The Economics of Science and...

The Economics of Science and Technology

David B. Audretsch1

Barry Bozeman2

Kathryn L. Combs3

Maryann Feldman4

Albert N. Link5

Donald S. Siegel6

Paula Stephan7

Gregory Tassey8

Charles Wessner9

ABSTRACT. This paper provides a non-technical, accessibleintroduction to various topics in the burgeoning literature onthe economics of science and technology. This is an interdisci-plinary literature, drawing on the work of scholars in the fieldsof economics, public policy, sociology and management. Theaim of this paper is to foster a deeper appreciation of the eco-nomic importance of science and technology issues. We alsohope to stimulate additional research on these topics.

JEL Classification: O3

1. Introduction

Science and technology have long been regardedas important determinants of economic growth.Edwin Mansfield (1971, pp. 1–2), a pioneer in theeconomics of technological change, noted:

Technological change is an important, if not the mostimportant, factor responsible for economic growth � � �

without question, [it] is one of the most important determi-nants of the shape and evolution of the American economy.

1Indiana University2Georgia Institute of Technology3University of St. Thomas4Johns Hopkins University5University of North Carolina at GreensboroGreensboro, NC 27402E-mail: [email protected] of Nottingham and Rensselaer Polytechnic Institute7Georgia State University8National Institute of Standards and Technology9National Academy of Sciences

Science and technology policy are even moreimportant in the “new” economy, with its greateremphasis on the role of intellectual property andknowledge transfer. Thus, it is unfortunate thatmost individuals rarely have the opportunity tostudy this subject. As a result, the general pub-lic poorly understands the antecedents and conse-quences of technological change.

It is clear in the report that most Americansare not well-informed about public policy issuesrelating to science and technology. As shown inTable I, individuals rank science and technologypolicy issues relatively high in terms of interest,yet noticeably lower in terms of their self-assessedknowledge about the issues. However, it is pre-cisely these issues that may be most critical indetermining long-run economic growth.

The purpose of this paper is to provide anoverview of salient topics in the economics ofscience and technology. We devote considerableattention to historical and institutional informa-tion concerning these issues, because we believethat an understanding of the current situationdepends to a large extent on an understanding ofhow this literature and the institutions that sup-port science and technology evolved.1

The remainder of this paper is organized asfollows. Section 2 provides some initial defini-tions. In Section 3, we summarize major pub-lic policy initiatives toward science and technol-ogy in the United States from the colonial periodto the present. Our up-front emphasis on policyunderscores the subtlety of partnerships involvingthe public and private sectors that have emerged

Journal of Technology Transfer, 27, 155–203, 2002©2002 Kluwer Academic Publishers. Manufactured in The Netherlands.

156 Audretsch et al.

Table IIndices of public interest and self-assessed knowledge in

selected policy issues, 1999

Publicinterest Policy issues Informedness

82 New medical discoveries 5371 Environmental pollution 4871 Local school issues 5867 Issues about new scientific

discoveries 4465 Use of new inventions and

technologies 4365 Economic issues and business

conditions 5064 Military and defense policy 4455 Use of nuclear energy to

generate electricity 2953 International and foreign policy 4051 Space exploration 3747 Agricultural and farm issues 33

Source: Science and Engineering Indicators—2000, AppendixTables 8-2 and 8-5.

over the last three centuries. These “public/privatepartnerships” have evolved from government’sdesire to steer private investment towards certaintypes of scientific activity and the developmentand use of new technologies. Thus, the federalgovernment has attempted to establish an environ-ment that is conducive for private sector invest-ment in research and development (R&D), as wellas one in which the public and private sectors canbe partners in undertaking innovative activity.

Section 4 emphasizes the role of technology ineconomic growth and sets the stage for under-standing the scope of science and technology pol-icy mechanisms used to maintain national growth.Fundamental to all such policy instruments isthe relationship among investment in R&D, tech-nological advancement, and economic growth.Dimensions of R&D are described in Section 5.The following section introduces the second partof the primer by emphasizing the entrepreneurialnature of firms, both to innovate and to respond topolicy initiatives. Section 6 advances an economicrationale for government’s role in the innovationprocess. An articulation of such a role has longbeen absent from science and technology policydebates. Having set forth this rationale, four pol-icy mechanisms are discussed from a historical aswell as an economic perspective: patent laws in

Section 7, tax incentives in Section 8, research col-laborations in Section 9, and public/private part-nerships that subsidize research in Sections 10 and11. These mechanisms, and their relationship tofirm behavior, are summarized in Section 12.

All of the policy mechanisms discussed in Sec-tions 7 through 11 are designed to stimulate theprivate sector’s demand for R&D resources. Fordemand mechanisms to be effective there mustalso be a supply response in terms of the edu-cation and increased availability of scientists andengineers. This is the topic of Section 13. As withall public-sector initiatives, there is also an issueof public accountability. How effectively are publicresearch funds being spent? The history of publicaccountability and related evaluation methods aredescribed in Section 14. Section 15 constitutes oursummary statement regarding the current state ofthe field of economics of science and technologyand future directions.

2. Fundamental concepts

As in any new field—and we view the economicsof science and technology as an emerging field thatdraws on concepts from numerous disciplines—there are several fundamental concepts. Thus, webegin with several definitions.

In everyday conversation, terms such as scienceand technology, as well as invention and innova-tion, are often used interchangeably. However, foracademics and policymakers there are importantdistinctions that give each of these terms a uniquemeaning.

Science, in a broad sense, is the search forknowledge, and that search is based on observedfacts and truths. Thus, science begins with knownstarting conditions and searches for unknownend results (Nightingale, 1998). Technology is theapplication of new knowledge learned throughscience to some practical problem. Technologicalchange is the rate at which new knowledge is dif-fused and put into use in the economy.

Closely related to science and technology arethe concepts of invention and innovation. Follow-ing Bozeman and Link (1983, p. 4):

The concepts commonly used in connection with innovationare deceptively simple. Invention is the creation of some-thing new. An invention becomes an innovation when it isput in use.

The Economics of Science and Technology 157

Innovations may be new products, new processes,or new organizational methods that are novel andadd value to economic activity. Thus, inventionparallels the concept of science and innovationparallels the concept of technology.

It is useful to think of an innovation as some-thing new that has been brought into use. Thus,this innovation represents, in a sense, a newunderlying technology.2 Embedded in this distinc-tion between invention and innovation is a processwhereby inventions become applied. This pro-cess is central to what we call entrepreneurship.Entrepreneurship is a process involving the organi-zation of resources, and the output of that processis an innovation.3 Of course, for entrepreneurshipto have economic value the resultant output orinnovation must have economic value.

From an economic perspective, the conceptof entrepreneurial innovation can be traced backto the Physiocrats in France in the mid-1700s.Baudeau (1910, p. 46) referred to a processguided by an active agent, which he called anentrepreneur, within a capitalistic system4:

Such is the goal of the grand productive enterprises: firstto increase the harvest by two, three, four, ten times if pos-sible; secondly to reduce the amount of labor employedand so reduce costs by a half, a third, a fourth, or a tenth,whatever possible.

Embedded in this conceptualization of entrepre-neurship is the notion of an innovative process,one perhaps as simple as the perception of newtechnology adopted from others so as to increaseagricultural yield, or one as refined as the actualdevelopment of a new technology to do the same.When the process is completed, and when theinnovation is put into use, there will be an increasein productivity, and possibly, substitution of capi-tal for labor.

We have defined entrepreneurship as a process:an output is the promotion of one’s own inno-vation or the adoption of another’s innovation.The term entrepreneurship is commonly used torefer to a businessman or even to a risk taker.We use the term entrepreneur in a much broadersense; an entrepreneur is one who perceives anopportunity and has the ability to act upon it.Hence, entrepreneurship is a process that involvesboth perception and action. The perception ofthe opportunity may be influenced by changes instrategic directions or competitive markets, but

perception of the opportunity is the fundamentalfirst step. The consequent step is the ability to acton that perception. What defines the entrepreneuris the ability to move technology forward intoinnovation. The technology may be discovered ordeveloped by others. The entrepreneur is able torecognize the commercial potential of the inven-tion and organize the capital, talent, and otherresources that turn an invention into a commer-cially viable innovation.

What are the requisite resources needed foraction, which takes the perception of an opportu-nity forward to result in an innovation? One obvi-ous answer is research and development (R&D),that is, the commitment of resources to inven-tion and innovation. R&D not only provides astock of knowledge to encourage perception butalso the ability for the firm to foster action. How-ever, firms that do not conduct R&D can still beentrepreneurial, as discussed above. In such firms,innovations are likely to be introduced rather thanproduced. Such firms act in an entrepreneurialmanner by hiring creative individuals and provid-ing them with an environment conducive for theblossoming of their talents.

Consider R&D-active firms. The R&D theyconduct serves two general purposes. First, it pro-vides the resource base from which the firm canrespond to an opportunity with perceived strate-gic merit or technical opportunity that allows thefirm to develop a commercial market. Second,those scientists involved in R&D are the inter-nal resource that facilitates the firm’s being ableto make decisions regarding the technical meritsof others’ innovations and how effectively thoseinnovations will interface with the existing tech-nological environment of the firm. The firm maychoose to purchase or license this technology orundertake a new R&D endeavor. In this lattersense, one important role of R&D is to enhancethe absorptive capacity of the firm.5

Thus, the role of R&D in enhancing theabsorptive capacity of the firm goes beyond simplyassessing the technical merits of potentially pur-chasable technology. It allows the firm to interpretthe extant technical literature, to interface whennecessary with the research laboratories of oth-ers, in a research partnership relationship, or toacquire technical explanations from, say, a federallaboratory or university laboratory; or simply tosolve internal technical problems.


Strategic direction of the firm

Entrepreneurial response Innovation Value Added Competitive market conditions

Figure 1. The entrepreneurial process: an initial look.

Figure 1 provides an initial view of what weterm the entrepreneurial process. This initial anal-ysis will be expanded upon, but here it introducesthe following concepts:

• The organization, typically a private firm, has afocus that results from its agreed upon strate-gic direction. This strategic direction, coupledwith competitive market conditions, generatesan entrepreneurial response.

• The purposive activity associated with theentrepreneurial response leads to an innova-tion.

• There are market forces at work that are, inpart, beyond the influence of the firm andthese forces determine the economic value ofthe innovation and hence the value added tothe project as well as to the user of the inno-vation.

There is a subtle distinction between entrepre-neurship or the innovation process and the processof science and discovery. As noted above, sciencemoves from starting conditions toward unknownresults whereas the innovation process starts withan anticipated intended result and moves towardthe unknown starting conditions that will pro-duce it.

3. Historical background6

The history of science, technology, and economicgrowth in the United States was greatly influencedby the scientific discoveries and university infras-tructure within Europe at the time of colonization.While difficult to pinpoint how or which specificelements of scientific and technical knowledge dif-fused across the Atlantic, certain milestone eventscan be dated and pivotal individuals can be sin-gled out. This background gives us an appreciationfor the role that science and technology resourcesplayed in the developing American nation andcontributed to shaping the preeminent role thatthe former colony achieved.

An understanding of these historical eventsand players is important because it allows one tounderstand the environment in which innovationtakes place and the genesis of the assumptionsthat underlie the public policies that influence thisenvironment. It also illustrates the evolving rolethat the government has taken in promoting sci-ence and technology.

The colonial period

The first member of the Royal Society of Londonto immigrate to the Massachusetts Bay Colonywas John Winthrop, Jr. in 1631, just a few yearsafter the founding of the Colony. As a scientist,he is credited with establishing druggist shops andchemistry laboratories in the surrounding villagesto meet the demand for medicine. According toUNESCO (1968, p. 9), these ventures were “per-haps the first science based commercial enterpriseof the New World.”

Before the turn of the eighteenth century,colonists made noticeable advances toward whatmay be called a scientific society, organizing sci-entists who came from England and other Euro-pean countries into communities that promotedscientific inquiry. In 1683, the Boston Philosoph-ical Society was formed to advance knowledge inphilosophy and natural history. Benjamin Franklinformed the American Philosophical Society ofPhiladelphia in 1742 to encourage correspon-dence with colonists in all areas of science. Itlater merged with the Franklin-created AmericanSociety to promote what Franklin called “use-ful knowledge,” and it still exists today. Thiscombined Society focused on making availableadvancements in agriculture and medicine to allindividuals by sponsoring the first medical schoolin America (also supported by the Pennsylva-nia House of Representatives). Thus, Franklin’sSociety was a hallmark of how public and privatesector interests could work together for the com-mon weal.7

Influenced by the actions of Pennsylvania andlater Massachusetts with regard to sponsorship ofscientific institutions, the establishment of nationaluniversities for the promotion of science wasfirst discussed at the Constitutional Conventionin 1787. However, at that time the founders of


the Constitution believed educational and scien-tific activities should be independent of directnational governmental control. But, they felt thatthe national government should remain an influ-ential force exerting its influence through indirectrather than direct means. For example, Article I,Section 8, of the Constitution permits the enact-ment of patent law:

The Congress shall have the power � � � To promote theprogress of science and useful arts, by securing for limitedtimes to authors and inventors the exclusive right to theirrespective writings and discoveries.8

However, Thomas Jefferson championed amore direct role for the government in the areaof science. While president, Jefferson sponsoredthe Lewis and Clark expedition in 1803 to advancethe geographic knowledge of the nation, thus mak-ing clear that “the promotion of the general wel-fare depended heavily upon advances in scientificknowledge” (UNESCO, 1968, p. 11). In fact, thisaction by Jefferson set several important prece-dents including the provision of federal funds toindividuals for scientific endeavors.

Although the Constitution did not set forthmechanisms for establishing national academicinstitutions based on the founders’ belief that thegovernment should have only an indirect influenceon science and technical advancement, the needfor a national institution related to science andtechnology was recognized soon after the Revolu-tionary War. For example, West Point was foundedin 1802 as the first national institution of a sci-entific and technical nature, although Connecticutestablished the first State Academy of Arts andSciences in 1799.

In the early 1800s, universities began to empha-size science and technical studies, and in 1824Rensselaer Polytechnic Institute was founded inNew York State to emphasize the application ofscience and technology. The American Journal ofScience was the first American scientific publica-tion, followed in 1826 by the American MechanicsMagazine.

The social importance of the government hav-ing a direct role in the creation and applicationof technical knowledge was emphatically demon-strated in the 1820s and 1830s through its supportof efforts to control the cholera epidemic of 1822.Also during that time period federal initiatives

were directed toward manufacturing and trans-portation. In fact, the Secretary of the Treasury—the Department of the Treasury being the moststructured executive department at that time—directly funded the Franklin Institute in Philadel-phia to investigate the causes of these problems.This action, driven by public concern as well asthe need to develop new technical knowledge, wasthe first instance of the government sponsoringresearch in a private-sector organization.

In 1838, the federal government again took alead in the sponsorship of a technological innova-tion that had public benefits. After Samuel Morsedemonstrated the feasibility of the electric tele-graph, Congress provided him with $30,000 tobuild an experimental line between Baltimore,Maryland, and Washington, DC. This venture wasthe first instance of governmental support to a pri-vate researcher.9

Public/private research relationships continuedto evolve in frequency and in scope. In 1829,James Smithson, gifted $500,000 to the UnitedStates to found an institution in Washington,DC for the purpose of “increasing and diffusingknowledge among men” (UNESCO, 1968, p. 12).Using the Smithson gift as seed funding, Congresschartered the Smithsonian Institution in 1846, andJoseph Henry became its first Executive Officer.Henry, a renowned experimental physicist, contin-ued the precedence of a federal agency directlysupporting research through grants to individ-ual investigators to pursue fundamental research.Also, the Institution represented a base for exter-nal support of scientific and engineering researchsince, during the 1850s, about 100 academic insti-tutions were established with science and engi-neering emphases.

Thus, the pendulum had made one completeswing in the hundred years since the signing of theConstitution. In the early years, the governmentviewed itself as having no more than an indirectinfluence on the development of science and tech-nology, but over time its role changed from indi-rect to direct. This change was justified in largepart because advances in science and technologyare viewed as promoting the public interest.

Toward a national infrastructure

Scientists had long looked toward the Europeanuniversities for training in the sciences, but now an


academic infrastructure was beginning to developin the United States. Harvard University awardedits first bachelor of science degree in 1850. Thedevelopment of an academic science base and thebirth of technology-based industries (e.g., the elec-trical industry) in the late 1850s established whatwould become the foundation for America’s tech-nological preeminence.

In 1863, during the Civil War, Congress estab-lished the National Academy of Sciences. The fed-eral government funded the Academy but not themembers affiliated with it who had “an obligationto investigate, examine, experiment, and reportupon any subject of science or art in response to arequest from any department of the Government”(UNESCO, 1968, p. 14). Then, as today, theAcademy is independent of governmental control.

The Morrill Act of 1862 established the landgrant college system thereby formally recognizingthe importance of trained individuals in the agri-cultural sciences. The Act charged each state toestablish at least one college in the agriculturaland mechanical sciences. Each state was given30,000 acres of federal land per each elected U.S.Senator and Representative. An important out-growth of this land grant system was a mechanismor infrastructure through which state and federalgovernments could financially support academicresearch interests.

Although the federal government was encour-aging an infrastructure to support science andtechnical research, it did not have a so-called in-house staff of permanent professionals who werecompetent to identify either areas of nationalimportance or areas of importance to specificagencies. In 1884, Congress established the Alli-son Commission to consider this specific issue.While many solutions were debated, including theestablishment of a Department of Science—anidea that resurfaces every few decades—the Com-mission soon disbanded without making any rec-ommendations much less reaching closure on thematter. One could conclude from the inaction ofthe Commission that it favored the decentralizedadministrative architecture that had evolved overtime as opposed to a centralized one.

Toward an industrial infrastructure

Most scientists in the United States in the 1870sand 1880s had been trained in Europe, Germany

in particular. What they experienced firsthandwere the strong ties between European industriesand graduate institutions. European companiesinvested in professors and in their graduate stu-dents by providing them with funds and accessto expensive materials and instruments. In return,the firms gained lead-time toward new discover-ies, as well as early access to the brightest grad-uate students as soon as they completed theirstudies.10 This form of symbiotic arrangementbecame the norm for the European-trained scien-tists who were working in U.S. industries and U.S.universities.

By the turn of the century, it was widelyaccepted among industrial leaders that scientificknowledge was the basis for engineering develop-ment and was the key to remaining competitive.Accordingly, industrial research laboratories soonbegan to blossom as companies realized their needto foster scientific knowledge outside of the uni-versity setting.11 There are a number of examplesof this strategy.

General Electric (GE) established the GeneralElectric Research Laboratory in 1900 in responseto competitive fears that improved gas lightingwould adversely affect the electric light business,and that other electric companies would threatenGE’s market share as soon as the Edison patentsexpired. Similarly, AT&T was facing increasingcompetition from radio technology at the sametime. In response, AT&T established Bell Labo-ratories to research new technology in the eventthat wire communications were ever challenged.And as a final example, Kodak realized at theturn of the century that it must diversify from syn-thetic dyes. For a number of years Kodak relied onGerman chemical technology, but when that tech-nology began to spill over into other areas suchas photographic chemicals and film, Kodak real-ized that their competitive long-term health restedon their staying ahead of their rivals. Kodak tooformed an in-house research laboratory.

Many smaller firms also realized the compet-itive threats that they could potentially face as aresult of technological competition, but because oftheir size they could not afford an in-house facility.So as a market response, contract research labo-ratories began to form. Arthur D. Little was onesuch contract research laboratory that specializedin the area of chemicals.


Just as industrial laboratories were growingand being perceived by those in both the pub-lic and private sectors as vitally important to theeconomic health of the nation, private founda-tions also began to grow and to support universityresearchers. For example, the Carnegie Institutionof Washington was established in 1902, the Rus-sell Sage Foundation in 1907, and the RockefellerFoundation in 1913.

In the early-1900s science and technologybegan to be embraced—both in concept and inpractice—by the private sector as the founda-tion for long-term competitive survival and gen-eral economic growth.

