they want to undertake. The fallacy then though, is that you can go to the
other extreme, and you can say, well I’ve now got 600,
I only need, need 50, because only 50 are productive. Well that’s an impossible equation to actually make work out. What you have to do is literally look at what you’re trying to accomplish, identify the skills that you need, identify with the outside competitive and supportive capabilities around that, and bring in what you need
or train what you have to be able to accomplish the
goals that you’ve set out. I’ve actually participated in starting an entire new business for DuPont where we didn’t
have the capability internally to do the science but we had
the manufacturing capability, if we could draw that connection. So, we had to hire people,
but it didn’t take 100. It took 15, and we got that done. Wesley Cohen: Great.
Ashish Arora: What scale? I mean so because these
numbers 15 out of like, is there a denominator here? Steven Freilich: How should, I think it depends on whether you’re trying to do percentages or not. If you’re looking at the
absolutes of what you need– Ashish Arora: How big was the business? If it’s not going to, in annual revenue? Steven Freilichy: All businesses start small. Ashish Arora: No, no, but 15 to support
a business of what size? Steven Freilich: It would have been a product line of three distinct products
that were all headed towards the same value
chain, but the key thing is– Ashish Arora: No I meant dollars. Steven Freilich: I don’t have that number. I left before I could
get those dollar figures, but the key thing is that if you have the manufacturing capability
in place and your R&D is now designed to be able
to connect the market need to the manufacturing that you have with new science that you’re bringing in. It actually gives you a
real leg up on being able to be strategic and accomplish the goal. Wesley Cohen: Great, good. Other questions? In back and then in front. So in back please, name and affiliation – [Jimmy] Jimmy [inaudible] I am a policy fellow in
the office of Senator [inaudible] My question is, is it better [inaudible]
research, mostly basic research. This has two parts. The main part is, what is the balance, for example, in relation to, for the government to invest more in applied research or basic research. (muffled speaking off mic) And the other question is
(muffled speaking off mic) you talk about research in general, but most of the decline, I wonder, and also, is being in applied research, has been in basic research, and which should have more
influence in innovation. Wesley Cohen: Okay. Want to take it? Ashish Arora: So, there’s this sort of, if
you look at the NSF numbers, they do distinguish
between basic and applied. I’m happy to, the folks at NSF, my understanding is that the distinction between basic and applied
in the NSF numbers is much less reliable,
reliably measured consistently, than if you just look at
research versus development, so that’s one. We could look at the evidence
we have on publications, and look at where the
publications come in. Do they come in in more basic journals, or more applied journals? And you could do it, again this is not an easy thing to do, but you could look at impact factors. Alright, so if you think
general interest journals have higher impact factor. The short answer there is, most of the decline that we
see in corporate publications, is in the basic research site,
or the more basic journals. That’s why you see the decline,
not in the more applied. Those are the facts. What the right balance is? I’ll leave it to my superiors to say. (laughter) Wesley Cohen: And I’ll leave it to my superiors. That’s a hard problem too, as is I think you already understand to really put a number to certainly. I think what we can say is that we’re probably not at the right
balances, our expectation. Would others like to speak to
that very difficult question? There was another question
right in front here please. I’m sorry, please. Arti Rai: So, I just wanted to make one point, which is I think that in terms
of at least the presentation this panel is making, including Ashish, what we’re all more concerned
about is R versus D, and R is in general, is
an area of great concern, whether basic or applied, and so the fact is that it’s
hard to draw distinctions but that entire area is a problem. Wesley Cohen: Okay, three rows back, then next two rows back, please. Go ahead. – [Audience Member] [inaudible] German Embassy, Thanks for making it possible. I would like to raise the question about, and fill the gap. You were mentioning venture capital and startups as probably not
being able to fill the gap. What about the states? What about philanthropy? What are things that can
come from philanthropy? And may there be a
shock for companies if they see the budget figures, that they will again
