|
06.09
Federal
Funding of Basic Research: Who Needs It?
By Dr.
James Gover, IEEE Fellow
It is widely accepted that a
highly trained work force, capital investment
and technology innovation are important inputs
to international economic growth. However, what
is not well understood is the most effective
role for federal, state and local governments to
fill in promoting each of these inputs to the
global economic system. For example, in the
United States, states fund public high schools,
community colleges and universities to provide
work force education and states offer tax
benefits, construction property and specialized
work force training valued up to $350,000 per
job created for companies to build or move
manufacturing or services facilities to their
state.
Despite the financial support of
states for public universities, in most rating
systems the highest rated U.S. universities are
private institutions that receive no state
funding. However, for the most part, these
private universities depend upon the federal
government to fund their research efforts. The
intent of this paper is to explore the optimum
role of government in promoting technology
innovation that leads to job creation. An
implicit assumption of those that argue for more
federal spending for basic research is that
doing more of what we have already been doing
will somehow get more economically relevant
results than we have been getting. It should be
clear to the reader that the currently depressed
international economic environment requires all
nations to examine how public funds are spent
and to maximize the economic return to the
public.
In addition to its regulator
function, funding of research at universities,
government owned laboratories and companies is
another way that the federal government promotes
innovation. The question, “What type of federal
funding is most effective in promoting
innovation?” is particularly difficult to
address. Despite passage of the Government
Performance and Results Act introduced by
the late Senator Roth, federal R&D agencies
always claim success and rarely apply objective,
quantitative metrics to determine the return to
the public of their research investment. Outputs
rather than outcomes are usually emphasized.
Consequently, public outcome data are sparse.
Surveys of the semiconductor manufacturing
equipment sector made during the first Clinton
administration revealed that well over 90
percent of the U.S. companies in this sector
were not impacted by any federal research
program and among the few that had benefited
from federally sponsored research, no federal
research program stood out as more successful
than others in promoting innovation in the
companies.
Federal R&D programs also tend
to focus only on the science and technology of
problems that are addressed. Unlike private
sector programs that fail, federal R&D programs
fail without any consequence to the agency that
led the program. For example, by focusing on
only the technical dimension of nuclear waste
storage and not the political dimension, the DOE
has spent $13 billion building a nuclear fuel
rod storage facility in Nevada that is almost
certain to never be used simply because the
citizens of Nevada do not want it. In fact, it
has been zeroed out in the proposed Fiscal Year
2010 budget. This program was a technical
success but a political and economic failure.
The public really does not care
what type of research Congress funds as long as
(a) it creates jobs at a public expenditure rate
less than $350,000 per job, promotes economic
development and provides for the common defense
or (b) the research funds are spent in their
state. Obviously defense funding and economic
development are linked; economic growth is
required in order for tax revenues to be
available to fund defense. The Defense Advanced
Research Projects Agency (DARPA) is particularly
favored by Congress because the research DARPA
has funded has led to numerous discoveries that
advanced defense as well as spin-off technology
that is economically relevant. Most noteworthy
of DARPA’s innovations was funding research that
eventually led to creation of the Internet which
has evolved to become a major industry sector.
DARPA’s motivation in developing the DARPA Net
was to promote electronic file sharing among
researchers it funded so that the productivity
of its defense research could be increased.
Congress has attempted, without success, to
clone DARPA in both the Department of Energy and
Department of Homeland Security. It is the
opinion of the author that DARPA’s success is
almost entirely attributable to management
innovation, the most important type of
innovation
according to respected business and management
authority,
Professor Gary Hamel, who addressed "Continuous
Management Innovation" in a 2006 Fortune
Innovation Forum. Unfortunately, the flat
management structure of DARPA is difficult for
most agencies to implement.
Republican members of Congress
like the idea of federal funding for basic
research because support for basic research is
widely accepted to not be “industrial policy”
that results in government picking “winners and
losers.” Democrat members of Congress, while not
as concerned about “industrial policy” as their
Republican colleagues, like the idea of funding
basic research if only because predominantly
democratic university researchers lobby for it.
Both parties generally accept the idea that
basic research is weakly appropriable so
companies are thought to under invest in it. The
fact is that companies invest in the basic
research that is needed for success in the
marketplace and if the company wants to hide its
basic research from other companies, it is able
to do so for an amount of time sufficient to
gain product or process leadership. The only
basic research that is weakly appropriable is
basic research conducted at a university and
that is because the tenure and promotion process
of universities require tenure-track faculty to
publish.