World War I and the years that followed

Increased pressure on the pace of scientificand technical advancements came at the begin-ning of World War I. The United States hadbeen cut off from its European research base.Congress, in response, established the Council ofNational Defense in 1916 to identify domesticpockets of scientific and technical excellence. TheNational Academy of Sciences recommended toPresident Woodrow Wilson the formation of theNational Research Council to coordinate cooper-ation between the government, industry, and theacademic communities toward common nationalgoals.12 The prosperity of the post-World War Idecade also created an atmosphere supportive ofthe continued support of science and technology.In 1920, there were about 300 industrial researchlaboratories, and by 1930 there were more than1,600.13 Of the estimated 46,000 practicing scien-tists in 1930, about half were at universities andover a third were in industry. Herbert Hoover wasSecretary of Commerce at this time. He adoptedthe philosophy that (UNESCO, 1968, p. 18):

� � � pure and applied scientific research constitute a foun-dation and instrument for the creation of growth and effi-ciency of the economy.

In response to the Great Depression and the sub-sequent national economic crisis, two importantevents occurred in 1933. One was the appoint-ment of a Science Advisory Board and the otherwas the establishment of a National PlanningBoard. Whereas the National Research Councilhad been organized around fields of science to

address governmental needs, the Science Advi-sory Board was multi-field and organized aroundimpending national problems. The National Plan-ning Board was formed on the presumptionthat there were areas of economic concern thatrequired a national perspective rather than afield-of-science perspective. In 1934, the NationalResources Committee replaced the National Plan-ning Board, and it then subsumed the ScienceAdvisory Board. The bottom line was, after all ofthe organizational issues were settled, that the fed-eral government recognized through the forma-tion of these committees and boards that it hadand would continue to have an important coordi-nating role to play in the science and technologyplanning toward a national goal of economic wellbeing. Hence, the pendulum began to swing awayfrom government having a hands-on role toward ithaving an indirect influence on planning the envi-ronment for science and technology.

In 1938, the Science Committee of the NationalResources Committee issued a multi-volumereport entitled, Research—A National Resource.Some important first principles were articulatedin that report. These principles have since thenformed a basis for economists and policy makersto rationalize/justify the role of government in sci-ence and technology. The report is explicit that:

• There are certain fields of science and tech-nology, which the government has a Consti-tutional responsibility to support. These fieldsinclude defense, determination of standards,and certain regulatory functions.

• The government is better equipped to carryon research in certain fields of science thanthe private sector. These are areas where“research is unusually costly in proportion toits monetary return but is of high practicalor social value” (p. 25). Examples cited inthe report include aeronautical and geologicalresearch.

• Research by the government “serves to stimu-late and to catalyze scientific activity by non-governmental agencies. In many fields, newlines of research are expensive and returnsmay be small or long delayed. Industry can-not afford to enter such fields unless there isreasonable prospect of definite financial gainwithin a predictable future, and it is under such


circumstances that the Government may leadthe way � � � ” (p. 26). One example cited wasthe Navy Department’s influence on the devel-opment of the steel industry.

World War II and the years that followed

The involvement of the United States in WorldWar II had a dramatic impact on the scope anddirection of government’s support of science andtechnology. Prior to the war, there were about92,000 scientists, with about 20 percent in govern-ment and the remaining 80 percent being almostequally divided between universities and the morethan 2,200 industrial laboratories. Clearly, theUnited States had a significant scientific resourcebase to draw upon for it war efforts.

In 1940, President Roosevelt established theNational Defense Research Committee and askedVannevar Bush, President of Carnegie Institutionof Washington, to be its chairman. The purposeof this committee was to organize scientific andtechnological resources toward enhancing nationaldefense. It soon became apparent that this taskrequired an alternative administrative structure.In 1941, Roosevelt issued an Executive Orderestablishing the Office of Scientific Research andDevelopment (OSRD) with Bush as Director. TheOSRD did not conduct research; rather it real-ized that there were pockets of scientific andtechnological excellence throughout the country,and through contractual relationships with uni-versities and industry and government agenciesit could harness national strengths with a focuson ending the war. One hallmark event from theefforts of the OSRD was the establishment ofthe Los Alamos Laboratory in New Mexico underthe management of the University of California.What came about from the collective efforts ofthe resources acquired by the Office were not onlyatomic weapons but also radar.

By 1944, it was clear that World War II wasalmost over. President Roosevelt then asked Bushto develop recommendations as to how scien-tific advancements could contribute in the largersense to the advancement of national welfare. Inhis November 17, 1944 letter to Bush, PresidentRoosevelt stated:

The Office of Scientific Research and Development, ofwhich you are the Director, represents a unique experi-ment of team-work and cooperation in coordinating sci-entific research and in applying existing scientific knowl-edge to the solution of the technical problems paramountin war. � � � There is � � � no reason why the lessons to befound in this experiment cannot be profitably employed intimes of peace. This information, the techniques, and theresearch experience developed by the Office of ScientificResearch and Development and by the thousands of scien-tists in the universities and in private industry, should beused in the days of peace ahead for the improvement ofthe national health, the creation of new enterprises bring-ing new jobs, and the betterment of the national standardof living. � � � New frontiers of the mind are before us, andif they are pioneered with the same vision, boldness, anddrive with which we have waged this war we can create afuller and more fruitful employment and a fuller and morefruitful life.

Shortly before asking Bush to prepare thisreport, Senator Kilgore from West Virginia hadintroduced a bill to create a National ScienceFoundation. The Kilgore bill recommended givingauthority to federal laboratories to allocate publicmoneys in support of science to other governmentagencies and to universities. Clearly, this recom-mendation gave a direct role to government inshaping the technological course of the countrynot only in terms of scientific direction but alsoin terms of what groups would conduct the under-lying research. The bill was postponed until afterthe war.

Bush submitted his report, Science—The End-less Frontier, to President Roosevelt on July 25,1945. In Bush’s transmittal letter to the presidenthe stated:

The pioneer spirit is still vigorous within this Nation. Sci-ence offers a largely unexplored hinterland for the pioneerwho has the tools for his task. The reward of such explo-ration both for the Nation and the individual are great.Scientific progress is one essential key to our security asa nation, to our better health, to more jobs, to a higherstandard of living, and to our cultural progress.

The foundations set forth in Science—The End-less Frontier are:

• “Progress � � � depends upon a flow of new sci-entific knowledge” (p. 5).

• “Basic research leads to new knowledge.14 Itprovides scientific capital. � � � New productsand new processes do not appear full-grown.They are founded on new principles and newconceptions, which in turn are painstakingly


developed by research in the purest realms ofscience” (p. 11).

• “The responsibility for the creation of new sci-entific knowledge � � � rests on that small bodyof men and women who understand the fun-damental laws of nature and are skilled in thetechniques of scientific research” (p. 7).

• “A nation which depends upon others for itsnew basic scientific knowledge will be slow inits industrial progress and weak in its compet-itive position in world trade, regardless of itsmechanical skill” (p. 15).

• “The Government should accept new responsi-bilities for promoting the flow of new scientificknowledge and the development of scientifictalent in our youth” (p. 7).

• “If the colleges, universities, and researchinstitutes are to meet the rapidly increas-ing demands of industry and Government fornew scientific knowledge, their basic researchshould be strengthened by use of public funds”(p. 16).

• “Therefore I recommend that a new agency forthese purposes be established” (p. 8).

Bush recommended in his report the creationof a National Research Foundation. Its proposedpurposes were to:

� � � develop and promote a national policy for scien-tific research and scientific education, � � � support basicresearch in nonprofit organizations, � � � develop scientifictalent in American youth by means of scholarships andfellowships, and � � � contract and otherwise support long-range research on military matters.

Bush envisioned a National Research Foun-dation that would provide funds to institutionsoutside government for the conduct of research.Thus, this organization differed from Kilgore’sproposed National Science Foundation in thatBush advocated an indirect role for government.There was agreement throughout government thatan institutional framework for science was needed,but the nature and emphases of that frameworkwould be debated for yet another five years.15

Science—The Endless Frontier affected the sci-entific and technological enterprise of this nationin at least two ways. It laid the basis for whatwas to become the National Science Foundationin 1950. Also, it set forth a paradigm that wouldover time influence the way that policy makers and

academic researchers thought about the processof creating new technology. The so-called linearmodel set forth by Bush is often represented by:

Basic Research→Applied Research→Development→Enhanced Production→Economic Growth

Complementing Science—The Endless Frontierwas a second, and often overlooked, report pre-pared in 1947 by John Steelman, then Chair-man of the President’s Scientific Research Board.As directed by an Executive Order from Pres-ident Truman, Steelman, in Science and PublicPolicy, made recommendations on what the fed-eral government could do to meet the challengeof science and assure the maximum benefits tothe Nation. Steelman recommended that nationalR&D expenditures should increase as rapidly aspossible, citing (p. 13):

1. Need for Basic Research. Much of the world isin chaos. We can no longer rely as we once didupon the basic discoveries of Europe. At thesame time, our stockpile of unexploited funda-mental knowledge is virtually exhausted in cru-cial areas.

2. Prosperity. This Nation is committed to a pol-icy of maintaining full employment and full pro-duction. Most of our frontiers have disappearedand our economy can expand only with moreintensive development of our present resources.Such expansion is unattainable without a stim-ulated and growing research and developmentprogram.

3. International Progress. The economic healthof the world—and the political health of theworld—are both intimately associated with ourown economic health. By strengthening oureconomy through research and development weincrease the chances for international economicwell-being.

4. Increasing Cost of Discovery. The frontiers ofscientific knowledge have been swept so farback that the mere continuation of pre-wargrowth, even in stable dollars, could not possi-bly permit adequate exploration. This requiresmore time, more men, more equipment thanever before in industry.


5. National Security. The unsettled internationalsituation requires that our military researchand development expenditures be maintainedat a high level for the immediate future. Suchexpenditures may be expected to decrease intime, but they will have to remain large for sev-eral years, at least.

An important element of the Steelman reportwas the recommended creation of a National Sci-ence Foundation, similar in focus to the NationalResearch Foundation outlined by Bush. And,Congress passed the National Science FoundationAct in 1950.

Renewed post-war attention toward science andtechnology came with the success of the SovietUnion’s space program and the orbit of its SputnikI in October 1957. In response, President Eisen-hower championed a number of committees andagencies to ensure that the United States couldsoon be at the forefront of this new frontier. Note-worthy was the National Defense Education Actof 1958, which authorized $1 billion in federalmoneys for support of science, mathematics, andtechnology graduate education. This proposal isprecisely the type of support that Bush recom-mended in his report.

As the post-World War II period came to aclose, there was a well-established national andindustrial infrastructure to support the advance-ment of science and technology. But, moreimportant than the infrastructure, there was animbedded belief that scientific and technologi-cal advancements are fundamental for economicgrowth, and that the government has an impor-tant supporting role—both direct and indirect—toensure such growth.

Every president since Eisenhower has initiatedmajor science policy initiatives.16 Kennedy set thegoal of sending a man to the moon by the endof the 1960s and funded the needed programsto make this a reality. Johnson emphasized theuse of scientific knowledge to solve social prob-lems through, for example, his War on Poverty.Nixon dramatically increased federal funding forbiomedical research as part of his War on Cancer.Ford created the Office of Science and Technol-ogy Policy (OSTP) within the Executive Branch.Carter initiated research programs for renewableenergy sources such as solar energy and fission.

During the Reagan administration, expenditureson defense R&D increased dramatically as part ofhis Star Wars system. President Bush (no relation-ship to Vannevar Bush) set forth this nation’s firsttechnology policy (see below) and increased thescope of the National Institute of Standards andTechnology (NIST, see below). President Clintonestablished important links between science andtechnology policy, championing programs to trans-fer public technology to the private sector.

4. Economic growth and technological change

In the previous section, we described VannevarBush’s paradigm of research and developmentleading to economic growth. As a practical mat-ter, economic growth is generally defined at themacroeconomic level in terms of Gross Domes-tic Product (GDP).17 Figure 2 shows GDP-relatedgrowth for a number of countries. GDP per capitais greater in the United States than in any of theother industrial nation shown in the figure.

An inspection of Figure 2 raises a number ofquestions, two of which are: Why do economiesgrow? and, Why has the U.S. economy outper-formed that of other industrial nations even aftercontrolling for national size?

Theories of economic growth

The early literature on economic growth is formu-lated analytically using what economists refer toas a production function. Simply put, a produc-tion function represents the relationship betweenthe output of an economic unit (a firm, industry,or economy) and the factors of production—orinputs or resources—used to produce that output.

0

5

10

15

20

25

30

35

1960 1965 1970 1975 1980 1985 1990 1995 2000

Year

Rea

l GD

P

U.S. Japan France Germany U.K.

Figure 2. Real GDP per capita for selected countries1960–1999 (U.S. $1000). Source: Science and EngineeringIndicators—2000, Appendix Table 7-3.


If output, Q, can be defined in terms of thetwo most basic factors of production, the stock ofcapital (plant and equipment), K, and the stock oflabor, L, then a basic production function can bewritten as:

Q = f �K�L� (1)

Equation (1) denotes that a firm’s, industry’s,or economy’s output will change in response tochanges in either the quantity or quality of capi-tal or the quantity or quality of labor. However,to account for other influences on output suchas new technology, the production relationshipin Equation (1) can be modified most simply toinclude a catch-all variable, A, as:

Q = AF�K�L� (2)

where, to be more specific, A is a shift factor toaccount for exogenous technological factors (asopposed to conventional factor inputs such as Kand L) that affect production, and where, becauseof the inclusion of A in Equation (2), F�K�L� isdistinct from f �K�L� in Equation (1).

By dividing both sides of Equation (2) by thecombination of K and L inputs (i.e., by total fac-tors) denoted by F�K�L�, variable A can be inter-preted as an index of output per unit of all inputsor of total factor productivity.

A=Q/F�K�L� (3)

Early on, Robert Solow (1957), who was sub-sequently awarded the Nobel Prize in economics,estimated a variation of Equation (3) using aggre-gate U.S. data. He calculated changes in thevalue of A between 1909 and 1949. His analysisshowed that more than 87 percent of the growthin the U.S. economy could not be explained bythe growth in capital and labor, and hence theresidual or unexpected portion of growth must

50

60

70

80

90

100

110

1948 1953 1958 1963 1968 1973 1978 1983 1988 1993 1998

Year

TF

P

Figure 3. Private nonfarm TFP, 1948–1997. Source: U.S.Bureau of Labor Statistics.

be attributable to something else.18 Solow specu-lated that what was captured in his residual calcu-lation may reflect technology advance over time.Changes in A from Equation (3) measure what iscalled total factor productivity growth, or techno-logical advancement.

Other researchers, using alternative frame-works, reached similar conclusions. Abramovitz(1956), for example, referred to the unexplainedportion of growth more cautiously as a measureof our “ignorance.” This implied that whileeconomists were able to calculate unexplainedgrowth, they were unable to provide a con-clusive explanation for what caused improve-ments in economic performance. Unlike Solow,Abramovitz speculated in some detail that growthnot attributable to capital and labor was likelydue to improvements in education and increasesin research and development activity (R&D).

The academic literature is replete with theo-ries to explain growth over time. The so-called“old growth” theory literature (Nelson and Phelps,1966) is based on more sophisticated versions ofEquation (1). That is, this literature emphasizesadditional inputs aside from K and L such asinvestments in R&D and education. As well, itemphasizes the greater specificity by which inputsare measured including consideration for the het-erogeneity of K and L (e.g., new vintages of Kembody others’ technological investments).

The so-called “new growth” theory (Romer,1986, 1994) emphasizes the influence of other fac-tors on growth that are not directly specified inan expanded version of Equation (1). These fac-tors include, for example, technologies or efficien-cies that spillover into a firm’s production functioneither from other firms or from general advancesin the economy (such as information technology)or that spillover into a nation’s production func-tion from trade policies. New growth theory is alsobased on careful, explicit analytical modeling ofthe incentives of agents to invest in new technol-ogy. Figure 4 expands upon Figure 1 to incorpo-rate these ideas. In particular, two new elementsare included in Figure 4 that were not in Figure 1.First, in-house or private investment in R&D isshown to have an influence on how the firmresponds to the interaction of its strategic direc-tion and its competitive environment. Second, wehave capsulated the essence of new growth theory


Strategic direction of the firm

Entrepreneurial response Innovation Value Added Competitive market conditions

In-house R&D External influences on economic performance

Figure 4. The entrepreneurial process: a second look.

by simply acknowledging that external influencesaffect firm performance directly as well as indi-rectly through innovation.

Regardless of whether one adheres to the morenarrow old theories or the broader new theo-ries, the evidence is overwhelming that technologydrives economic growth. There was renewed inter-est in promoting economic growth in the postwaraftermath of the destruction of the industrial baseof many nations. Thus, it is not surprising thatgreater attention was devoted to the analysis ofR&D, which was hypothesized early on to be animportant determinant of economic growth.

5. Dimensions of R&D

For purposes of measurement, there are three fun-damental dimensions of R&D. The first relatesto the source of funding of R&D (who financesR&D), the second to the performance of R&D(who actually does the work), and the third tothe character of use of R&D (whether the workis basic research, applied research, or develop-ment). These three fundamental dimensions arenot mutually exclusive.

Table IINational R&D expenditures, by performer and funding source: 1999 ($millions)

Source of R&D funds

Other PercentFederal Universities Non-federal nonprofit distribution,

Performer government Industry and colleges government institutions Total by performer

Total R&D 65�853 169�312 5�838 2�085 3�912 247�000 1000%Federal government 17�362 — — — — 17�362 70%Industry 19�937 165�955 — — — 185�892 753%Industry FFRDCs 2�166 — — — — 2�166 09%Universities and colleges 16�137 2�163 5�838 2�085 2�032 28�255 114%University FFRDCs 6�169 — — — — 6�169 25%Other nonprofit institutions 3�246 1�194 — — 1�880 6�320 26%Nonprofit FFRDCs 836 — — — — 836 03%Percent distribution, by source 26.7% 68.5% 2.4% 0.8% 1.6% 100%

Source: National Science Foundation.

There are also other important aspects ofR&D. One such dimension is related to the sizeof firms that conduct R&D, a second to the geo-graphical distribution of R&D, and a third to therelationship between R&D and total factor pro-ductivity growth. We now consider each of thesedimensions in turn.

Sources of funding of R&D

The top row of Table II shows the sources for the$247.0 billion of R&D expenditures in the UnitedStates in 1999. Industry accounted for nearly 69percent of those expenditures; the federal govern-ment another 27 percent; and all other sources,including state and local governments, universi-ties and colleges, and other nonprofit institutions,about 5 percent.

The primacy of industry in funding R&D hasnot always held, as shown in Figure 5. In theaftermath of World War II up through the early1980s, the federal government was the leadingprovider of R&D funds in the Nation. Althougha federal R&D presence existed before then, dur-ing the war the federal government dramaticallyexpanded its R&D effort by establishing a networkof federal laboratories, including atomic weaponslaboratories. It was at that time that the federalgovernment also greatly increased its support toextramural R&D performers, especially to a selectgroup of universities and large industrial firms.After the war (along with the widespread influence


0

50

100

150

200

250

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

Year

$R&

D

Total Federal Industry

Figure 5. U.S. R&D funding by source, 1953–1998 (bil-lions $1992). Source: Science and Engineering Indicators—2000,Appendix Table 2-6.

of the Bush and Steelman reports), federal R&Dsupport continued to expand for both defense andnon-defense purposes, including health R&D inthe National Institutes of Health and—after theestablishment of the National Science Founda-tion in 1950—a broad portfolio of fundamentalresearch activities. As a result of a post-Sputniknational commitment to catch up to the Sovietspace successes, federal support for space-relatedR&D mushroomed in the late 1950s and early1960s. By 1960, the federal government accountedfor 65 percent of the nation’s total investment(80 percent of which was for defense), and indus-try accounted for 32 percent of the total.

Over the next twenty years the federal govern-ment continued to be the leading source of R&Dfunding, although the direction shifted over time.In the early 1960s, the relative defense share offederal R&D funding dropped precipitously from80 percent in 1960 to about 50 percent in 1965,where it fluctuated narrowly until 1980. Early on,R&D for space exploration was the primary non-defense recipient of federal R&D funding. Indeedmore than three-fourths of federal non-defenseR&D funds were in support of NASA’s missionactivities by 1965. By 1970, however, after the suc-cess of several lunar landings, support for othernon-defense purposes began to claim an increas-ingly larger share of the federal R&D totals, andcontinued to do so throughout the 1970s; notablygrowth in federal energy R&D occurred as aresponse to the several oil embargoes. Also by1970, R&D support from industry was on the rise,and it accounted for just over 40 percent of thetotal national R&D effort. As a result of relativelyflat federal funding in the 1970s and continual

slow growth from the industrial sector, the fed-eral government and industry accounted for aboutequal shares by the early 1980s.