start invest in R&D because they can no longer
rely on the public money. Wesley Cohen: Yeah. Steve, go ahead. Stephen Merrill: So, states are important but a very small portion
of research funding. They would fall in the category of own institution funds that I showed. Their predominant field of research is biomedical just like
philanthropy is, so it’s, again, an even greater imbalance. Wesley Cohen: And if I can take the
last part of your question. The point is is it the lags are enormous between the
conduct of upstream research and its material benefits
in the marketplace. Indeed, what an article published in 1990 that was quite nice by Jim Adam showing the lags of what he considered for his purposes basic research, and that study showed lags
on the order of 20 years. So the point is, is by
pulling down funding now, we’re doing something that’s probably, in any, you know, meaningful way, irreversible in many of its effects. And so given those very long lags, we need to pay attention now, because that’s not a problem that’s easily fixed down the road. Please. Steven Freilich: Actually before we go on, I’m sorry, there’s one other important point, which is we tend to think
about the Valley of Death as being a cliff that you fall off on, never to climb back out again. There are some incremental
bridges that can be built, and there was an important study done, I believe the author was Sabrina Howell, where she showed that at the right time, with an injection of about $150,000 in materials innovation
in the energy space, you can do a world of good, that maybe later on, a million
dollars wouldn’t help with. Wesley Cohen: Good, please in front. – My name is Rosalyn Suggero. (muffled speaking off mic) focus on startups and
the innovation manufacturing. Looking at your research, and you mentioned about China, he’s talking about the Germans. How do we hope that in
the national research of looking at the actual innovation and research together so that we can exchange the ideas? I was only making a medical claim, because eventually, we will come in and the other companies,
(muffled speaking off mic). I have had so many Germans
come in on innovation. (muffled speaking off mic) So how do we make that in
the national and global area on looking at financing,
investment, and other things? Thank you. Wesley Cohen: Does anybody want to take it? I’m not sure. So the question is then, I’m not exactly sure what
the precise question is. – [Rosalyn] Is the person that comment how do we bring in the
international researchers into what we are doing in
innovation, manufacturing, and looking at maybe an
international company or profession would say, yes
I would fund the research. If I fund the research,
or finance the research, would you refuse because
it’s not American company? (both talking at once)
Wesley Cohen: I see, I see, okay. Ashish Arora: Knowing what I know about
American Universities, we would not refuse money from anybody. (laughing) Wesley Cohen: Yes, and I would absolutely
endorse that observation, so I think that’s not,
would not be a problem. But go ahead. Pian Shu: I think one interesting observation that I think Chinese startups have been the recipients on the beneficiary side of international innovation, so there a lot of companies, you know like for a period in time for the Chinese internet companies, the motto is, you know, copy after the successful US firm, and obviously, you have a ton of copycats, but one person is gonna get lucky, and they’re gonna be doing very well in the Chinese internet space. I think that motto is not probably as prevalent now as before. Interestingly, I can’t
think of a counter example from the US side, which is a US company copying a Chinese internet company
and doing well in the US. I can’t think of that. But there are many examples
on the Chinese side. Wesley Cohen: Great, and yes please. – [James] Hi yes my name is James [inaudible]. I am with the American (muffled speaking off mic) I just wanted to put off something that I was listening for
but I didn’t hear it, the impact of future workforce. If you decline university research, then you decline the opportunity to teach the workforce that will
be coming down the line, because with the rapid
changes to innovation globally, if you don’t, the student that’s coming as a professional today
needs to be taught by people who are about five, six, 10 years ahead. But I just wanted you
guys to comment on that because I haven’t heard
anything specific on that. Wesley Cohen: Steve, I think you
spoke to that a little. Arti Rai: Yeah. Steven Freilich: Yeah. I think your comment in itself is the wisdom of the group, so thank you. One of the things that
I think is going to be extremely helpful here
is not to think about these disciplines as pure disciplines. One of the things that you see is is mathematicians starting to evolve into, as opposed to going into the