Most members of Congress of both
parties also at least implicitly believe in the
validity of the linear model of innovation in
which basic research is believed to be the first
step in a process that evolves from basic
research to applied research to product
development. The linear model of innovation is
inadequate because: (1) it has no feedback paths
linking basic research to manufacturing and
marketing; (2) the central process of innovation
is not science but a design that often draws on
science, but also forces the creation of
science; (3) the notion that innovation is
initiated by research is wrong most of the time;
and (4) innovation occurs when inventions are
transformed into socially relevant products or
products even in the absence of science.
Technology innovation has been
studied by scholars and the model that is
preferred by these scholars is the chain-link
model. This model is far more complex than the
linear model. In fact, if one accepts the
chain-link model it calls into question the way
most federal agencies select research projects (DARPA
is the exception) and the way most federal
agencies are organized to manage their research
projects (again, DARPA is the exception). The
model of innovation proposed by Kline and
further refined by Rosenberg that is generally
accepted to be representative of how innovation
really works is illustrated in Figure 1.
Note that while research is an
important element of the chain link model of
innovation, it is not the only element and it
likely is not the most important element. While
the linear model of innovation starts with basic
research that leads to a new science or
technology discovery, the chain-link model
starts with a market in mind. The market may be
for a product that serves the purposes of U.S.
defense or it may be a market for a commercial
product. In the chain-link model as one moves
along the path toward marketing and research
needs are identified, if that research is not
available, then a firm or agency may conduct the
needed research.
Figure 1: The
Klein and Rosenberg Chain Link Model of
Innovation as Depicted by Phillips.[i]
In my 25 years working as an
engineer or engineering supervisor at Sandia
National Laboratories, there were several
occasions while working on development programs
that a problem was encountered that required
basic research, so if we could not find the
research results in the literature, we did the
basic research at an accelerated pace. Other
times the basic research needed for a
development project was anticipated in advance
and the basic research was available in a timely
manner. It was my experience — and my experience
is consistent with the chain link model of
innovation — when research organizations were
isolated from development projects, all too
often the research outputs, although highly
publishable and technically impressive, were
irrelevant to product development. This outcome
has been experienced by most firms;
consequently, large central research
organizations have been dismantled and research
has been integrated into firms’ development
projects. This was an appropriate decision made
by companies and it did not create a deficiency
that the federal government needs to correct.
So what kind of research should
governments fund? Obviously, government should
fund research relevant to national defense and
other government missions; however, special care
must be taken to link these research activities
to where the research is needed. DARPA stands as
proof that this method not only serves national
defense, it results in spin-off of commercially
relevant products and processes.
Funding of basic research at
universities is often justified on the basis
that it results in the education of scientists
and engineers and it does educate a few;
however, funding universities to develop courses
and to put them on the world-wide web would
likely educate more BS level working scientists
and engineers than conducting basic research
which only educates Ph.D. candidates (only 25
percent of U.S. Ph.D. candidates in engineering
are U.S. citizens.). Of course, putting
educational innovations on the world-wide web
would raise the level of international
competition and accelerate product and process
innovations around the world, not just in the
United States; however, the U.S. public should
benefit from the competition. It should be noted
that one of the world’s greatest engineering
schools, MIT, already displays the notes from
many of its electrical engineering courses on
the web.
Another justification is that
conducting basic research keeps professors at
the cutting edge of science and technology and
this elevates the quality of their teaching;
however, conducting applied research or
development programs also helps keep professors
at the cutting edge and their work could also
provide practical education experiences for
undergraduates.
In summary, I think that the
public would benefit more from the following
than it benefits from funding stand-alone basic
research: (1) government fund development
projects within the mission portfolio of the
federal government provided these funds are
sufficient to support the research, including
basic research, needed to successfully complete
the development project; (2) government fund the
development of science and engineering courses
that are available on the world wide web to
professors and students around the world and (3)
government enforce the Government Performance
and Results Act.
References
[i] Peter W.B. Phillips,
“Innovation and the entrepot cluster,”
www.utoronto.ca/isrn/publications/NatMeeting/NatSlides/
Nat04/Phillips04_Entrepot.ppt. Note that
this is only one of many references one may find
on the web which serve to illustrate that
nations lacking the semi-infinite pool of
research funds present in the United States have
made use of the chain link model to determine
how programs their governments’ sponsor may spur
economic development. The work reported in this
particular reference was done by Professor
Phillips at the University of Saskatchewan to
determine the best way for the Saskatchewan
government to support the Saskatoon biological
entrepot.
Dr. James Gover is an IEEE
Fellow and a professor of electrical engineering
at Kettering University in Flint, Mich.
Comments on this article may
be submitted to todaysengineer@ieee.org.
Opinions expressed are the
author's.
|
E-mail
this page to a friend