Since then the federal government’s share ofR&D decreased to about 40 percent of totalin 1990 to its current share, somewhat below30 percent. Initially, the decreasing federal sharecame about even though federal dollar supportfor R&D—in absolute terms—was increasing.Between 1980 and 1987, federal R&D rose about40 percent after adjusting for inflation. Most ofthis growth, however, was in support of defenseactivities so that by 1987, the defense R&D sharehad grown to two-thirds of the federal R&D total(its highest share since 1963). After the break-upof the Soviet Union, the imperative for continualgrowth in federal defense R&D support was notas strong and the federal R&D total once againslowed (and even fell in constant dollars).

In terms of which agencies provide the R&Dfunds, federal sources are highly concentratedamong just a few agencies. According to the latestdata provided by the agencies themselves, of the$74 billion obligated for R&D and R&D plant infiscal year 1998, just five accounted for 94 per-cent of all funds: Department of Defense (48%),Department of Health and Human Services, pri-marily the National Institutes of Health (19%),National Aeronautics and Space Administration(13%), Department of Energy (9%), and NationalScience Foundation (3%).

Concurrent with recent reductions in federalR&D spending, major changes have also occurredin industrial R&D spending patterns. After lack-luster funding in the early 1990s (reflecting theimpact of mild recessions on its R&D activities)industry R&D support has grown rapidly since1994 and now accounts for almost 70 percent ofthe national R&D total. As a result, and comparedwith the funding patterns of the mid-1960s, indus-try and government have reversed positions.

R&D performers

R&D is performed in what has been termed theU.S. national innovation system. The system is,according to Crow and Bozeman (1998, p. 42):

� � � the complex network of agents, policies, and institu-tions supporting the process of technical advance in aneconomy.


The performers of R&D within the system areresearch laboratories. The laboratory performersof R&D correspond to the sectors that financeR&D, but not all R&D funded by a sector isperformed in that sector. For example, industryperformed approximately $186 billion of R&D in1999, of which $166 billion came from industryitself. The additional amount of R&D performedby industry came from the federal government.

Almost one-third of the R&D funded by thefederal government is performed in industry, andmore than one-half of those dollars are spent inthe aircraft, missiles, and transportation equip-ment industries. Universities and colleges fundonly about 20 percent of the R&D they perform.Fifty-seven percent of the R&D performed inuniversities and colleges comes from the federalgovernment and the rest equally from industry,nonprofit institutions, and nonfederal governmentsources.

As shown in Figure 5, since the late-1980sthe federal government has decreased its fund-ing of national R&D. The lion’s share of thatdecrease has come in the form of federal alloca-tions for R&D performed in industry, for whichthe R&D level of support displays a somewhatroller-coaster-like pattern. The latest peak in fed-eral support for industrial R&D was a result ofmajor defense-related funding increases for Presi-dent Reagan’s Strategic Defense Initiative prior tothe collapse of the Soviet Union. By contrast, fed-eral funding to universities and colleges, adjustedfor inflation, has increased slightly each year sinceat least the late 1970s.

There are other important dimensions to theperformance of industrial R&D. About three-fourths of industrial R&D is performed in manu-facturing industries. The dominant manufacturingindustries in terms of dollars of R&D performedare chemicals and allied products, electrical equip-ment (including computers), and transportationequipment. The remaining one-fourth is per-formed in the non-manufacturing sector, includingservices. Computer-related services are the lead-ers therein. The steep growth in R&D performedin the services is a relatively recent phenomenon.As recently as 15 years ago, manufacturers stillaccounted for more than 90 percent of the indus-trial R&D total.

Also, not all industry-performed R&D occurswithin the geographical boundaries of the United

States. Of the nearly $186 billion in R&D per-formed by industry in 1998 (the latest year forwhich the foreign-performed data are available),about $16 million, or about 9 percent, was con-ducted in other countries. Foreign investmentsin R&D are not unique to U.S. firms; the out-flow of U.S. industrial R&D into other countriesis approximately offset by an inflow of others’R&D to be performed in the United States.Most (68 percent) of U.S.-funded R&D abroadwas performed in Europe—primarily in Germany,the United Kingdom, and France. The currentEuropean share of U.S. industry’s offshore R&Dactivity, however, is somewhat less than the 75 per-cent share reported for 1982 (peak year). Overall,U.S. R&D investments abroad have generallyshifted away from the larger European coun-tries and Canada, and toward Japan, several ofthe smaller European countries (notably Swedenand the Netherlands), Australia, and Brazil. Phar-maceutical companies accounted for the largestindustry share (18 percent of U.S. 1997 over-seas R&D), which was equivalent to 21 per-cent of their domestically-financed R&D. Muchof this pharmaceutical R&D took place in theUnited Kingdom.

Foreign firms in the United States make sub-stantial R&D investments. From 1987 to 1996,inflation-adjusted R&D growth from majority-owned affiliates of foreign firms averaged 10.9 per-cent per year, and are now roughly equivalent toU.S. companies’ R&D investment abroad. Affili-ates of firms domiciled in Germany, Switzerland,the United Kingdom, France, and Japan collec-tively account for 72 percent of this foreign fund-ing. Foreign-funded R&D in the United Statesin 1996 was concentrated in drugs and medicines(mostly from Swiss, German, and British firms),industrial chemicals (funded predominantly byGerman and Dutch firms), and electrical equip-ment (one-third of which came from Frenchaffiliates).

R&D by character of use

Vannevar Bush is credited for first using theterm “basic research,” which he defined to meanresearch performed without thought of practicalends in his 1945 report to President Roosevelt,Science—The Endless Frontier. Since that time,


policy makers have been concerned about defini-tions that appropriately characterize the variousaspects of scientific inquiry that broadly fall underthe label of R&D and that relate to the linearmodel that Bush proffered.

Definitions are important to the National Sci-ence Foundation because it collects expendituredata on R&D. For those data to accurately reflectindustrial and academic investments in technolog-ical advancement, and for those data to be com-parable over time, there must be a consistent setof reporting definitions.

The classification scheme used by the NationalScience Foundation for reporting purposes wasdeveloped for its first industrial survey in 1953–1954.19 While minor definitional changes weremade in the early years, namely to modify thecategory originally referred to as “basic or funda-mental research” to simply “basic research,” theconcepts of basic research, applied research, anddevelopment have remained much as was implic-itly contained in Bush’s 1945 linear model.

The objective of basic research is to gain morecomprehensive knowledge or understanding of thesubject under study, without specific applicationsin mind. Basic research is defined as researchthat advances scientific knowledge but does nothave specific immediate commercial objectives,although it may be in fields of present or poten-tial commercial interest. Much of the scientificresearch that takes place at universities is basicresearch. Applied research is aimed at gainingthe knowledge or understanding to meet a spe-cific recognized need. Applied research includesinvestigations oriented to discovering new sci-entific knowledge that has specific commercialobjectives with respect to products, processes,or services. Development is the systematic useof the knowledge or understanding gained fromresearch directed toward the production of use-ful materials, devices, systems, or methods, includ-ing the design and development of prototypes andprocesses.20

Approximately 61 percent of national R&D isdevelopment, with 23 percent of R&D being allo-cated to applied research and 16 percent beingallocated to basic research. Different sectors con-tribute disproportionately to the Nation’s funding

and performance of these R&D component cate-gories. Applied research and development activ-ities are primarily funded by industry and per-formed by industry. Basic research, however, isprimarily funded by the federal government andgenerally performed in universities and colleges.The decline in federal support of R&D over thepast decade has primarily come at the expense ofapplied research and development performed inindustry.

R&D activity in large and small firms

Table III shows the level of R&D expendi-tures from all sources for several of the largerR&D-performing firms in the United Statesin 1997. These data come from a variety of pub-lic sources: several patterns can be seen from thedata in the table:

• Microsoft (the tenth-largest R&D-active com-pany) invests about one-fifth of the amountof R&D invested by General Motors (thelargest R&D-active company). Thus, evenamong the R&D giants, R&D expendituresvary dramatically.

Table IIILargest R&D-active U.S. companies

Rank $R&D $R&D/in 1997 Company (millions) $sales (%)

1 General Motors 8�2000 492 Ford Motor Company 6�3270 413 IBM 4�3070 524 Lucent Technologies 3�1006 1185 Hewlett-Packard 3�0780 726 Motorola 2�7480 867 Intel 2�3470 948 Johnson & Johnson 2�1400 959 Pfizer 1�9280 15410 Microsoft 1�9250 169

95 Imation 1949 8996 Dana 1930 2297 Thermo Electron 1916 5498 Eastman Chemical 1910 4199 Cabletron Systems 1818 132100 Whirlpool 1810 21

Source: Science & Engineering Indicators—2000, AppendixTable 2-58.


• Microsoft is, however, more than three timesas R&D intensive as General Motors, meaningthat its invests nearly three times the amountas General Motors relative to its sales.

• In general (with notable exceptions such asCabletron Systems), larger R&D performersalso spend more on R&D relative to their sizethan do the lower-ranked firms in the list ofthe top 100.

• The level of R&D expenditures is not unre-lated to the industry of the R&D performer.For example, first-ranked General Motors andFord Motor Company are in the transporta-tion industry. Second-ranked IBM, LucentTechnologies, Hewlett-Packard, Motorola, andIntel are in information systems.

Company-specific data on R&D expendituresfor small-sized firms are not readily available.However, aggregate NSF data show that $98 bil-lion of the $169 billion that industry spent onR&D in 1998 was performed in firms with 10,000or more employees. Thirty billion dollars (18% ofthe industry total) was performed in firms withfewer than 500 employees. Similarly, 70 percent ofthe funds expended were part of company R&Dbudgets that exceeded $100 million.

Stylized facts aside, there is a more subtleand perhaps more important R&D-related issue.Since the early 1980s, policy makers have beenconcerned that critical American industries werelosing their competitive dominance of world mar-kets. During the 1990s, these same industriesseem to have reemerged as major internationalcompetitors. While some of this resurgence is aresponse to purposive policies, a portion of itcan also be attributable to small firms, many ofwhich were not in existence in the early 1980sto be affected by policy and many of which donot even conduct R&D. Still, during the 1990s,small firms were a driving engine of growth,job creation, and renewed global competitivenessthrough innovation.

There is a rich literature related to the per-formance of R&D in small firms as comparedto large firms. Some of the conclusions from thisresearch are:

• Large firms have a greater propensity to patentthan do small firms.

• Small firms are just as innovative as large firms,in general. But, in some industries, large firmshave the innovative advantage (pharmaceuti-cals, aircraft), while in other industries smallfirms have the innovative advantage (software,biotechnology).

• Small-firm and large-firm innovative activitiesare complementary.

Table IV provides a selected summary of the find-ings from the literature on innovation and firmsize.

The economic importance of small firms,including the innovative differences between smallfirms and large firms, requires an explanationsince the share of overall economic activityattributable to small firms is small and it did notincrease during the 1990s. The explanation rel-evant to the focus of this primer begins with amodel of the knowledge production function.21

Table IVSelected studies of the relationship between innovation and

firm size

Innovationmeasure Findings Authors

R&D R&D spending in Mueller (1967)positively related Grabowski (1968)to firm size Mansfield (1968)

Patents Patenting is Schererpositively or (1965, 1983)proportionally Pakes andrelated to Griliches (1980)firm size Hall et al. (1986)

Schwalback andZimmermann(1991)

New product Parity across firm Acs and Audretschinnovations size, although (1990)

there are Audretsch (1995)differencesaccording toindustry

Adoption of Positive relationship Romeo (1975)advanced between firm size Dunne (1994)manufacturing and the probability Siegel (1999)technologies of adopting an

advancedmanufacturingtechnology


The simplified production function in Equa-tion (1) above can be expanded conceptually andanalytically to include the stock of knowledge as adiscrete input along with K and L. One investmentin knowledge that many firms make is in R&D.However, there are other key factors that generateknowledge for the firm besides R&D, and in factmany small firms do not even conduct R&D yetthey are very innovative. Some such firms rely onknowledge that spills over from external sourcesincluding universities, and small firms are rela-tively more adept at absorbing knowledge fromexternal sources than large firms. Table V pro-vides a brief summary of this spillover literaturewith respect to small firms.

Included in Table V under the source categoryof individual spillovers are new employees. Whyare, for example, small firms able to exploit knowl-edge embodied in new employees to a greaterextent than large firms? New and small firms pro-vide the opportunity for creative individuals to

Table VSelected studies on knowledge spillovers

Spilloversource Findings Authors

Industry Spillovers vary across Jaffe (1989)spillovers industries; greater Saxenien (1990)

spillovers in Acs et al. (1992)knowledge-intensive Trajtenberg andindustries Henderson (1993)

Audretsch andFeldman (1996)

University University spillovers Link andspillovers more important to Rees (1990)

small firms than Audretsch andlarge firms Feldman (1996)

Firm Firm spillovers Acs et al. (1994)spillovers more important to Feldman (1994)

large firms than Eden et al. (1997)small firms

City spillovers Diversity generates Glaeser et. almore spillovers than (1992)specialization; Almeida andlocalized Kogut (1997)competition more Feldman andthan monopoly Audretsch (1999)

Individual Spillovers shaped Audretsch andspillovers by role and Stephan (1996)

mobility of Prevezer (1997)knowledge workers

implement new ideas that otherwise would berejected or would remain unexploited in an orga-nizationally rigid firm. New firms thus serve asagents of change. In a global economy where com-parative advantage is based in large part on inno-vation, small firms are a critical resource. Publicpolicies to enhance innovation in small firms arediscussed below.

R&D activity by geographic location

R&D activities in the United States are highlyconcentrated in a small number of states. In 1997,the 20 highest-ranking states in R&D accountedfor about 86 percent of the U.S. total; the lowest20 states accounted for only 4 percent. Califor-nia, at nearly $42 billion, had the highest level ofR&D expenditures; it alone accounted for approx-imately one-fifth of the $199 billion U.S. total.The six states with the highest levels of R&Dexpenditures—California, Michigan, New York,New Jersey, Massachusetts, and Texas (in decreas-ing order of magnitude)—accounted for nearlytwo-thirds of the national effort. Among thesetop ten states, California’s R&D effort exceeded,by nearly a factor of three, the next-higheststate, Michigan, with $14 billion in R&D expen-ditures. After Michigan, R&D levels declined rel-atively smoothly to approximately $7 billion forMaryland.

States that are national leaders in total R&Dperformance are usually ranked among the lead-ing sites in industrial and academic R&D per-formance. For industrial R&D, nine of the topten states were among the top ten for totalR&D, with Ohio of the top industrial R&D statesreplacing Maryland. For academic R&D, NorthCarolina and Georgia replaced New Jersey andWashington. There was less commonality with thetop ten for total R&D among those states thatperformed the most federal intramural research.Only four states were found in both top-ten lists:Maryland, California, Texas, and New Jersey.

Competition for resources is a fundamentalexplanation for the skewed distribution of R&Dand science resources within the United States.States that lack such resources lag others in inno-vation, scholarship, graduate education, and over-all economic growth.22


The relationship between R&D andproductivity growth

As previously noted, Robert Solow’s seminalarticle in 1957 established that an extremely largepercentage of U.S. economic growth (over 87percent) could not be explained by growth inconventional inputs, i.e., capital and labor. Sincethen, researchers have searched for statisticalcorrelates of this unexplained growth, which iscommonly referred to as technological advance-ment or change. Hence, what these investigatorshave done is to posit that technological change,measured as total factor productivity growth, iscausally related to increased investments in R&D.

Using manufacturing sector, industry, and firm-level data, researchers have examined the strengthof the statistical relationship between R&D andtotal factor productivity growth. These analysesare based on models that correlate estimates ofthe growth of A from Equation (3) with measuresof R&D investment undertaken by firms, indus-tries, or aggregate sectors (depending on the unitof analysis). Mathematics aside, the extent of thecorrelation can be shown to be a measure of therate of return to R&D. This literature is consistentin terms of the following findings:

• the rate of return to privately-funded R&D isrelatively large, ranging on average between 30percent and 50 percent;

• the rate of return to privately-funded basicresearch is significantly greater than toprivately-funded development, the differencesbeing over 100 percent to basic research com-pared to about 15 percent to 20 percent forapplied research plus development; and

• the rate of return to federally-funded researchperformed in industry varies by characterof use; the returns to federally-funded basicresearch performed in industry is over 100 per-cent, while federally-funded development has anegligible return on productivity growth.

See summary Table VI.Related studies have attempted to evaluate the

social benefits, that is the spillover benefits to soci-ety, from industrial R&D. The rates of return toapplied research and development described justabove are for the most part private rates of return.More limited in number than the private rate ofreturn studies, the findings from the social rate

Table VISelected studies of the relationship between R&D and

productivity growth

Findings Authors

87.5% of the increase in Solow (1957)aggregate output between 1909and 1949 can be attributedto technical change

R&D has positive impact on total Terleckyj (1974)factor productivity growth as Scherer (1983)does R&D embodied in purchased Siegel (1997)intermediate and capital goods

Rate of return to privately-financed Mansfield (1980)basic research greater than for Link (1981)applied research or development Griliches (1986)

Lichtenberg andSiegel (1991)

Small direct impact of Link (1981)federally-financed R&D on total Griliches (1986)factor productivity growth

of return studies clearly indicate that the spilloverbenefits to society were somewhere in the 50 per-cent to 100 percent range.

Because of these findings, namely that the pri-vate and social rate of return to R&D is rela-tively high, policy makers have remained focusedon R&D investments in the private sector as a tar-get variable for stimulating economic growth. Theargument underlying such a focus is that, throughincentives, firms will continue to invest in addi-tional R&D projects and thus continue to stim-ulate economic growth and enhance standards ofliving through additional spillover effects.

Most of the academic studies associated withthis line of research were funded by the NationalScience Foundation during the late 1970s andearly 1980s, motivated in large part by a slowdownin industrial productivity growth that began in theearly 1970s23 and increased in the late 1970s andearly 1980s (see Figure 3) and by the fact that U.S.industries were losing their competitive advantagein global markets.24 It is not surprising then that inthe early 1980s, given the findings that the privateand social rates of return to R&D were very high,that there were several important policy initiativesdesigned specifically to stimulate industrial R&D.


6. Government’s role in innovation

The government should have an important roleto play in fostering innovation, especially private-sector innovation. The following reasoning hasbeen used to justify government intervention inthe innovation process:

• Innovation results in technological advance.• Technological advance is the prime driver of

economic growth.• Government has a responsibility to encourage

economic growth.

However, the economic underpinnings of gov-ernment’s role in innovation are more complexthan might first appear. From an economic per-spective, the justification for the role of govern-ment in innovation rests on a comparison of theefficiency of market resources with and withoutgovernment intervention.

Economic rationale for government involvement

Even today, many policy makers and academicspoint to Science—The Endless Frontier to date theorigins of U.S. science and technology policy.25

Vannevar Bush did not articulate a science andtechnology policy, and he did not articulate aneconomic rationale for government’s role. Rather,Bush like his predecessors and contemporariessimply assumed that the government had a roleto play in the innovation process, and then he setout to describe that role (as opposed to a rationalefor that role).

The first U.S. policy statement on science andtechnology was issued in 1990 by President Bush,U.S. Technology Policy. As with any initial policyeffort, this was an important general document.However, precedent aside, it too failed to articu-late a rationale for government’s intervention intothe private sector’s innovation process. Rather,like Science—The Endless Frontier and subsequentreports, it implicitly assumed that government hadsuch a role, and it then set forth a rather generalgoal (1990, p. 2):

The goal of U.S. technology policy is to make the best useof technology in achieving the national goals of improvedquality of life for all Americans, continued economicgrowth, and national security.