fields of math or modeling, going into the financial sector. I think one of the ways that you break that trend out of the science
and into the finance world is by recognizing the multi
and interdisciplinary nature of the world that we live in today, and beginning to tailor
the education that way, it becomes much more
attractive to students, and it also becomes attractive
to industry as well, because now they can become part of that learning process that’s
starting at the universities, and has a real home in industry. Wesley Cohen: Steve? Stephen Merrill: I just want to say a word
about the workforce issue. So, it depends on what aspect of the technical workforce
you’re talking about. If you’re talking about the skilled but non-baccalaureate technical workforce, I refer you to a report that
just came out of the STEP Board at the Academy on that subject. If you’re talking about the graduate level and PhD technical workforce, then there’s a very close correspondence between federal spending in those fields and the number of people being trained, and that’s simply because we
support graduate education predominantly through
research associate positions. Wesley Cohen: We have time, I gather, for maybe just one more question and then we’re going to have to call it. You’ve had your hand up for a few moments. – [Will] I wanted to ask about
this idea of optimality. We talked about it a little bit. So, first off, my name is Will Reinhart. I work for a think tank here in DC called the American Action Forum. I do a lot of research
in this space as well. One of the things that I was
very surprised last time, I saw the 2015 numbers come out of the NSF is that essentially, we really haven’t changed all that much. When you do break out the
research and development, you look at say the 1950s,
it’s right around 30 percent, and today, it’s a little
higher, like 36 percent. So I’m wondering, and especially now when
you actually look at, as a percentage of GDP, that we’re actually at
about all time high levels. I’m just wondering what, I mean, we haven’t really talked about this idea of optimality, but I’m just wondering what
specifically should we, what numbers should we
actually be looking for when we’re talking about a percentage, or what types of statistics should we be trying to understand? Because we talk about patentability, that’s kind of a, that’s again, Arti set a little bit of information about the problems with that, and a lot of people
have said there’s issues with trying to do this
percentage of our GDP. What kinds of data do we need, and similarly, what kinds of things that should we be directed
to be looking for. Because everyone says there’s a problem. When I look at the data, there doesn’t seem to
be a huge dramatic … I mean, there are shifts clearly, but there’s, the data doesn’t show this huge dramatic
problem we need to solve. Stephen Merrill: I’m not sure what you’re referring to when you say 30 percent. – [Will] When you look at
R&D as a percentage, the changes haven’t, I mean– Wesley Cohen: Percent, what’s your denominator?
– [Will] What’s your denominator… Wesley Cohen: When you say 30 percent? – [Will] When you look
at R&D as compared, so when you look at R&D as just a percentage of all R&D, and you look at what their
changes have been over time, I mean, this is the problem
with the NSF numbers is I know that there’s this real delineation between the two, but when you look at the percentage between R and then D, it doesn’t seem to change
all that much over the last– Wesley Cohen: Ashish, what do you want to say to that? Ashish Arora: So you’re right. If you look at the very long-term, there’s a kind of an inverted U, and we peak around some, late ’70s maybe mid ’80s or even ’90s, and so since then we’ve
been declining dramatically. So, I’m of two minds. One way is, the 1950s
were very different right? America, America had, if you look at the state of
American universities in 1950s to where we are now, there’s been a massive expansion of American research capability. In 1950, arguably, Europe was much closer, in many cases far ahead
of US research capability, which is not the case now. So today, in that sense, I don’t think 1950 is a fair comparison. If you compare with 1980s, for example, it’s been a steady down, the share of research
has steadily declined. I’m thinking specifically
now about private, in the private sector, but it may well be also
in the public sector. Wesley Cohen: Yes.
Arti Rai: Steve’s data show that. Wesley Cohen: Yeah. Good. Well, on that note, I’d like to thank all
of you for coming today, and I’d like to thank
the panelists as well for their presentations and thank you. (applause)


Reader Comments

Leave a Reply

Your email address will not be published. Required fields are marked *