President Clinton took a major step forward inhis 1994 Economic Report of the President by firstarticulating principles about why the governmenthas a role in innovation and in the overall techno-logical process (p. 191):

Technological progress fuels economic growth � � � TheAdministration’s technology initiatives aim to promotethe domestic development and diffusion of growth- andproductivity-enhancing technologies. They seek to correctmarket failures that would otherwise generate too littleinvestment in R&D � � � The goal of technology policy isnot to substitute the government’s judgment for that of pri-vate industry in deciding which potential “winners” to back.Rather the point is to correct [for] market failure � � � .

Although the 1994 Economic Report of thePresident did not expand on how to correct formarket failure much less discuss appropriate pol-icy mechanisms for doing so, it did for the firsttime posit an economic rationale for governmentinvolvement in the innovation process. Even morerecently, the 1998 so-called Ehlers report, Unlock-ing Our Future, fails to articulate a clear rationalefor government’s involvement although it vastlydeparts in a positive way from the linear modelof technology development proffered by Bush.

Barriers to technology and market failure

Market failure refers to conditions under whichthe market, including those performing R&Dand those adopting the R&D outputs of others,underinvests from society’s perspective in any par-ticular technology. Such underinvestment occursbecause of conditions that prevent firms from fullyrealizing or appropriating the benefits expectedfrom their investments. Stated alternatively, firmsunderinvest in R&D when they determine, basedon their expectations of post-innovation activity,that their private return is less than their privatehurdle rate (minimum acceptable rate of returnon their R&D investment). This is of public con-cern when the R&D investment not undertaken issocially desirable.

There are a number of factors that explain whya firm might perceive its private return to be lessthan its hurdle rate. These factors are what we callbarriers to technology and they relate to technicaland market risk, where risk is defined to mea-sure the possibility that an actual outcome willdeviate from an expected outcome. Also, a firm


may believe that it cannot appropriate a sufficientreturn on its R&D investment, even if it can over-come technical uncertainty. This may be due toa perceived inability to maintain proprietary con-trol of the technology, thus enabling rivals to imi-tate their invention and reducing any resultingprofitability. Individually or in combination, thefollowing factors contribute to why a firm mayperceive a private rate of return as being less thanits hurdle rate26:

• Because of high technical risk, the outcome ofthe R&D may not solve the technical problemadequately to meet perceived needs.

• Even if the R&D is technically successful, themarket may not accept the technology becauseof competing alternatives or interoperabilityconcerns.

• Even absent technical and market risk, it maybe difficult to assign intellectual property rightsto the technology, and it might be quickly imi-tated so that the innovator may not receiveadequate return on the R&D investment.

These factors create barriers to investing in tech-nology, which thus lead to market failure.

From an economic perspective, the role of gov-ernment is to correct for these market failures inthose instances where society will benefit from thetechnology. This latter situation occurs when therate of return to society is greater than the socialhurdle rate (minimum accepted rate of returnfor society from investments of resources withalternative social uses), as illustrated in Figure 6.27

The social rate of return is measured on thevertical axis along with society’s hurdle rate oninvestments in R&D. The private rate of returnis measured on the horizontal axis along with theprivate hurdle rate on investments in R&D. A 45-degree line (long dashed line) is imposed on thefigure. This is the point at which the social rate ofreturn from an R&D investment equals the pri-vate rate of return from that same investment. Thearea above the 45-degree line and to the left of theprivate hurdle rate is the area of policy interest.

Three R&D projects are labeled in the figure.As drawn, the private rate of return exceeds theprivate hurdle rate for project C, and the socialrate of return exceeds the social hurdle rate. Thegap, measured vertically, between the social andprivate returns reflects the spillover benefits to

Social Rateof Return

Private Hurdle Rate

A B C

S

Private Rate of Return

ocial Hurdle Rate

450

Figure 6. Gap between social and private rates of return toR&D projects.

society from the private R&D investment. How-ever, the inability of the private sector to appropri-ate all benefits from its investment is not so greatas to prevent the project from being adequatelyfunded by the private firm.

In general, then, any R&D project that isexpected to lead to a technology with a privaterate of return to the right of the private hurdlerate and on or above the 45-degree line is nota candidate for public support, and hence gov-ernment does not have an intervening role froman economic perspective. Even in the presence ofspillover benefits, project C will be funded by theprivate firm.

For projects A and B, the gap between thesocial and private returns is larger than in the caseof project C; neither project A nor B will be ade-quately funded by the private firm as evidenced bythe private rate of return being less than the pri-vate hurdle rate. To address this market failure,the government has several policy mechanisms.

Before examining specific policy mechanisms,a critical question is: Do private-sector firms, ordoes industry, underinvest in R&D? Alternativelystated: Are there many projects like A and B thatthe private sector considers but rejects for somereason?

Private firms may not pursue promising techni-cal opportunities for the following reasons:

• R&D scientific and technical frontiers are riskyand the chances of failure are high.

• An individual firm may not have the capa-bilities required to develop the technology.


Complex new technologies may require col-laboration and information sharing; however,the cost of establishing research and develop-ment partnership and making then work pro-ductively may provide disincentives to under-taking the project.

• Private incentives may not be sufficient toinduce a firm to undertake the project in theface of difficulties in appropriating the result-ing benefits (i.e., the resulting knowledge mayfollow to others who may benefit from theR&D without sharing the cost).

Government has at its disposal at least four pol-icy mechanisms to reduce risk and market fail-ure and thus overcome an underinvestment inR&D, where underinvestment refers to the situ-ation where the private sector invests less in R&Dthan society would like it to invest (such as forproject A and project B). These policy mecha-nisms include:

• Patent laws• Tax incentives• Improved environment for collaborative re-

search• Subsidies to fund the research

Each mechanism is discussed in the sections thatfollow.

The U.S. patent system is more of an institu-tion than a policy mechanism. Thus, patent laws,in a sense, characterize the overall innovative envi-ronment in which firms operate. In contrast, taxincentives, efforts to stimulate collaborative R&D,and direct and indirect government subsidies arein a sense policy levers.

7. The patent system

The history of the U.S. patent system dates to theauthority given to Congress in the Constitution ofthe United States.28 Article I, section 8 states:

Congress shall have power � � � to promote the progressof science and useful arts, by securing for limited times toauthors and inventors the exclusive right to their respectivewritings and discoveries.

Based on this authority, Congress initiated a num-ber of patent laws.29 The version of law that is nowin effect was enacted on July 19, 1952 (to be effec-tive January 1, 1953). It is Title 35 in the UnitedStates Code.

The Patent and Trademark Office issuespatents for inventions. The patent term is 20 years,and it grants exclusive property rights to the inven-tor over that period of time. Patents are effectiveonly within the United States and its territoriesand possessions.

U.S. law is clear about what can be patented.Any person who:

invents or discovers any new or useful process, machine,manufacture, or composition of matter, or any new anduseful improvement thereof, may obtain a patent.

It is important to note the word “useful,” recallingtoo that Franklin created the American Philosoph-ical Society of Philadelphia in 1742 to promote“useful knowledge.” One cannot patent an idea.30

While the U.S. Code applies to patents in theUnited States, and in the territories and posses-sions of the United States, treaties have been pro-mulgated to extend protection beyond nationalboundaries. The Paris Convention for the Pro-tection of Intellectual Property of 1883 providedthat each of the 140 signatory nations recognizesthe patent rights of other countries. Subsequenttreaties have extended such coverage and madefilings in other countries more efficient.

Figure 7 illustrates the economics of patentingfrom the perspective of the firm. The marginalprivate rate of return to R&D is measured onthe vertical axis and the level of R&D spendingis measured on the horizontal axis. The marginalprivate return schedules are downward slopingreflecting diminishing returns to R&D in any giventime period, and for simplicity we assume themarginal private return schedule to be linear.

Marginal Private Rate of Return

Marginal Private Cost

Marginal Private Return with a patent

Marginal Private Return without a patent

R&D Expenditures RD0 RD1

Figure 7. Economics of patenting: increasing marginal privatereturn for the firm.


Absent the patent system, the firm will choose toinvest RD0 in R&D. This is an optimal invest-ment for the firm; it invests to the point whereits marginal cost of R&D (assumed to be constantfor simplicity) equals its expected marginal return.One might think of RD0 as a level of R&D thatthe firm is willing to invest. Is it, however, at alevel insufficient to generate a socially desirabletechnology? If so, we then have the case of projectB in Figure 6: the R&D expenditure is insufficientfor the project to be undertaken. Level RD1 inFigure 7 is sufficient for project B to be under-taken, but the firm does not have an incentive toinvest in R&D at that level. The key point is thatthe existence of the patent, or more precisely, theexpectation that the firm will be awarded a patent,can induce the firm to devote additional resourcesto R&D (to reach level RD1.)

Receipt of monopoly power for 20 yearsthrough a patent increases the firm’s marginal pri-vate return from its investments in R&D, thusshifting the marginal private return schedule tothe right.

Several trends in patenting activity in theUnited States are noteworthy:

• In the early 1980s, the number of patentsawarded to U.S. inventors began to declineand the number of U.S. patents awarded toforeign inventors began to rise, thus causingsome policy makers to question the inventiveenvironment in U.S. firms. This trend was yetanother indicator that U.S. global competitive-ness was declining.

• During the 1980s, the number of U.S. patentsawarded to foreign inventors was greatest forJapanese inventors. In fact, in 1995, over20,000 patents were awarded to Japaneseinventors compared to about 7,000 for the nexthighest represented country, Germany.

• The share of total patents awarded to foreigninventors is low in the United States comparedto other countries. It is highest in Italy andCanada and lowest in Japan and Russia.

• During the past decade, Japanese inven-tors have more international patents in threeimportant technologies than any country, withthe United States being second. These tech-nologies are: robotics, genetic engineering, andadvanced ceramics.

• Since the enactment of the Bayh-Dole Act of1980, which transferred ownership of intellec-tual property from federal agencies to univer-sities, there has been a rapid rise in universitypatenting (see Henderson et al., (1998).

Researchers have investigated a number ofaspects related to patenting activity, and the eco-nomic role of patents in the innovation pro-cess. Some significant findings from this bodyof research are summarized in Table VII. Onekey result is that the value of patents is highlyskewed, when citations (see Trajtenberg, 1990a;Jaffe et al., 1993; and Henderson et al., 1998) areused as an indicator of value. This result sug-gests that the use of counts of patents as an indi-cator of innovative output can be misleading ifthey are not properly deflated. Another criticalfinding is the existence of a strong positive cor-relation between R&D expenditure and patents.Finally, most studies report a positive correla-tion between patent activity and various measuresof economic performance, including productivity,

Table VIISelected literature on patent activity

Findings Authors

Strong positive correlation Scherer (1965)between R&D expenditure Schmookler (1966)(or employment) and patents Scherer (1983)

Bound et al. (1984)Pakes andGriliches (1984)

Hall et al. (1986)Acs and Audretsch(1989)

Positive correlation between patents Griliches (1981)and market value (stock market Hirschey (1982)rate of return and Tobin’s Q) Ben-Zion (1984)

Pakes (1985)Cockburn andGriliches (1988)

Austin (1993)

Value of patents is highly skewed, Trajtenberg (1990a)where value is determined by Jaffe et al. (1993)citations Henderson et al.

(1998)Jaffe et al. (1998)

Citation-weighted measures of Hall et al. (2000)patents are more highly correlatedwith market value thanunweighted measures


measures of accounting profitability, Tobin’s Q,and stock prices.

A recent paper by Hall et al. (2000) pro-vides an important extension of this literatureon estimation of the private returns (the returnsthat accrue to firms) to patenting by attemptingto link citation-weighted patents to the marketvalue of firms. Their preliminary results suggestthat citation-weighted measures of patents aremore highly correlated with market value thanunweighted measures.

8. Tax incentives as a policy tool

Tax incentives are another mechanism that gov-ernment uses to increase private sector R&D.Like any policy tool, tax incentives have advan-tages and disadvantages.31 Advantages include thefollowing:

• Tax incentives entail less interference in themarketplace than do other mechanisms, thusaffording private-sector recipients the abilityto retain autonomy regarding the use of theincentives.

• Tax incentives require less paperwork thanother programs.

• Tax incentives obviate the need to directly tar-get individual firms in need of assistance.

• Tax incentives have the psychological advan-tage of achieving a favorable industry reaction.

• Tax incentives may be permanent and thus donot require annual budget review.

• Tax incentives have a high degree of politicalfeasibility.

Some disadvantages of tax incentives are:

• Tax incentives may bring about unintendedwindfalls by rewarding firms for what theywould have done in the absence of the incen-tive.

• Tax incentives often result in undesirable in-equities.

• Tax incentives raid the federal treasury.• Tax incentives frequently undermine public

accountability.• The effectiveness of tax incentives often varies

over the business cycle.

The economics of tax credits

Figure 8 illustrates the economics of a tax credit.The marginal rate of return is measured onthe vertical axis and the level of R&D spend-ing is measured on the horizontal axis. Both themarginal social return and the marginal privatereturn schedules are downward sloping reflect-ing diminishing returns to R&D investments in agiven time period. The social return schedule isdrawn greater than the private return schedule forall levels of R&D because firms cannot appropri-ate all the benefits from conducting R&D; someof those benefits spillover to other firms in the cur-rent time period and in the post-innovation timeperiod thus generating additional benefits to soci-ety. The marginal cost to the firm to undertakeR&D is shown to be constant (horizontal).

As drawn, the firm will equate its marginalcost to conduct R&D with its marginal return,and the firm will invest RD0. One might think ofRD0 as corresponding to a partial level of fundingfor project B in Figure 6, much like the case ofthe patent illustration in Figure 7. Society, giventhe firm’s marginal cost schedule, would like thefirm to invest in R&D to maximize social benefits.Hence, the optimal credit is one that provides anincentive to the firm to increase its R&D to pointRD1. Receipt of a tax credit can be thought ofas a reduction in the marginal cost of undertak-ing additional R&D, and the firm will re-equateits new marginal cost with its marginal returnand invest at RD1. Unlike patents, the researchand experimentation (R&E) credit does not cor-rect for market failure. It simply increases thefirm’s private return on marginal R&D projects

Marginal Rate of Return

Marginal Private Cost

Marginal Social Return

Marginal Private Return

R&D ExpendituresRD0 RD1

Figure 8. Economics of a tax credit: decreasing marginal pri-vate cost.


by reducing its marginal private cost to undertakesuch projects. Thus, tax incentives will increase thefirm’s level of R&D from RD0 to RD1 but will notalleviate technical or market risk.

However, because R&D is not a homogeneousactivity and because the research (R) portion ofR&D has a greater impact on productivity growthand hence economic growth than does develop-ment (D), any uniform tax incentive that treatsR&D as if it were a homogeneous activity willlikely encourage more of the same mix of R&D.That may not be bad in the case of R&D sinceeconomic studies suggest that the marginal privatereturn from R&D is greater than the marginalcost of conducting R&D. However, a tax credit onresearch as opposed to development, while con-ceptually more desirable, could be cumbersome toadminister.

R&E tax credit

The adoption of Section 174 of the InternalRevenue Code in 1954 codified and expandedtax laws pertaining to the R&D expendituresof firms. This provision permitted businesses todeduct fully R&E expenditures but not devel-opment or research application expenditures inthe year incurred.32 Businesses under Section 174are not allowed to expense R&E related equip-ment. Such equipment must be depreciated. How-ever, the Economic Recovery and Tax Act of1981 (ERTA) provided for a faster depreciationof R&E capital assets than other business capitalassets.

ERTA also included a 25 percent tax creditfor qualified R&E expenditures in excess of theaverage amount spent during the previous threetaxable years or 50 percent of the current year’sexpenditures (the R&E base). The initial R&Etax credit had several limitations including thefact that it did not cover expenses related to theadministration of R&D or to research conductedoutside of the United States. The Tax Reform Actof 1986 modified these limitations, but reducedthe marginal rate from 25 percent to 20 percent.Over the years the credit had been modified, pri-marily in terms of the definition of the R&E base,but the credit has never been made permanent. Ithas expired a number of times, only to be renewedretroactively.

In 1996, the Congressional Office of Technol-ogy Assessment released a report on the effec-tiveness of the R&E tax credit. The report con-cluded:

• There is not sufficient information availableto conduct a complete benefit-cost analysis ofthe effectiveness of the R&E tax credit on theeconomy.

• The econometric studies that have been doneto date conclude that the credit has been effec-tive in the sense that for every dollar lost infederal revenue there is an increase of a dollarin private sector R&D spending. These studiesconclude that the credit would be more effec-tive if it were made permanent.33

• The R&E tax credit represents a small fractionof federal R&D expenditures, about 2.6 per-cent of total federal R&D funding and about6.4 percent of federal R&D for industry.

The R&E tax credit is not unique to the UnitedStates. Japan’s tax credit is marginal, and it wasinitiated in 1966. Canada also initiated a pro-gram in the 1960s, but their program is a flat taxprogram.

Economic theory predicts that firms that coop-erate in a research joint venture type of arrange-ment have an incentive to cooperate at theresearch end of the R&D spectrum rather than atthe development end. Thus, some have proposeda tax credit for cooperative research involvementas a viable alternative to the R&E tax credit.34

9. Research collaborations

Research partnerships, meaning formal or infor-mal collaboration among firms in the conduct ofresearch, are an organizational form that mayovercome elements of market failure by reduc-ing technical risk to the R&D-conducting firms byenlarging the underlying knowledge base. Thesepartnerships may also limit market risk by helpingto ensure that particular elements of the technol-ogy are standardized and thus interoperate in asystem. As well, to the extent that collaborationreduces redundant research, there may be costsavings to each partner, reduced time to market,and better appropriability of R&D results.

Research partnerships assume many forms inpractice. Partners may aim to develop or refine


a new product, improve production processes, setstandards, or develop technology to meet environ-mental regulations. The collaboration may takeplace between partners who compete in the mar-ketplace, or between partners who produce com-plementary products. Some partnerships includegovernment or university partners as well.

Research partnerships are certainly not a neworganizational form, but since the mid-1980s gov-ernment has provided a favorable environment forthem. Before reviewing the related policies, wedescribe one research collaboration in detail as anillustration of the types of collaborations that areprevalent and that have had a major impact onresearch.

Semiconductor research corporation

One of the first formal research collaborationsin the United States was the SemiconductorResearch Corporation (SRC). A brief history ofthe SRC will serve to illustrate that many researchcollaborations or partnerships are formed toaddress industry-wide technological issues, or atleast issues that affect a sizeable segment of theindustry. This history is also interesting because itillustrates a purposeful entrepreneurial responseto competitive market conditions.35

In the late 1950s, an integrated circuit (IC)industry emerged in the United States. Thefledgling industry took form in the 1960s and expe-rienced rapid growth throughout the 1970s. In1979, when Japanese companies captured 42 per-cent of the U.S. market for 16 kbit DRAMs(memory devices) and converted Japan’s inte-grated circuit trade balance with the United Statesfrom a negative $122 million in 1979 to a posi-tive $40 million in 1980, the U.S. industry becamepainfully aware that its dominance of the IC indus-try was being seriously challenged. It was clear toall in the industry that it was in their collectivebest interest to invest in an organizational struc-ture that would strengthen the industry’s positionin the global semiconductor marketplace.

The Semiconductor Industry Association (SIA)was formed in 1977 to collect and assemble reli-able information on the industry and to developmechanisms for addressing industry issues withthe federal government. In a presentation at anSIA Board Meeting in June 1981, Erich Bloch of

IBM described to the industry the nature of thegrowing competition with Japan and proposed thecreation of a “semi-conductor research coopera-tive” to assure continued U.S. technology leader-ship. This event witnessed the birth of the SRC. InDecember 1981, Robert Noyce, then SIA chair-man and vice-chairman of Intel, announced theestablishment of the SRC for the purpose of stim-ulating joint research in advanced semiconduc-tor technology by industry and U.S. universitiesand to reverse the declining trend in semiconduc-tor research investments. The SRC was formallyincorporated in February 1982 with a stated pur-pose to36:

• Provide clearer view of technology needs.• Fund research to address technology needs.• Focus attention on competition.• Reduce research redundancy.

Policy makers soon noticed the virtues of coop-erative research in part because such organiza-tional structures had worked well in Japan andin part because the organizational success of theSRC demonstrated that cooperation among com-petitive firms at the fundamental research levelwas feasible.

Public policy toward research collaborations

To place the activities surrounding the SRC’s for-mation in a broader perspective, recall that in theearly 1980s there was growing concern about thepersistent slowdown in productivity growth thatfirst began to plague the U.S. industrial sectorin the mid-1970s and about industry’s apparentloss of its competitive advantage in world markets,especially firms in the semiconductor industry.37

As noted in a November 18, 1983 House reportabout the proposed Research and DevelopmentJoint Ventures Act of 1983:

A number of indicators strongly suggest that the position ofworld technology leadership once firmly held by the UnitedStates is declining. The United States, only a decade ago,with only five percent of the world’s population was gener-ating about 75 percent of the world’s technology. Now, theU.S. share has declined to about 50 percent and in anotherten years, without fundamental changes in our Nation’stechnological policy � � � the past trend would suggest thatit may be down to only 30 percent. [In hearings,] many dis-tinguished scientific and industry panels had recommendedthe need for some relaxation of current antitrust laws to


encourage the formation of R&D joint ventures. � � � Theencouragement and fostering of joint research and devel-opment ventures are needed responses to the problem ofdeclining U.S. productivity and international competitive-ness. According to the testimony received during the Com-mittee hearings, this legislation will provide for a significantincrease in the efficiency associated with firms doing similarresearch and development and will also provide for moreeffective use of scarce technically trained personnel in theUnited States.

In an April 6, 1984 House report on competinglegislation, the Joint Research and DevelopmentAct of 1984, the supposed benefits—and recallthat at this time it was still too soon for there tobe visible benefits coming from the SRC’s activi-ties on behalf of the IC industry—of joint researchand development were for the first time clearlyarticulated:

Joint research and development, as our foreign competitorshave learned, can be procompetitive. It can reduce duplica-tion, promote the efficient use of scarce technical person-nel, and help to achieve desirable economies of scale. � � �[W]e must ensure to our U.S. industries the same eco-nomic opportunities as our competitors, to engage in jointresearch and development, if we are to compete in theworld market and retain jobs in this country.

The National Cooperative Research Act(NCRA) of 1984, after additional revisions in theinitiating legislation, was passed on October 11,198438:

� � � to promote research and development, encourage inno-vation, stimulate trade, and make necessary and appropri-ate modifications in the operation of the antitrust laws.

The NCRA created a registration process, laterexpanded by the National Cooperative Researchand Production Act (NCRPA) of 1993, underwhich research joint ventures (RJVs) can disclosetheir research intentions to the Department ofJustice. RJVs gain two significant benefits fromsuch voluntary filings: if subjected to criminal orcivil action they are evaluated under a rule of rea-son that determines whether the venture improvessocial welfare; and if found to fail a rule-of-reasonanalysis they are subject to actual rather thantreble damages.

One of the more notable RJVs formed andmade public through the NCRA disclosure pro-cess was SEMATECH (SEmiconductor MAnufac-turing TECHnology). It was established in 1987as a not-for-profit research consortium with anoriginal mission to provide a pilot manufacturing

facility where member companies could improvetheir semiconductor manufacturing process tech-nology. Its establishment came after the DefenseScience Board recommended direct governmentsubsidy to the industry in a 1986 report com-missioned by the Department of Defense (there-fore SEMATECH is discussed below under thebroader heading of a public/private technologypartnership). It was thought that SEMATECHwould be the U.S. semiconductor industry’s/U.S.government’s response to the Japanese govern-ment’s targeting of their semiconductor indus-try for global domination. Since its inception,SEMATECH’s stated mission has evolved andbecome more general. The consortium currentlydefines its mission around solving the technicalchallenges presented in order to sustain a lead-ership position for the United States in the globalsemiconductor industry.

Trends in RJVs

To date, there have been over 800 formal RJVsfiled under the NCRA. Certainly, this number isa lower bound on the total number of researchpartnerships in the United States, even since 1984.Not all are as publicly visible as SEMATECH.Some are quite small, with only two or three mem-bers, and others are quite large with hundredsof members. On average, a joint ventures has 14members.39

While informal cooperation in research mayhave been prevalent in the United States fordecades, formal RJV relationships are new and itwill take longer than a decade and a half to detectmeaningful trends.

Albeit that research partnerships as formalentities are relatively new to the technology strat-egy arena, the literature concludes that thereare both benefits and costs to members of theventure.40 The benefits include:

• the opportunity for participants to captureknowledge spillovers from other members,

• reduced research costs due to a reduction induplicative research,

• faster commercialization since the fundamen-tal research stage is shortened, and

• the opportunity to form, in some cases,industry-wide competitive vision.


The costs include:

• a lack of appropriability since research resultsare shared among the participants, and

• managerial tension, in some cases, as partic-ipants learn to trust each other and to worktogether.

Research partnerships are correctly viewed asa complementary source of technical knowledgeand technical efficiency for the firm. Thus, firmsthat participate in a research partnership enhancetheir own R&D process through interactions.

Universities as research partners

Especially noticeable in the RJVs filed with theDepartment of Justice are the presence of univer-sities as research partners and that the numberof RJVs with at least one university partner hasincreased over the past 15 years.41 On average,about 15 percent of RJVs have at least one uni-versity research partner, and of these over 90 per-cent are U.S. universities. RJVs with universitiesas research partners have, on average, five univer-sity partners.42

The literature on universities as research part-ners is sparse. However, some stylized conclusionscan be drawn from the limited investigations43:

• Firms that interact with universities generallyhave greater R&D productivity and greaterpatenting activity.

• One key motive for firms to maintain jointresearch relationships with universities is tohave access to key university personnel—facul-ty as well as students as potential employees.

Many speculate that university participationmay increase in importance and frequency in thefuture. According to the Council on Competitive-ness (1996, pp. 3–4):

Over the next several years, participation in the U.S. R&Denterprise will have to continue experimenting with differ-ent types of partnerships to respond to the economic con-straints, competitive pressures and technological demandsthat are forcing adjustment across the board. � � � [and inresponse] industry is increasingly relying on partnershipswith universities, while the focus of these partnerships isshifting progressively toward involvement in shorter-termresearch.

And (Council on Competitiveness, 1996, p. 11):

For universities, cutbacks in defense spending have resultedin a de facto reallocation of funding away from the physicalsciences and engineering and shifted the focus of defenseresearch away from the frontiers of knowledge [e.g., basicscience] to more applied efforts. � � � Although defensespending is clearly not the only viable mechanism to sup-port frontier research and advanced technology, the UnitedStates has yet to find an alternative innovation paradigm toreplace it.

Given this spending trend, and the increasingease of global technology transfer, it is conceiv-able, at least according to the Council, that theUnited States may lose its technological leader-ship in some important areas such as health andadvanced materials since innovation in these fieldsis closely linked to improvements in basic science.

There is some indication that scholars arebeginning to think more deeply and more broadlyabout the social, economic, and technological con-sequences of university involvement in private sec-tor research partnerships. This thinking reflectssome major concerns about the impact of theserelationships on the research university’s missionto conduct basic research. Unfortunately, moreinformation is necessary in order for researchersto examine the ramifications of this trend from awide variety of disciplinary perspectives.

It is likely that the increasing trend towarduniversity private-sector research partnerships willcontinue. A 1993 national survey of U.S. biology,chemistry, and physics faculty members revealedthat many academic scientists desired more suchinvolvement. An earlier survey of engineering fac-ulty members reached the same conclusion. How-ever, one of the authors (Morgan, 1998, p. 169)of the surveys was quick to point out one area ofmajor concern:

� � � a diminution of the role of the university as an indepen-dent voice to help look out for the broader societal goodand to guard against industrial as well as other excesses.An independent science, engineering and public policy roleis essential to ensure an adequate supply of well educatedscientists and engineers prepared to work with the publicsector in public interest groups. Having industry assumea more central role as customer and client for university-based scientific and engineering research, while in someway a natural and desirable step, needs to be balancedagainst the need for independence, oversight and serviceto society and the larger public good.


Government laboratories as a research partner

The federal government also enters directly intoresearch partnerships with firms through the fed-eral laboratory system. This relationship can takevarious forms ranging from informal relationshipswhereby a firm(s) interacts with a federal lab-oratory scientist, or more formal relationshipswhereby a firm(s) utilizes federal laboratory facili-ties and is jointly involved with the laboratory sci-entists in the research. Or, the relationship canbe nothing more than an informational transferwhereby the firm utilizes public information thatwas generated within a government agency.

While very few studies have systematicallylooked at the economics of federal laboratoriesas research partners, three generalizations can bemade44:

• Federal laboratories are generally associatedwith research joint ventures that are large interms of other member companies.

• One key advantage to partnering with a fed-eral laboratory is access to specialized techni-cal equipment.

• Firms’ research with laboratories tends to benearer the basic research end of the R&Dspectrum and, related, firms generally viewaccess to the laboratories’ research scientists asa more important benefit than the technologiesavailable for licensing.

One of the few broad-based research pro-grams examining federal laboratories as researchpartners is work performed by Bozeman andcolleagues (summarized in Crow and Bozeman,1998). According to a study by Bozeman andPapadakis (1995), companies’ objectives for inter-acting with federal laboratories include engagingin strategic pre-commercial research (51 percent),interest in access to the unique resources of thelab (45 percent) and a desire to develop new prod-ucts and services (42 percent). Their decisions towork with the federal laboratories are particularlyrelated to the skills and knowledge of the federallab’s scientists and engineers (61 percent). Manycompanies are “repeat customers” and 42 percentindicate that prior experience with the lab is oneof the major reasons for choosing to work with thelab in the more recent project.

Regarding the incidence of commercial out-comes, more than 22 percent of the projects

led to the development and marketing of a newproduct, a product development rate comparingfavorably to firms operating independently. Thosecompanies most likely to develop products fromtheir partnerships with federal laboratories hadthe following characteristics: (1) smaller than theaverage for all companies in the data base, (2)high levels of R&D intensity (R&D employeesas a percentage of total employees), (3) morerecently established firms (Bozeman and Wittmer,2002). Interestingly, while 89 percent of partici-pating companies reported a high degree of sat-isfaction with their federal laboratory partnership,those developing products were actually somewhatless satisfied, perhaps because the federal labora-tories are generally a better source for upstreambasic and applied research than for technologydevelopment (Rogers and Bozeman, 1997). Thisinterpretation comports with a finding from a sep-arate study using a different data set (Roessnerand Bean, 1991).

Comparing universities and governmentlaboratories as research partners

In recent times, much science and technologypolicy has striven to determine the respectiveadvantages and disadvantages of government lab-oratories and universities in cooperative researchand technology transfer. Absent such knowledge,it is difficult to make decisions about the alloca-tions of resources among institutional actors.

According to the directors of the federal labo-ratories, a major comparative advantage of federallaboratories is their ability to perform interdis-ciplinary team research. Generally, universities,organized on the disciplinary lines, have difficultyperforming research that cuts across disciplinesand traditional academic departments. This haschanged somewhat in recent years, however, hasuniversities have developed a wide array of inter-disciplinary centers, often in response to NationalScience Foundation initiatives for creating sciencecenters or engineering research centers. A secondmajor advantage of the federal laboratories, espe-cially the national labs, is that extremely expensive,often unique, scientific equipment and facilitiesare located on their premises. The “user facilities”at federal laboratories are designed explicitly toshare resources and these user facilities can be an


important instrument for technology transfer. Fewuniversities, or even university consortia, have theresources to build facilities national in scope.

The most obvious advantage of universitiesover federal laboratories is a vitally importantone—students. The presence of students makes aremarkable difference in the output, culture andutility of research. In recommending that federalfunding for science and technology give strongestemphasis to academic institutions, the NationalAcademy of Sciences’ Committee on Criteria forFederal Support of Research and Development(National Academy of Sciences, 1995, p. 20) con-cluded that university R&D funding supports pro-duction of “well prepared scientists and engineerswho not only will be the next generation of fac-ulty, but who will also work productively in, andtransfer technology to, industry and government.”

Students are a reservoir of cheap labor sup-porting university research, bartering their belowmarket wage rate for training. More importantfor present purposes is that students are a meansof technology transfer (through postgraduate jobplacements) and they often provide enduring linksas the social glue holding together many fac-ulty scientists and the companies they work with.Roessner and Bean (1991) found that the singlemost important benefit to industry from partic-ipation in the NSF Engineering Research Cen-ters, according to the industrial participants them-selves, is the ability to hire ERC students andgraduates. In some cases, the vast benefits accru-ing from students are enjoyed by governmentlaboratories, but chiefly at such institutions asLawrence Berkeley Lab or Ames Laboratory,those actually located on university campuses. Weshall return to this issue subsequently in a discus-sion of the role of “scientific and technical humancapital” (Rogers and Bozeman, 1997).

10. Public/private technology partnerships45

One can trace the origins of public/privatepartnerships—federal grants assistance technologypartnerships—in the United States at least as farback as the Lincoln Administration. In 1862, theMorrill Act established what was known as theland-grant college system. The Act created a part-nership between the federal and state govern-ments to cooperate with the private sector in tech-nology development. The Act charged states to

develop colleges to offer curricula in agricultureand mechanical arts. Then in 1887, the Hatch Actprovided resources for a system of state agricul-ture experiment stations that would be under theauspices of land-grant colleges and universities. Apartnership among the various levels of govern-ment was established by the Smith-Level Act of1914. The Cooperative Agriculture Extension ser-vice was charged to deliver the practical benefits ofresearch to citizens through an extension service.

According to Carr (1995, p. 11):

Until the end of the 1970s, the philosophy behind the dis-semination of federally-funded research was that if the pub-lic paid for the research, the resulting intellectual propertyshould be made equally available to all interested parties.

Because the 1970s and 1980s witnessed manyforeign competitors beginning to successfully chal-lenge the long-standing dominance of the UnitedStates both in world and domestic markets, itbecame clear to public and private policymakersthat a change in the philosophy of federal R&Dsupport was needed. The Office of TechnologyPolicy (1996) reflected on the motivation for thischange in policy mindset as:

• Global competitors were better able toappropriate the output of U.S. basic andmission-oriented research as their technicalsophistication grew.

• Traditional public-sector mechanisms of tech-nology development, transfer, and develop-ment took too long in an era of acceleratingprivate sector development.

• U.S. federal R&D represented a declin-ing share of the world R&D as globally-competitive nations increased their publicfunding; hence the marginal benefits to indus-try from additional public moneys (allocated inthe same historical manner) declined.

Stated alternatively, but maintaining the samegeneral theme, Link and Tassey (1987, pp. 4, 131)reflected on the changing competitive environ-ment of U.S. industry in the early 1980s:

� � � there is a new order of competition in the world. Aninescapable element of the competition is technology � � �

With the advent of technology-based economies [through-out the world], the increase in the number of world com-petitors has been greater than the increase in the size of theworld market. What has resulted from this is a significantshortening of technology life cycles � � � As such, effective


long-run competitive strategies will have to deal explicitlywith technology � � � [C]ompetitive survival will depend ontechnology-based strategies. These strategies will have toevolve from new philosophies about interdependence � � �

The importance of interdependence arises from the needof a domestic industry to rapidly and efficiently developcomplex technological elements from which specific appli-cations (innovations) are drawn for competitive activity� � � [Accordingly,] government must expand and adapt itsrole � � � with industry for more effective joint planning inresearch.

Beginning with legislation in the 1980s, as sum-marized in Table VIII, a new era in federal tech-nology policy began. This new era was basedon the belief that the global competitiveness ofU.S. firms can be enhanced through legislation tobolster the commercial impact of federal R&Dinvestments. As such, using the terminology of theOffice of Technology Policy (1996, p. 26), “a newparadigm for public/private technology partner-ships emerged.” This new paradigm viewed indus-try as a partner in the formation and execution oftechnology programs rather than a passive recipi-ent of the output from federal research.

Several public/private partnerships are over-viewed below from an institutional perspective.SEMATECH is discussed in detail because it wasone of the earliest public/private partnerships andfor years was heralded as the model organizationalform for other public/private partnerships to fol-low. The Small Business Innovation Research Pro-gram and the Advanced Technology Program arealso singled out for discussion because these aretwo of the programs that have conducted in-depthprogram evaluations and hence more is knownabout their economic impacts than about thoseassociated with other partnerships.

SEMATECH46

In 1986 when the Semiconductor Industry Associ-ation (SIA) and the Semiconductor Research Cor-poration (SRC) began to explore the possibilityof joint industry/government cooperation, the U.S.semiconductor industry was not in a favorable eco-nomic position. During 1986, Japan overtook theUnited States for the first time in terms of theirshare of the world semiconductor market. Japanhad about 45 percent of the world market com-pared to about 42 percent for the United States.The U.S. semiconductor industry expected Japan’s

share to grow at the expense of that of the UnitedStates. In January 1997, President Reagan recom-mended $50 million in matching federal fundingfor R&D related to semiconductor manufactur-ing, and this was to be part of the Department ofDefense’s 1988 budget. Soon thereafter, the SIAapproved the formation of SEMATECH and theconstruction of a world-class research facility.47

In September 1987, Congress authorized $100 mil-lion in matching funding for SEMATECH.

SEMATECH and its members have a missionto:

� � � create a shared competitive advantage by workingtogether to achieve and strengthen manufacturing technol-ogy leadership.

This shared vision is accomplished by joint spon-sorship of leading edge technology developmentin equipment supplier companies. As these com-panies become world-class manufacturers, so willthe members of SEMATECH.

By 1988, Japan’s world market share reachedover 50 percent, and that of the United Statesfell to about 37 percent. The U.S. share declinedagain in 1989 and then it began to increase atthe expense of that of Japan. Early in 1992, theUnited States was again at parity with Japan atabout 42 percent, and stayed slightly ahead ofJapan until 1995 when the gap began to widen.

The mid-1990s saw increasing cooperationbetween U.S. and Japanese semiconductor com-panies, and in fact, in 1998 International SEMAT-ECH began operations with Hyundai (Japan) andPhilips (Amsterdam) as important members. Also,beginning in 1998, all funding came only frommember companies.

Small Business Innovation Research Program48

The Small Business Innovation Research (SBIR)program began at the National Science Founda-tion (NSF) in 1977. At that time the goal ofthe program was to encourage small businesses,long believed to be engines of innovation in theU.S. economy, to participate in NSF-sponsoredresearch, especially research that had commercialpotential. Because of the early success of the pro-gram at NSF, Congress passed the Small BusinessInnovation Development Act of 1982. The Actrequired all government departments and agencieswith external research programs of greater than


Table VIIISelected public/private technology partnership legislation

Enabling legislation Characteristics of the program

Stevenson-Wylder TechnologyInnovation Act of 1980

The Act was predicated on the premise that federal laboratories embody important andindustrially-useful technology. Accordingly, each federal laboratory was mandated toestablish an Office of Research and Technology Application to facilitate the transferof public technology to the private sector.

University and Small BusinessPatent Procedure Act of 1980

This Act is also known as the Bayh-Dole Act. It reformed federal patent policy by pro-viding increased incentives for the diffusion of federally-funded innovation results.Universities, non-profit organizations, and small businesses were permitted to obtaintitles to innovations they developed with the use of governmental financial support,and federal agencies were allowed to grant exclusive licenses to their technology toindustry.

Small Business InnovationDevelopment Act of 1982

This Act required federal agencies to provide special funds to support small businessR&D that complemented the agency’s mission. These programs are called SmallBusiness Innovation Research (SBIR) programs. The Act was reauthorized in 1992.

National Cooperative ResearchAct of 1984

NCRA was legislated in an effort to reduce antitrust barriers and thus encourage theformation of joint research venture among U.S. firms. This Act was amended bythe National Cooperative Research and Production Act of 1993, thereby expandingantitrust protection to joint production ventures.

Trademark ClarificationAct of 1984

This Act set forth new licensing and royalty regulations to take technologyfrom federally-funded facilities into the private sector. It specifically permittedgovernment-owned, contractor-operated (GOCO) laboratories to make decisionsregarding which patents to license to the private sector, and contractors could receiveroyalties on such patents.

Federal Technology TransferAct of 1986

This Act was amended by the Stevenson-Wylder Act. It made technology transfer anexplicit responsibility of all federal laboratory scientists and engineers. It autho-rized cooperative research and development agreements (CRADAs). This Act wasamended by the National Competitiveness Technology Transfer Act of 1989 whichexpanded the definition of a federal laboratory to include those that are contractoroperated.

Omnibus Trade and CompetitivenessAct of 1988

This Act established two competitiveness programs, the Advanced Technology Program(ATP) and the Manufacturing Extension Partnership (MEP) within the re-namedNational Institute of Standards and Technology (NIST).

Defense Conversion, Reinvestment,and Transition Assistance Actof 1992

This Act created an infrastructure for dual-use partnerships. Through Technology Rein-vestment Project partnerships the Department of Defense was given the ability toleverage the potential advantages of advanced commercial technologies to meetdepartmental needs.

$100 billion to establish their own SBIR programsand to set aside funds equal to 0.2 percent ofthe external research budget.49 Currently, agenciesmust allocate 2.5 percent of the external researchbudget to SBIR.

The 1982 Act states that the objectives of theprogram are:

1. to stimulate technological innovation2. to use small business to meet Federal research

and development needs

3. to foster and encourage participation by minor-ity and disadvantaged persons in technologicalinnovation

4. to increase private sector commercialization ofinnovations derived from federal research anddevelopment.

The Act was reauthorized in 1992.SBIR awards are of three types. Phase I awards

are small, generally less than $100,000. The pur-pose of these awards is to assist firms to assessthe feasibility of the research they propose to


undertake for the agency in response to theagency’s objectives. Phase II awards can rangeup to $750,000. These awards are for the firmto undertake and complete its proposed research,hopefully leading to a commercializable productor process.

The Department of Defense’s (DoD’s) SBIRprogram has been studied in some detail. It can beconcluded that DoD’s SBIR program in encour-aging commercialization from research that wouldnot have been undertaken without SBIR support.And moreover, the structure of DoD’s SBIR pro-gram is overcoming reasons for market failure thatpreviously have caused small firms to underinvestin R&D.50

Advanced Technology Program

The Advanced Technology Program (ATP) wasestablished within the National Institute ofStandards and Technology (NIST) through theOmnibus Trade and Competitiveness Act of 1988,and modified by the American Technology Preem-inence Act of 1991. The goals of the ATP, as statedin its enabling legislation, are to assist U.S. busi-nesses in creating and applying the generic tech-nology and research results necessary to:

1. commercialize significant new scientific discov-eries and technologies rapidly

2. refine manufacturing technologies.

These same goals were restated in the FederalRegister on July 24, 1990:

The ATP � � � will assist U.S. businesses to improve theircompetitive position and promote U.S. economic growthby accelerating the development of a variety of pre-competitive generic technologies by means of grants andcooperative agreements.

The ATP received its first appropriation fromCongress in FY 1990. The program fundsresearch, not product development. Commercial-ization of the technology resulting from a projectmight overlap the research effort at a nascentlevel, but generally full translation of the tech-nology into products and processes may take anumber of additional years. ATP, through costsharing with industry, invests in risky technologiesthat have the potential for spillover benefits to theeconomy.

Appropriations to ATP increased from $10 mil-lion in 1990 to a peak of $341 million in 1995.Funding decreased in 1996 to $221 million, andhas averaged about $200 million per year until2000 when it fell to just under $150 million. Todate, ATP has funded through competitive pro-cesses approximately 450 research projects.

Much like the case of DoD’s SBIR program,ATP has provided incentives to firms to under-take research that would not otherwise have beenpursued—such as projects A or B in Figure 6.

11. Infrastructure technology

Infrastructure technology is a term that refersto the technological environment in which firmsinnovate. Two important parts of the techno-logical environment of firms are patent laws(previously discussed) and the federal laboratorysystem (publicly funded). The federal laboratorysystem’s output is infrastructure technology, whichby its nature is a public good, freely accessible byfirms.

Federal laboratory system

As shown above in Table II, the federal govern-ment allocated $65.8 billion to R&D in 1999, ofwhich $17.4 billion was performed by the fed-eral government. Most of this research occurredin the more than 700 federally-funded R&D lab-oratories in the country.51 The Department ofDefense’s laboratories account for almost one-half of this intramural research, and the Depart-ment of Health and Human Services accounts forabout 15 percent. Table IX provides some sum-mary information about a few of the laboratoriesin the federal laboratory system.

The governmental agency that provides a sig-nificant amount of direct infrastructure technologyto the industrial economy is the National Instituteof Standards and Technology (NIST) within theDepartment of Commerce.52 NIST is discussedherein in more detail than any other federal lab-oratory for two reasons. One, its research mis-sion is varied by field of science, as opposed tobeing focused on one major scientific area such asenergy; and two, the Program Office within NISThas a long history of evaluation of the outputs


Table IXOverview of selected national laboratories

Laboratory Year established Research Ownership

Ames Laboratory 1942 Energy Department of Energy operatedby Iowa State University

Brookhaven National Laboratory 1947 Energy Department of Energy operatedby SUNY Stony Brook

Los Alamos National Laboratory 1943 National security Department of Energyand energy

National Institute of 1901 Standards Department of CommerceStandards and Technology

Rome Laboratory 1951 Communications Department of Defense

Source: Crow and Bozeman (1998).

from its research programs, and as such its con-tribution to economic growth is more readily doc-umented.

National Institute of Standards and Technology

Historical overview of NIST.53 A standard is a pre-scribed set of rules, conditions, or requirementsconcerning:

• Definitions of terms• Classification of components• Specification of materials, their performance,

and their operations• Delineation of procedures• Measurement of quantity and quality in

describing materials, products, systems, ser-vices, or practices.

To understand the current activities that takeplace at NIST, its public good mission must beplaced in an historical perspective. The concept ofthe government’s involvement in standards tracesto the Articles of Confederation signed on July 9,1778. In Article 9, §4:

The United States, in Congress assembled, shall also havethe sole and exclusive right and power of regulating thealloy and value of coin struck by their own authority, or bythat of the respective States; fixing the standard of weightsand measures throughout the United States � � �

This responsibility was reiterated in Article 1, §8of the Constitution of the United States:

The Congress shall have power � � � To coin money, regulatethe value thereof, and of foreign coin, and fix the standardof weights and measures � � �

On July 20, 1866, Congress and PresidentAndrew Johnson authorized the use of the metricsystem in the United States. This was formalizedin the Act of 28 July 1866—An Act to Authorizethe Use of the Metric System of Weights and Mea-sures:

Be it enacted � � � , That from and after the passage of thisact it shall be lawful throughout the United States of Amer-ica to employ the weights and measures of the metric sys-tem; and no contract or dealing, or pleading in any court,shall be deemed invalid or liable to objection because theweights or measures expressed or referred to therein areweights and measures of the metric system.

As background to this Act, the origins of themetric system can be traced to the research ofGabriel Mouton, a French vicar, in the late 1600s.His standard unit was based on the length of anarc of 1 minute of a great circle of the earth. Giventhe controversy of the day over this measure-ment, the National Assembly of France decreedon May 8, 1790, that the French Academy of Sci-ences along with the Royal Society of Londondeduced an invariable standard for all the mea-sures and all the weights. Within a year, a stan-dardized measurement plan was adopted basedon terrestrial arcs, and the term mètre (meter),from the Greek metron meaning to measure, wasassigned by the Academy of Sciences.

Because of the growing use of the metric systemin scientific work rather than commercial activ-ity, the French government held an internationalconference in 1872, which included the participa-tion of the United States, to settle on proceduresfor the preparation of prototype metric standards.Then, on May 20, 1875, the United States partici-pated in the Convention of the Meter in Paris and


was one of the eighteen signatory nations to theTreaty of the Meter.

In a Joint Resolution before Congress onMarch 3, 1881, it was resolved that:

The Secretary of the Treasury be, and he is hereby directedto cause a complete set of all the weights and measuresadopted as standards to be delivered to the governor ofeach State in the Union, for the use of agricultural collegesin the States, respectively, which have received a grant oflands from the United States, and also one set of the samefor the use of the Smithsonian Institution.

Then, the Act of 11 July 1890 gave authority tothe Office of Construction of Standard Weightsand Measures (or Office of Standard Weights andMeasures), which had been established in 1836within the Treasury’s Coast and Geodetic Survey:

For construction and verification of standard weights andmeasures, including metric standards, for the custom-houses, and other offices of the United States, and for theseveral States � � �

The Act of 12 July 1894 established standardunits of electrical measure:

Be it enacted � � � , That from and after the passage of thisAct the legal units of electrical measure in the UnitedStates shall be as follows: � � � That it shall be the dutyof the National Academy of Sciences [established in 1863]to prescribe and publish, as soon as possible after the pas-sage of this Act, such specifications of detail as shall benecessary for the practical application of the definitions ofthe ampere and volt hereinbefore given, and such specifica-tions shall be the standard specifications herein mentioned.

Following from a long history of our Nation’sleaders calling for uniformity in science, trace-able at least to the several formal proposals fora Department of Science in the early 1880s, andcoupled with the growing inability of the Officeof Weights and Measures to handle the explosionof arbitrary standards in all aspects of federal andstate activity, it was inevitable that a standards lab-oratory would need to be established. The politi-cal force for this laboratory came in 1900 throughLyman Gage, then Secretary of the Treasury underPresident William McKinley. Gage’s original planwas for the Office of Standard Weights and Mea-sures to be recognized as a separate agency calledthe National Standardizing Bureau. This Bureauwould maintain custody of standards, comparestandards, construct standards, test standards, andresolve problems in connection with standards.Although Congress at that time wrestled with the

level of funding for such a laboratory, the impor-tance of the laboratory was not debated. Finally,the Act of 3 March 1901, also known as theOrganic Act, established the National Bureau ofStandards within the Department of the Treasury,where the Office of Standard Weights and Mea-sures was administratively located:

Be it enacted by the Senate and House of Representatives ofthe United States of America in Congress assembled, That theOffice of Standard Weights and Measures shall hereafter beknown as the National Bureau of Standards � � � That thefunctions of the bureau shall consist in the custody of thestandards; the comparison of the standards used in scien-tific investigations, engineering, manufacturing, commerce,and educational institutions with the standards adopted orrecognized by the Government; the construction, when nec-essary, of standards, their multiples and subdivisions; thetesting and calibration of standard measuring apparatus;the solution of problems which arise in connection withstandards; the determination of physical constants and theproperties of materials, when such data are of great impor-tance to scientific or manufacturing interests and are not tobe obtained of sufficient accuracy elsewhere.

The Act of 14 February 1903, established theDepartment of Commerce and Labor, and in thatAct it was stated that the National Bureau of Stan-dards be moved from the Department of the Trea-sury to the Department of Commerce and Labor.Then, in 1913, when the Department of Labor wasestablished as a separate entity, the Bureau wasformally housed in the Department of Commerce.

In the post World War I years, the Bureau’sresearch focused on assisting in the growth ofindustry. Research was conducted on ways toincrease the operating efficiency of automobileand aircraft engines, electrical batteries, and gasappliances. Also, work was begun on improv-ing methods for measuring electrical losses inresponse to public utility needs. This latterresearch was not independent of internationalefforts to establish electrical standards similar tothose established over 50 years earlier for weightsand measures.

After World War II, significant attention andresources were devoted to the activities of theBureau. In particular, the Act of 21 July 1950established standards for electrical and photomet-ric measurements:

Be it enacted by the Senate and House of Representatives ofthe United States of America in Congress assembled, Thatfrom and after the date this Act is approved, the legal unitsof electrical and photometric measurements in the United


States of America shall be those defined and established asprovided in the following sections. � � � The unit of electri-cal resistance shall be the ohm � � � The unit of electricalcurrent shall be the ampere � � � The unit of electromotiveforce and of electrical potential shall be the volt � � � Theunit of electrical quantity shall be the coulomb � � � Theunit of electrical capacity shall be the farad � � � The unitof electrical inductance shall be the henry � � � The unit ofpower shall be the watt � � � The units of energy shall bethe (a) joule � � � and (b) the kilowatt-hour � � � The unitof intensity shall be the candle � � � The unit of flux lightshall be the lumen � � � It shall be the duty of the Secretaryof Commerce to establish the values of the primary elec-tric and photometric units in absolute measure, and thelegal values for these units shall be those represented by,or derived from, national reference standards maintainedby the Department of Commerce.

Then, as a part of the Act of 20 June 1956, theBureau moved from Washington, DC to Gaithers-burg, Maryland. The responsibilities listed in theAct of 21 July 1950, and many others, weretransferred to the National Institute of Stan-dards and Technology (NIST) when the NationalBureau of Standards was renamed under theguidelines of the Omnibus Trade and Competi-tiveness Act of 1988:

The National Institute of Standards and Technology[shall] enhance the competitiveness of American industrywhile maintaining its traditional function as lead nationallaboratory for providing the measurement, calibrations,and quality assurance techniques which underpin UnitedStates commerce, technological progress, improved prod-uct reliability and manufacturing processes, and publicsafety � � � [and it shall] advance, through cooperativeefforts among industries, universities, and government lab-oratories, promising research and development projects,which can be optimized by the private sector for commer-cial and industrial applications � � � [More specifically, NISTis to] prepare, certify, and sell standard reference materialsfor use in ensuring the accuracy of chemical analyses andmeasurements of physical and other properties of materials� � �

The organizational structure at NIST. NIST’s mis-sion is to promote U.S. economic growth by work-ing with industry to develop and apply technology,measurements, and standards. It carries out thismission through four major programs includingATP, but its centerpiece program is the measure-ment and standards laboratories program. It pro-vides technical leadership for vital components ofthe nation’s technology infrastructure needed byU.S. industry to continually improve its productsand services.

The economics of standards.54 An industry stan-dard is a set of specifications to which all ele-ments of products, processes, formats, or proce-dures under its jurisdiction must conform. Theprocess of standardization is the pursuit of thisconformity, with the objective of increasing theefficiency of economic activity.

The complexity of modern technology, espe-cially its system character, has lead to an increasein the number and variety of standards that affecta single industry or market. Standards affect theR&D, production, and market penetration stagesof economic activity and therefore have a signif-icant collective effect on innovation, productivity,and market structure. Thus, a concern of govern-ment policy is the evolutionary path by which anew technology or, more accurately, certain ele-ments of a new technology become standardized.

Standardization can and does occur withoutformal promulgation as a standard. This distinc-tion between de facto and promulgated standardsis important more from an institutional processthan an economic impact perspective. In onesense, standardization is a form rather than a typeof infrastructure because it represents a codifi-cation of an element of an industry’s technologyor simply information relevant to the conduct ofeconomic activity. And, because the selection ofone of several available forms of a technology ele-ment as the standard has potentially importanteconomic effect, the process of standardization isimportant.

While economics is increasingly concerned withstandards due to their proliferation and perva-siveness in many new high-technology industries,the economic roles of standards are unfortunatelypoorly understood. Standards can be grouped intotwo basic categories: product-element standardsand nonproduct-element standards. This distinc-tion is important because the economic role ofeach type is different.

Product-element standards typically involve oneof the key attributes or elements of a product,as opposed to the entire product. In most cases,market dynamics determine product-element stan-dards. Alternative technologies compete intenselyuntil a dominant version gains sufficient marketshare to become the single de facto standard. Mar-ket control by one firm can truncate this com-petitive process. Conversely, nonproduct-element


standards tend to be competitively neutral withinthe context of an industry. This type of stan-dard can impact an entire industry’s efficiencyand its overall market penetration rate. Industryorganizations often set nonproduct-element stan-dards using consensus processes. The technicalbases (infratechnologies) for these standards havea large public good content. Examples includemeasurement and test methods, interface stan-dards, and standard reference materials.

From both the positions of a strategically-focused firm as well as a public policy maker, stan-dardization is not an all-or-nothing proposition.In complimented system technologies, such as dis-tributed data processing, telecommunications, orfactory automation, standardization typically pro-ceeds in an evolutionary matter in lock step withthe evolution of the technology. Complete stan-dardization too early in the technology’s life cyclecan constrain innovation.

The overall economic value of a standard isdetermined by its functionality (interaction withother standards at the systems level) and the costof implementation (compliance costs). Standardsshould be competitively neutral, which meansadaptable to alternative applications of a generictechnology over that technology’s life cycle.

12. Toward a more integratedentrepreneurial process

An integrated entrepreneurial process is illus-trated in Figure 9. As in the earlier schematicsin Figures 1 and 4, the strategic direction of thefirm and the competitive pressures that it facesmotivate an entrepreneurial response. R&D activ-ity is the primary resource that the firm relies uponto investigate the appropriate response and to actupon it, but Figure 9 includes additions to the ear-lier schematics that reflect the complexity of theentrepreneurial process.

One addition to Figure 9 is the inclusion ofseveral factors that influence the level of in-houseR&D. The first of these is the infrastructure tech-nology that comes from federal laboratories, suchas NIST, or from the environment created by beinglocated in a science park. The second is involve-ment in research partnerships, with other firms orperhaps with either a university or a federal lab-oratory. The third addition is the recognition that

Strategic direction of thefirm

Entrepreneurial response Innovation Value Added

Competitive market conditions

In-house R&D

Research Partnerships

Infrastructure Technology

Science Base

Spillovers to society

External influences oneconomic performance

Figure 9. The entrepreneurial process: an integrated look.

many innovations diffuse into society and generatespillover benefits to other firms in outside indus-tries. Finally, the science base also influences thelevel of R&D activity. The science base conceptu-alizes the stock of knowledge generated from basicresearch. The science base resides in the publicdomain—and the public domain is internationalin scope—generally in the form of scientific jour-nals but also it is in part embodied in universityscientists.

Two internal feedback mechanisms, depicted bydashed lines, are also added to Figure 9. Thefirst feedback in the model flows from innovationto the science base. Once an innovation exists,knowledge has been created and it too will residein the public domain.

The second internal feedback extends frominnovation to competitive market conditions. Itreflects the extent to which innovation can alterthe competitive landscape. This can be seen, forexample, in the evolution of industries, where newtechnologies eclipse old.55 Although profit oppor-tunities create an incentive for innovation, inno-vation can subsequently alter the structure of amarket and in so doing, the profit opportunitiesin that marketplace. Generally speaking, marketstructure and market behavior are jointly endoge-nous. More recent literature on industry structure,applied to issues such as innovation, often usesa game-theory approach to separate out endoge-nous effects from true exogenous conditions.56

Forward-looking entrepreneurial strategy takesseriously the feedback from innovation to mar-ket conditions. Such strategy shapes the natureof rivalry in the innovation process. For example,in one scenario, a significant innovation can cre-ate a monopoly for the successful entrepreneur,


either due to a patent award or from the fail-ure of rivals to quickly imitate. In this case,absent other market failures such as spillovers, theprofit incentive likely leads to intense competitionbetween competing entrepreneurs in their “race”to innovate. It follows that the initial competi-tive landscape ultimately changes to a monopoly.Interestingly, game-theory modeling shows thatexcessive duplicative R&D may occur in this sce-nario. However, as discussed in previous sections,and confirmed by game theoretic models, an oppo-site result can occur if significant spillovers ofknowledge flow from one firm to another. Thatis, strategic R&D competition may yield too lit-tle innovative activity because the flow of knowl-edge to competitors has a negative effect on theentrepreneur’s position.57

The usefulness of Figure 9 is not only as a sum-mary device but also it is a means to highlight themyriad sources of scientific and technical informa-tion that firms rely on to support their innovativeactivity. Certainly, not all firms rely on each sourceto the same degree. Larger firms in competitiveenvironments generally rely more heavily on theirin-house R&D than smaller firms. Small firms relymore on external sources of technical expertise.

Figure 9 is also a useful device for summa-rizing public policies toward R&D. The patentsystem and the R&E tax credit provide directincentives to the firm to increase its level ofR&D. The NCRA and the NCRPA affect the effi-ciency of in-house R&D by reducing duplicativeresearch costs and shortening the fundamentalresearch stage. Government-provided infrastruc-ture technology through standards reduces trans-action costs in the market, thus lowering themarginal cost of R&D. Finally, federal supportof university research continually enriches the sci-ence base thus facilitating the in-house R&Dprocess.

13. Labor market for R&D scientistsand engineers

Key inputs into the R&D process are scientistsand engineers (S&E) working in R&D laborato-ries. Because of the need for these individuals tobe highly trained, the relevant segment of thatworkforce consists of those who have received adoctorate or are studying for it.58

The majority of doctoral scientists trained inthe United States are employed in institutionsof higher education although over time, indus-try is employing more and more scientists. Whileacademe once employed 60 percent of all Ph.D.scientists, this percentage has been below 50 since1990. In contrast, industry, which used to employfewer than one in four scientists, now employsapproximately one in three.

The production or supply of new doctoratesin the U.S. can be summarized in terms of theratio of Ph.D.s granted to U.S. citizens and perma-nent residents to the U.S.-population aged 25–34.The proportion of 25–34 aged individuals receiv-ing a Ph.D. in both the physical and life sciencesincreased throughout the 1960s, declined in the1970s, was fairly stable in the 1980s, and againincreased during the 1990s.

Foreign-born scientists and engineers

Approximately 18 percent of the highly-skilled(doctoral or medical degree) scientists in the U.S.in 1980 were foreign-born.59 The percentage offoreign-born was highest among physical scien-tists (20.4 percent) and lowest among life sci-entists (15.4 percent). By 1990, the proportionforeign-born increased to nearly 25 percent. Morethan one in four physical scientists and mathand computer scientists working in the U.S. wereborn abroad. For life scientists, the proportionhad increased from approximately one in sevento one in five. The proportion of engineers whoare foreign born is substantially smaller than thatof highly-trained (bachelor’s degree) scientists. In1980 approximately 14 percent were foreign-born;this had crept up to about 16 percent by 1990.

A large number of immigrants who receivetheir doctoral training abroad initially come to theUnited States to take postdoctoral positions. Forexample, of the 14,918 postdoctoral appointeestraining in the life sciences in doctorate-traininginstitutions in the United States in 1996, NSF esti-mates that 7,425—almost exactly half—were notU.S. citizens.60 In the physical sciences the propor-tion was even higher, at 57 percent. While someof these are permanent non-residents, a numbercome as temporary residents.


The effect of policy to stimulate R&D on S&Elabor markets

The federal government has several policy mech-anisms to stimulate R&D activity. In industry,research is stimulated through the use of theR&E tax credit, as well as through direct subsi-dies. University research is encouraged throughthe provision of a large funding pool for fac-ulty grants. Depending on the sector targeted, thelabor market consequences of these policies varyconsiderably.

The government’s policy intent to stimulateR&D in the private sector increases the demandfor R&D processes and, by extension, the demandfor highly-trained scientists and engineers. Thisincrease in demand, however, only results in anincrease in the number of S&E workers in indus-try if the supply of S&E workers is not fixed butinstead is responsive to higher wage rates. If thesupply of S&E workers is not wage responsive,then public policies to stimulate R&D will onlyhave a compensation effect rather than an employ-ment effect.

The government policy to support R&D in theuniversity sector, however, can affect S&E labormarkets quite differently. This is because gradu-ate students are especially responsive to the leveland availability of research support, and a signif-icant portion of the R&D funding that supportsuniversity research provides support for graduateresearch assistants.

There are a number of barriers to attractingS&E doctorates to industry:61

• A lack of good information to graduate stu-dents concerning the rewards to working inindustry.

• Attitudes among faculty that employment inindustry makes the student a second-class sci-entist.

• Opportunities for recent Ph.D.s to remain inacademe and work as a post doctorate fellows.

• Funding patterns that tie students and post doc-torates to their mentor’s laboratory, therebydecreasing the opportunity to have differ-ent research experiences and potentially learnmore about positions in industry.

The existence of such barriers suggests thatgovernment policies as they relate to the S&E

workforce are less than efficient. Possible solutionsto these inefficiencies include:

• The provision of information concerning em-ployment opportunities in industry to aspiringstudents as well as to students enrolled in pro-grams.

• The creation of training grants which fund thestudent, instead of the faculty member, therebyproviding students greater independence.

• The reshaping of graduate education along thelines followed by professional schools, includ-ing the provision of information about careerprospects and the provision of opportunities towork in industry while in graduate school.62

Forecasting scientific labor markets

Although economists’ models of scientific labormarkets have been somewhat successful in provid-ing insight into factors affecting demand and sup-ply, reliable forecasts of scientific labor marketsdo not exist, partly because of the unavailabilityof reliable predictions of variables affecting sup-ply and demand.63 While this problem is endemicto forecasting in general, the ups and downs offederal funding make forecasts of scientific labormarkets particularly unreliable.

Despite these problems, forecasts of scientificlabor markets are somewhat common, in partbecause they are mandated by Congress, suppos-edly in an effort to keep the United States’ innova-tion process healthy and its industry competitive.

14. Public accountability64

The concept of public accountability can betraced as far back as President Woodrow Wil-son’s reforms, and in particular to the Budgetand Accounting Act of 1921. This Act of June 10,1921 not only required the President to transmit toCongress a detailed budget on the first day of eachregular session, but also it established the GeneralAccounting Office (GAO) to settle and adjust allaccounts of the government. We note this fiscalaccountability origin because the GAO has had asignificant role in the evolution of accountability-related legislation during the past decade.

What follows is a review the legislative historyof legislation that falls broadly under the rubric ofpublic accountability. As Collins (1997, p. 7) notes:


As public attention has increasingly focused on improv-ing the performance and accountability of Federal pro-grams, bipartisan efforts in Congress and the White Househave produced new legislative mandates for managementreform. These laws and the associated Administration andCongressional policies call for a multifaceted approach—including the provision of better financial and performanceinformation for managers, Congress, and the public andthe adoption of integrated processes for planning, manage-ment, and assessment of results.

Fundamental to any evaluation of a public insti-tution is the recognition that the institution isaccountable to the public, that is to taxpayers,for its activities. With regards to technology-basedinstitutions, this accountability refers to being ableto document and evaluate research performanceusing metrics that are meaningful to the institu-tions’ stakeholders, meaning to the public.

Performance accountability

Chief Financial Officers Act of 1990. The GAOhas a long-standing interest and a well-docu-mented history of efforts to improve governmen-tal agency management through performance mea-surement. For example, in February 1985 theGAO issued a report entitled “Managing the Costof Government—Building An Effective FinancialManagement Structure” which emphasized theimportance of systematically measuring perfor-mance as a key area to ensure a well-developedfinancial management structure.

On November 15, 1990, the 101st Congresspassed the Chief Financial Officers Act of 1990.As stated in the legislation as background for thisAct:

The Federal Government is in great need of fundamentalreform in financial management requirements and practicesas financial management systems are obsolete and ineffi-cient, and do not provide complete, consistent, reliable, andtimely information.

The stated purposes of the Act are to:

1. Bring more effective general and financial man-agement practices to the Federal Governmentthrough statutory provisions which would estab-lish in the Office of Management and Bud-get a Deputy Director for Management, estab-lish an Office of Federal Financial Managementheaded by a Controller, and designate a ChiefFinancial Officer in each executive department

and in each major executive agency in the Fed-eral Government.

2. Provide for improvement, in each agency of theFederal Government, of systems of accounting,financial management, and internal controls toassure the issuance of reliable financial infor-mation and to deter fraud, waste, and abuse ofGovernment resources.

3. Provide for the production of complete, reli-able, timely, and consistent financial informa-tion for use by the executive branch of theGovernment and the Congress in the financing,management, and evaluation of Federal pro-grams.

The key phrase in these stated purposes is inpoint (3) above, “evaluation of Federal programs.”Toward this end, the Act calls for the establish-ment of agency Chief Financial Officers, whereagency is defined to include each of the Fed-eral Departments. And, the agency Chief Finan-cial Officer shall, among other things, “developand maintain an integrated agency accounting andfinancial management system, including financialreporting and internal controls,” which, amongother things, “provides for the systematic mea-surement of performance.”

While the Act does outline the many fiscalresponsibilities of agency Chief Financial Officers,and their associated auditing process, the Act’sonly clarification of “evaluation of Federal pro-grams” is in the above phrase, “systematic mea-surement of performance.” However, neither adefinition of “performance” nor guidance on “sys-tematic measurement” is provided in the Act. Still,these are the seeds for the growth of attention toperformance accountability.

Government Performance and Results Act of 1993.Legislative history is clear that the GovernmentPerformance and Results Act (GPRA) of 1993builds upon the February 1985 GAO reportand the Chief Financial Officers Act of 1990.The 103rd Congress stated in the August 3, 1993legislation that it finds, based on over a year ofcommittee study, that:

1. waste and inefficiency in Federal programsundermine the confidence of the Americanpeople in the Government and reduce the Fed-eral Government’s ability to address adequatelyvital public needs;


2. Federal managers are seriously disadvantagedin their efforts to improve program efficiencyand effectiveness, because of insufficient artic-ulation of program goals and inadequate infor-mation on program performance; and

3. congressional policymaking, spending decisionsand program oversight are seriously handi-capped by insufficient attention to program per-formance and results.

Accordingly, the purposes of GPRA are to:

1. improve the confidence of the American peo-ple in the capability of the Federal Govern-ment, by systematically holding Federal agen-cies accountable for achieving program results;

2. initiate program performance reform with aseries of pilot projects in setting program goals,measuring program performance against thosegoals, and reporting publicly on their progress;

3. improve Federal program effectiveness andpublic accountability by promoting a new focuson results, service quality, and customer satis-faction;

4. help Federal managers improve service deliv-ery, by requiring that they plan for meetingprogram objectives and by providing them withinformation about program results and servicequality;

5. improve congressional decisionmaking by pro-viding more objective information on achievingstatutory objectives, and on the relative effec-tiveness and efficiency of Federal programs andspending; and

6. improve internal management of the FederalGovernment.

The Act requires that the head of each agencysubmit to the Director of the Office of Manage-ment and Budget (OMB):

� � � no later than September 30, 1997 � � � a strategicplan for program activities. Such plan shall contain � � � adescription of the program evaluations used in establishingor revising general goals and objectives, with a schedule forfuture program evaluations.

And, quite appropriately, the Act defines pro-gram evaluation to mean “an assessment, throughobjective measurement and systematic analysis, ofthe manner and extent to which federal programsachieve intended objectives.” In addition, eachagency is required to:

� � � prepare an annual performance plan [beginning withfiscal year 1999] covering each program activity set forthin the budget of such agency. Such plan shall � � � establishperformance indicators to be used in measuring or assess-ing the relevant outputs, service levels, and outcomes ofeach program activity;

where “performance indicator means a particularvalue or characteristic used to measure output oroutcome.”

Cozzens (1995) correctly notes that one fearabout GPRA is that it will encourage agenciesto ignore what is difficult to measure, no mat-ter how relevant. Alternatively, one could weara more pessimistic hat and state that GPRA willencourage agencies to emphasize what is easy tomeasure, no matter how irrelevant.

Fiscal accountability

Legislation following GPRA emphasizes fiscalaccountability more than performance account-ability. While it is not our intent to suggest thatperformance accountability is more or less impor-tant than fiscal accountability, for we believe thatboth aspects of public accountability are impor-tant, the emphasis in the case studies conducted atNIST that are summarized in this paper is on per-formance accountability. Nevertheless, our discus-sion would not be complete in this chapter with-out references to the Government ManagementReform Act of 1994 and the Federal FinancialManagement Improvement Act of 1996.

Government Management Reform Act of 1994.The Government Management Reform Act of1994 builds on the Chief Financial Officers Actof 1990. Its purpose is to improve the manage-ment of the federal government though reforms tothe management of federal human resources andfinancial management. Motivating the Act is thebelief that federal agencies must streamline theiroperations and must rationalize their resources tobetter match a growing demand on their services.Government, like the private sector, must adoptmodern management methods, utilize meaningfulprogram performance measures, increase work-force incentives without sacrificing accountability,and strengthen the overall delivery of services.


Federal Financial Management Improvement Actof 1996. The Federal Financial ManagementImprovement Act of 1996 follows from the beliefthat federal accounting standards have not beenimplemented uniformly through federal agen-cies. Accordingly, this Act establishes a uni-form accounting reporting system in the federalgovernment.

This overview of what we call public account-ability legislation makes clear that governmentagencies are becoming more and more account-able for their fiscal and performance actions. And,these agencies are being required to a greaterdegree than ever before to account for their activ-ities through a process of systematic measure-ment. For technology-based institutions in particu-lar, internal difficulties are arising as organizationslearn about this process.

Compliance with these guidelines is causingincreased planning and impact assessment activ-ity and is also stimulating greater attention tomethodology. Perhaps there is no greater vali-dation of this observation than the diversity ofresponse being seen among public agencies, ingeneral, and technology-based public institutions,in particular, as they grope toward an understand-ing of the process of documenting and assess-ing their public accountability. Activities in recentyears have ranged from interagency discussionmeetings to a reinvention of the assessment wheel,so to speak, in the National Science and Tech-nology Council’s (1996) report, “Assessing Funda-mental Science.”

Systematic approaches to the evaluation oftechnology-based programs

GPRA is directionally, as opposed to methodolog-ically, clear about the evaluation process; pub-lic institutions/research programs will identify out-puts and quantify the economic benefits of theoutcomes associated with such outputs. In ouropinion, agencies should attempt to quantify out-come benefits and then compare those quanti-fied benefits to the public costs to achieve thebenefits. Although these are GPRA’s directions,the methodological hurdle that has been plagu-ing most public agencies is how to quantify ben-efits. And even with an acceptable quantificationof benefits, will the confidence of the American

people in public sector research be strengthenedby simply comparing benefits to costs?

We now consider two different approachesto program evaluation. It appears that the bestevaluation technique for publicly-funded, publicly-performed research is based on a counterfactualmethod. In contrast, we conjecture that the bestevaluation method for publicly-funded, privately-performed research consists of an analysis ofspillovers. These techniques are described in thefollowing sections.

Traditional evaluation methods. Griliches (1958)and Mansfield et al. (1977) pioneered the applica-tion of fundamental economic insight to the devel-opment of measurements of private and socialrates of return to innovative investments. Streamsof investment outlays through time—the costs—generate streams of economic surplus throughtime—the benefits. Once identified and measured,these streams of costs and benefits are used tocalculate rates of return, benefit-to-cost ratios, orother related metrics.

In the Griliches/Mansfield model, the innova-tions evaluated are conceptualized as reducingthe cost of producing a good sold in a com-petitive market at constant unit cost. For anyperiod, there is a demand curve for the good,representing its marginal benefit to consumers,and a horizontal supply curve. Innovation low-ers the unit cost of production, shifting down-ward the horizontal supply curve and thereby,at the new lower equilibrium price, resulting ingreater consumer surplus (economists’ measureof the value of the difference between the priceconsumers are willing to pay and the price theyactually pay, summed over all purchases). Addi-tionally, the Griliches/Mansfield model accountsfor producer surplus, measured as the differencebetween the price the producers receive and theactual marginal cost, summed over the outputsold, minus any fixed costs. Social benefits are thenthe streams of new consumer and producer sur-pluses, while private benefits are the streams ofproducer surplus, not all of which are necessarilynew because the surplus gained by one producermay be cannibalized from the pre-innovation sur-plus of another producer. Social and private costswill, in general, also be divergent.


The Griliches/Mansfield model for calculatingeconomic social rates of return add the publicand the private investments through time to deter-mine social investment costs, and then the streamof new economic surplus generated from thoseinvestments is the benefit. Thus, the evaluationquestion that can be answered from such an eval-uation analysis is: What is the social rate of returnto the innovation, and how does that compare tothe private rate of return? We argue that this isnot the most appropriate question to ask from apublic accountability perspective. The fact that thesocial rate of return is greater than the privaterate of return may validate the role of governmentin innovation if the private sector would not haveundertaken the research; but it ignores, for exam-ple, consideration of the cost effectiveness of thepublic sector undertaking the research as opposedto the private sector.

The counterfactual evaluation method. A dif-ferent question should be considered whenpublicly-funded, publicly-performed investmentsare evaluated. Holding constant the very streamof economic surplus that the Griliches/Mansfieldmodel seeks to measure, and making no attemptto measure that stream, one should ask the coun-terfactual question: What would the private sectorhave had to invest to achieve those benefits in theabsence of the public sector’s investments? Theanswer to this question gives the benefits of thepublic’s investments—namely, the costs avoidedby the private sector. With those benefits—obtained in practice through extensive interviewswith administrators, federal research scientists,and those in the private sector who would haveto duplicate the research in the absence of publicperformance—counterfactual rates of return andbenefit-to-cost ratios can be calculated to answerthe fundamental evaluation question: Are the pub-lic investments a more efficient way of generat-ing the technology than private sector investmentswould have been? The answer to this question ismore in line with the public accountability issuesimplicit in GPRA, and certainly is more in linewith the thinking of public sector stakeholders,who may doubt the appropriateness of govern-ment’s having a role in the innovation process inthe first place.

The spillover evaluation method. There are impor-tant projects where economic performance canbe improved with public funding of privately-performed research. Public funding is neededwhen socially valuable projects would not beundertaken without it. If the expected private rateof return from a research project falls short ofthe required rate called the hurdle rate, then theprivate sector firm will not invest in the project.Nonetheless, if the benefits of the research spillover to consumers and to firms other than theones investing in the research, the social rate ofreturn may exceed the appropriate hurdle rate. Itwould then be socially valuable to have the invest-ments made, but since the private investor willnot make them, the public sector should. By pro-viding some public funding, thereby reducing theinvestment amount needed from the private firmor firms doing the research, the expected privaterate of return can be increased above the hurdlerate. Thus, because of this subsidy, the private firmis willing to perform the research, which is sociallydesirable because much of its output spills over toother firms and sectors in the economy.

The question asked in the spillover method isone that facilitates an economic understanding ofwhether the public sector should be underwriting aportion of private-sector firms’ research, namely:What proportion of the total profit stream gen-erated by the private firm’s R&D and innovationdoes the private firm expect to capture; and hence,what proportion is not appropriated but is insteadcaptured by other firms that imitate the innova-tion or use knowledge generated by the R&D toproduce competing products for the social good?The part of the stream of expected profits cap-tured by the innovator is its private return, whilethe entire stream is the lower bound on the socialrate of return. In essence, this method weighs theprivate return, estimated through extensive inter-views with firms receiving public support abouttheir expectations of future patterns of events andfuture abilities to appropriate R&D-based knowl-edge, against private investments. The social rateof return weights the social returns against thesocial investments.

The application of the spillovers model to theevaluation of public funding/private performanceof research is appropriate since the output of theresearch is only partially appropriable by the pri-vate firm with the rest spilling over to society. The


extent of the spillover of such knowledge with pub-lic good characteristics determines whether or notthe public sector should fund or partially fund theresearch.

Program evaluation

Many public technology-based institutions haveconducted program evaluations both as part oftheir overall management of the program and inresponse to GPRA.

The U.S. General Accounting Office (GAO)monitors the performance evaluation progressof federal government agencies. Based on theirassessment, “agencies’ fiscal year 2000 perfor-mance plans show moderate improvements overthe fiscal year 1999 plans � � � [h]owever, keyweaknesses remain � � � ” (GAO, 1999, p. 3). Oneweakness is that agencies’ performance data lackcredibility.

Not only is GAO expected to continue to mon-itor the performance of research programs, butalso GPRA-like frameworks are beginning to beused at the state level.

15. Conclusions

Our primary purpose in writing this primer wasto survey the landscape of topics that are relatedto the field of economics of science and technol-ogy. We have provided this survey based on histor-ical and applied perspectives, all within the contextof what we call the entrepreneurial process. Webelieve that this material is useful because moststudents and policymakers are not well informedabout the economics of science and technology,despite the fact that the field’s knowledge includesmany issues fundamental to economic growth.

After reading this survey, we hope that readerswill reach many of the same conclusions as we didconcerning the state of the field, its emphases and,just as important, the gaps in knowledge. In ourjudgment, and from the perspective of researchersin this field, the study of the economics of scienceand technology has made important advances inthe past two decades or so. In some instances,the areas of study we examined here were essen-tially unknown as formal research topics just afew short years ago. For example, the study of

patents and intellectual property included rela-tively few important contributions among appliedeconomists until fairly recently. Patent databasesand measurement techniques have greatly facili-tated the field of study, but equally important hasbeen the influence of public policy. With changesin the laws and statutes pertaining to patents, andparticularly the ownership of intellectual propertyamong university and government researchers, thetopic has become manifestly more important dur-ing recent years. Similarly, the study of domestictechnology transfer could barely be considered afield even twenty years ago. But the convergenceof research techniques and policy initiative has ledto a growth of research on this topic.

We believe that there are several issues thatrequire further exploration. The first of these is therelationship between globalization and the devel-opment and ownership of science and technol-ogy. Although there is a long-standing literatureon international technology transfer, the typicalmodel employed in these studies focuses on under-standing relations between a technology or knowl-edge donor nation and a recipient nation. Unfor-tunately, this type of model is not as relevant oruseful for understanding the complexity of the cur-rent state of technology transfer. That is becausethere are now numerous instances where the firms’nation of charter is almost inconsequential, wherecapital flows in a dizzying array of vehicles, insti-tutions and multinational forums, and where tech-nology is simultaneously marketed along differentchannels with different firms in a great number ofcountries. The complexity of this process does notlend itself easily to current models.

Another area that should prove fruitful for stu-dents of the economics of science and technologyis e-commerce. Indeed, there is already an increasein interest in electronic commerce and fundingagencies such as the National Science Founda-tion are girding themselves to support programs ofstudy for this topic. But thus far e-commerce hasnot been a topic of much interest to more thana handful of researchers using the theories andtools of the economics of science and technology.We suspect that this will change in the next fewyears.

A third area that requires more attention isthe intersection between the economics of scienceand technology and distributional and social equity


issues (beyond the question of how new technol-ogy affects the workforce (Siegel, 1999). One ofthe most interesting questions of the economicsof science and technology is who “wins” and who“loses” with innovation and the introduction ofnew technology. We have, for example, begun totalk about the “digital divide,” but there are alsohealth care technology divides, among others. Ifwe understand the economic forces that allow usefficiently to produce and market health care tech-nologies and pharmaceuticals but do not under-standing the distributional issues relating to accesstechnology, it is possible to encourage, at the sametime, more and more innovation and greater divi-sions in society.

These are some of the “big questions.” Unfor-tunately, the field is not yet prepared to providemuch evidence on these questions. But as needschange and resources shift, tomorrow’s economicsof science and technology will almost certainly lookvery different than today’s.

Acknowledgments

We are grateful for comments and suggestions byJohn Jankowski, Jamie Link, and John Scott onearlier versions of this paper. All remaining short-comings are our own.

Notes

1. We do not emphasize the analytical development of theacademic literature herein, but rather we summarize salientconclusions from the literature so as to provide a broad con-text for understanding attendant policies.2. When the innovation is itself the final marketable result,it is sometimes referred to as a product innovation. When theinnovation is applied in a subsequent production process it issometimes referred to as a process innovation (meaning thatits application affects a production process).3. It is not uncommon to see this process referred to as theinnovation process.4. See Hébert and Link (1988) for a complete history of theconcept of the entrepreneur.5. See Cohen and Levinthal (1989) for a detailed develop-ment of this idea.6. This section draws directly from UNESCO (1968) andNational Science Board (2000).7. There is a strong similarity between this early public/private partnership and the establishment of the land-grantcollege system under the Morrill Act in 1862.8. Congress passed the first patent act in 1790.

9. One could make an argument that Jefferson’s funding ofLewis and Clark was the first instance of public support forpure research, whereas Morse was funded to conduct appliedresearch. There are other historical examples of governmentalsupport to individuals for research that has the potential tobenefit society, such as the Longitude Act of 1714. The BritishParliament offered a prize (equal to several million dollars intoday’s terms) for a practicable solution for sailing vessels todetermine longitude (Sobel, 1995).10. This presumption has been the genesis for the formationof many science parks around the world. See Link (1995).11. An excellent history of the growth of U.S. industrialresearch organizations is in Hounshell (1996).12. The Allison Comission failed in 1884 to formulate aninfrastructure to undertake this task. See above.13. Historical data on the number of industrial research labo-ratories is less than precise. Crow and Bozeman (1998) report,based on secondary sources, that in 1920 there were about 500research laboratories, and just over 1,000 in 1930. Regard-less of the precise number that existed in 1920, the direc-tion of growth in industrial research laboratories since then isclear.14. The term “basic research” is credited to Vannevar Bush.He proffered the definition: “Basic research is performed with-out thought of practical ends.” “Basic research” thus is equiv-alent to “science.”15. See National Science Board (2000) for background infor-mation.16. See National Science Board (2000).17. Gross Domestic Product (GDP) is the value of all thegoods and services produced in an economy. In order to makecross-country comparisons we normalize GDP by dividing forthe size of the country’s population. This adjustment allows acomparison of the standard of living for the average person.Economic growth is often thought of in terms of increases inGDP per capita over time.18. In a sense, A in Equation (3) is a residual. It captureschanges in output over time that are not explained by changesin inputs.19. See Link (1996b) for a detailed history of this classifica-tion scheme.20. These descriptions come from Science and EngineeringIndicators—2000, and they are published in other NationalScience Foundation documents.21. This model was first formalized by Griliches (1979).22. See Kerr (1994) and Clark (1995).23. Researchers at that time were quick to point out thatthe slowdown in total factor productivity in the early 1970swas preceded by several years of flat and even declining R&Dspending in the economy.24. Lichtenberg and Siegel (1991) reported that the returnsto company-funded R&D remained high during the produc-tivity slowdown of the 1970s.25. While the historical overview presented above indicatesthat the roots of our national science and technology effortspre-date Bush’s report, Science—The Endless Frontier remainsthe frequently heralded origin.26. See Tassey (1999) and National Science Board (2000).27. See Tassey (1997), and Link and Scott (2001) for a com-plete discussion of variations of this graphical model.


28. Kaufer (1989) provides an excellent historical perspectiveon patents throughout the history of modern civilization.29. The first U.S. patent was issued in 1790.30. To be granted a patent, three criteria must hold: utility,novelty, and non-obviousness. Utility means that the inventionshould be useful; novelty means that the invention be new andnot merely a copy or repetition of another invention; and,non-obviousness—the most difficult criterion—means that theinvention is neither suggested by previous work nor totallyanticipated given existing practices.31. This section draws from Bozeman and Link (1984).32. There is a slight distinction between R&D expendituresfrom a NSF-reporting perspective and R&E expenditures froma tax perspective. R&E expenditures are somewhat more nar-rowly defined to include all costs incident to development.R&E does not include ordinary testing or inspection of mate-rials or products for quality control of those for efficiencystudies, etc. R&E, in a sense, is the experimental portion ofR&D. That said, in practice it is often difficult to distinguishone category from the other.33. More recently, Hall and van Reenen (2000) concludefrom their review of the literature that the tax elasticity ofR&D is about unity, meaning that a 1 percent increase in thecredit will increase industry R&D by about 1 percent.34. See Bozeman and Link (1984).35. This historical information draws from http://www.src.org.36. The eleven founding members were Advanced MicroDevices, Control Data Corporation, Digital Equipment Cor-poration, General Instrument, Honeywell, Hewlett-Packard,IBM, Intel, Monolithic Memories, Motorola, National Semi-conductor, and Silicon Systems.37. The declining U.S. position in the semiconductor industrywas well known and in other industries there was widespreadconcern although the empirical evidence about the competi-tive position of the United States in international markets wasincomplete. However, when the U.S. Department of Com-merce (1990) released its 1990 report on emerging technolo-gies, it was apparent to all that the concerns expressed in theearly 1980s were quite valid.38. This purpose is stated as a preamble to the Act.39. As an illustration of the research activity that can success-fully occur through a small, less visible research partnership,consider the Southwest Research Institute Clean Heavy DieselEngine II joint venture, noticed in the Federal Register inearly-1996. The eleven member companies, from six countriesincluding the United States, joined together to solve a com-mon set of technical problems. Diesel engine manufacturerswere having difficulties, on their own, meeting desired emis-sion control levels. The eleven companies were coordinated bySouthwest Research Institute, an independent, non-profit con-tract research organization in San Antonio, Texas, to col-laborate on the reduction of exhaust emissions. The jointresearch was successful, and each member company took withit fundamental process technology to use in their individ-ual manufacturing facilities to meet desired emission con-trol levels. The joint venture was formally disbanded in mid-1999.40. For a review of the academic literature on RJVs seeHagedoorn et al. (2000).

41. See Hall et al. (2001) for a discussion of barriers thatprevent firms from partnering with universities.42. However, the range is large. The number of universitiesin those joint ventures with universities as research partnersranges from one to over 100.43. See Hall et al. (2000).44. See Leyden and Link (1999).45. This section draws from Link (1999).46. For more information about SEMATECH, see www.sematech.org.47. The thirteen charter members of SEMATECH were:Advanced Micro Devices, AT&T, Digital Equipment Cor-poration, Harris Corporation, Hewlett-Packard Company,IBM Corporation, Intel Corporation, LSI Logic Corporation,Micron Technology, Inc., Motorola, Inc., National Semicon-ductor Corporation, Rockwell International Corporation, andTexas Instruments, Inc.48. This section draws from Tibbetts (1999).49. As a set aside program, the SBIR program redirects exist-ing R&D rather than appropriating new monies for R&D.50. See Audretsch et al. (2002).51. Some of these laboratories are government owned butcontractor operated (GOCO) and others are governmentowned and government operated (GOGO).52. Emphasis is placed on direct technology infrastructurebecause a significant amount of indirect technology infrastruc-ture comes from all departments and agencies as technologicalknowledge spillover and is transferred to industry in one formor another.53. This historical overview draws from Link and Scott(1998).54. This section draws from Tassey (2000).55. See Audretsch (1995).56. The market structure and behavior relationship isdescribed in general in Tirole (1988), pp. 1–4, and in terms ofinnovative activity in Kamien and Schwartz (1982), Chapter 6.57. See Tirole (1988), Chapter 10, for analysis of both inno-vation races and spillovers.58. There are certain fields in S&E where doctoral trainingis not a necessary condition for work in R&D. Engineering isa good case in point. Moreover, as innovation becomes moreand more imbedded in non-R&D functions of the firm, thedoctoral trained workforce may become less key to under-standing patterns of innovation.59. If they immigrate prior to receiving their doctoral train-ing, and if they indicate at the time that they receive theirdegree that they plan to remain in the U.S., they arecaptured in the National Science Foundation’s Survey ofDoctorate Recipients (the database from where the abovestatistics came). If, however, they immigrate after receivingtheir doctoral training, they are not included in this databaseand are only captured in a sampling frame created everyten years, based on the decennial census and known as theNational Survey of College Graduates (NSCG). See Stephanand Levin (2000) for a discussion.60. See National Science Foundation (1998).61. See Romer (2000) and Stephan and Levin (2000).62. See Romer (2000).63. The variables usually found to affect the supply ofenrollees (or the number of graduates) in a specific field are


the salary paid in that field, salary in an alternative occupa-tion, such as law or business and (for men) the draft defermentpolicy.64. This section draws directly from Link and Scott (1998).

References

Abramovitz, M., 1956, ‘Resource and Output Trends in theUnited States Since 1870’, American Economic Review46 (2), 5–23.

Acs, Z. and D.B. Audretsch, 1989, ‘Patents as a Measure ofInnovative Activity’, Kyklos 42 (2), 171–180.

Acs, Z. and D.B. Audretsch, 1990, Innovation and Small Firms,Cambridge, MA: MIT Press.

Acs, Z., D.B. Audretsch, and M.P. Feldman, 1992, ‘RealEffects of Academic Research: Comment’, American Eco-nomic Review 82, 363–367.

Acs, Z., D.B. Audretsch, and M.P. Feldman, 1994, ‘R&DSpillovers and Recipient Firm Size’, Review of Economicsand Statistics 100, 336–340.

Almeida, P. and B. Kogut, 1997, ‘The Exploitation of Tech-nological Diversity and the Geographical Localization ofInnovation’, Small Business Economics 9, 21–31.

Association for University Related Research Parks (AURRP),1998, Worldwide Research & Science Park Directory, Wash-ington, DC: Association for University Related ResearchParks.

Audretsch, D.B., 1995, Innovation and Industry Evolution,Cambridge, MA: MIT Press.

Audretsch, D.B. and M.P. Feldman, 1996, ‘R&D Spillovers andthe Geography of Innovation and Production’, AmericanEconomic Review 86 (3), 630–640.

Audretsch, D.B., A.N. Link, and J.T. Scott, 2002, ‘Pub-lic/Private Partnerships: Evaluating SBIR-SupportedResearch’, Research Policy 31 (1), 145–158.

Audretsch, D.B. and P.E. Stephan, 1996, ‘Company-ScientistLocational Links: The Case of Biotechnology’, AmericanEconomic Review 86 (3), 641–652.

Austin, D.H., 1993, ‘An Event Study Approach to MeasuringInnovative Output: The Case of Biotechnology’, AmericanEconomic Review Papers and Proceedings 83 (2), 253–258.

Barnett, H.G., 1953, Innovation: The Basis of Cultural Change,New York: W.W. Norton.

Baudeau, N., 1910 [originally 1767]; in A. Dubois (ed.),Premier Introduction à la Philosophie Économique, Paris:P. Geuthner.

Ben-Zion, U., 1984, ‘The R&D and Investment Decisionand Its Relationship to the Firm’s Market Value: SomePreliminary Results’, in Z. Griliches (ed.), R&D, Patents,and Productivity, Chicago: University of Chicago Press, pp.134–162.

Bound, J., C. Cummins, Z. Griliches, B. Hall, and A. Jaffe,1984, ‘Who Does R&D and Who Patents?’, in Z. Griliches(ed.), R&D, Patents, and Productivity, Chicago: Universityof Chicago Press, pp. 21–54.

Bozeman, B. and A.N. Link, 1983, Investments inTechnology: Corporate Strategies and Public Policy Alterna-tives, New York: Praeger Publishers.

Bozeman, B. and A.N. Link, 1984, ‘Tax Incentives for R&D:A Critical Evaluation’, Research Policy 13, 21–31.

Bozeman, B. and M. Papadakis, 1995, ‘Firms’ Objectivesin Industry-Federal Laboratory Technology DevelopmentPartnerships’, Journal of Technology Transfer 20, 64–74.

Bozeman, B. and D. Wittmer, 2002, ‘Technical Roles and Suc-cess of Federal Laboratory-Industry Partnerships’, Scienceand Public Policy, in press.

Brod, A.C. and A.N. Link, 2001, ‘Trends in CooperativeResearch Activity: Has the National Cooperative ResearchAct Been Successful?’, in M. Feldman and A. Link(eds.), Innovation Policy in the Knowledge-Based Economy,Norwell, MA: Kluwer Academic Publishers, pp. 105–119.

Bush, V., 1945, Science—The Endless Frontier, Washington,DC: U.S. Government Printing Office.

Carr, R.K., 1995, ‘U.S. Federal Laboratories and TechnologyTransfer’, mimeograph.

Clark, B., 1995, Places of Inquiry, Berkeley: University ofCalifornia Press.

Cockburn, I. and Z. Griliches, 1988, ‘Industry Effects andAppropriability Measures in the Stock Market’s Valuationof R&D and Patents’, American Economic Review Papersand Proceedings 78 (2), 419–423.

Cohen, W.M. and D.A. Levinthal, 1989, ‘Innovation andLearning: The Two Faces of R&D’, Economic Journal569–596.

Collins, E., 1997, ‘Performance Reporting in Federal Man-agement Reform’, National Science Foundation SpecialReport, mimeographed.

Council on Competitiveness, 1996, Endless Frontiers, LimitedResources: U.S. R&D Policy for Competitiveness, Washing-ton, DC: Council on Competitiveness.

Cozzens, S.E., 1995, ‘Assessment of Fundamental Science Pro-grams in the Context of the Government Performance andResults Act (GPRA)’, Washington, DC: Critical Technolo-gies Institute.

Crow, M. and B. Bozeman, 1998, Limited by Design: R&DLaboratories in the U.S. National Innovation System, NewYork: Columbia University Press.

Dunne, T., 1994, ‘Plant Age and Technology Usage in U.S.Manufacturing Industries’, Rand Journal of Economics 25(3), 488–499.

Economic Report of the President, 1994, Washington, DC: Gov-ernment Printing Office.

Eden, L., E. Levitas, and R.J. Martinez, 1997, ‘The Production,Transfer and Spillover of Technology: Comparing Largeand Small Multinationals as Technology Producers’, SmallBusiness Economics 9, 53–66.

Feldman, M.P., 1994, ‘Knowledge Complementarity and Inno-vation’, Small Business Economics 6, 363–372.

Glaeser, E., H. Kallal, J. Scheinkman, and A. Shleifer, 1992,‘Growth of Cities’, Journal of Political Economy 100,1126–1152.

Grabowski, H.G., 1968, ‘The Determinants of IndustrialResearch and Development: A Study of the Chemical,Drug, and Petroleum Industries’, Journal of Political Econ-omy 76, 156–159.


Griliches, Z., 1958, ‘Research Costs and Social Returns:Hybrid Corn and Related Innovations’, Journal of Politi-cal Economy 66 (5), 501–522.

Griliches, Z., 1979, ‘Issues in Assessing the Contribution ofR&D to Productivity Growth’, Bell Journal of Economics10, 92–116.

Griliches, Z., 1981, ‘Market Value, R&D, and Patents’, Eco-nomics Letters 7, 183–187.

Griliches, Z. (ed.), 1984, R&D, Patents, and Productivity,National Bureau of Economic Research for the Universityof Chicago Press, Chicago: University of Chicago Press.

Griliches, Z., 1986, ‘Productivity, R&D, and Basic Research atthe Firm Level in the 1970s’, American Economic Review76 (1), 141–154.

Griliches, Z., 1990, ‘Patent Statistics as Economic Indicators:A Survey’, Journal of Economic Literature 28, 1661–1707.

Griliches, Z., 1998, R&D and Productivity: The Economet-ric Evidence, National Bureau of Economic Research forthe University of Chicago Press, Chicago: University ofChicago Press.

Hagedoorn, J., A.N. Link, and N.S. Vonortas, 2000, ‘ResearchPartnerships’, Research Policy 29, 567–586.

Hall, B.H., Z. Griliches, and J.A. Hausman, 1986, ‘Patents andR&D: Is There a Lag?’, International Economic Review 27,265–302.

Hall, B.H., A. Jaffe, and M. Trajtenberg, 2000, ‘Market Valueand Patent Citations: A First Look’, National Bureau ofEconomic Research Working Paper No. 7741.

Hall, B.H., A.N. Link, and J.T. Scott, 2000, ‘Universities asResearch Partners’, NBER Working Paper No. 7643.

Hall, B.H., A.N. Link, and J.T. Scott, 2001, ‘Barriers InhibitingIndustry from Partnering with Universities: Evidence fromthe Advanced Technology Program’, Journal of TechnologyTransfer 26, 87–98.

Hall, B.H. and J. van Reenen, 2000, ‘How Effective Are FiscalIncentives for R&D? A Review of the Evidence’, ResearchPolicy 449–469.

Hébert, R.F. and A.N. Link, 1988, The Entrepreneur: Main-stream Views & Radical Critiques, 2nd ed., New York:Praeger Publishers.

Henderson, R., A. Jaffe, and M. Trajtenberg, 1998, ‘Universi-ties as a Source of Commercial Technology: A DetailedAnalysis of University Patenting, 1965–1988’, Review ofEconomics and Statistics, 80 (1), 119–127.

Hirschey, M., 1982, ‘Intangible Capital Assets of Advertisingand R&D Expenditures’, Journal of Industrial Economics30, 370–390.

Hounshell, D.A., 1996, ‘The Evolution of Industrial Researchin the United States’, in R. Rosenbloom and W. Spencer(eds.), Engines of Innovation: U.S. Industrial Research atthe End of an Era, Boston: Harvard Business School Press,pp. 13–86.

House Committee on Science, 1998, Unlocking Our Future:Toward a New National Science Policy, Report to Congress,September 24.

Jaffe, A.B., 1986, ‘Technological Opportunity and Spillovers ofR&D: Evidence from Firm’s Patents, Profits, and MarketValue’, American Economic Review 76, 984–1001.

Jaffe, A., 1989, ‘Real Effects of Academic Research’, AmericanEconomic Review 79, 957–970.

Jaffe, A.B., M.S. Fogarty, and B.A. Banks, 1998, ‘Evidencefrom Patents and Patent Citations on the Impact of NASAand Other Federal Labs on Commercial Innovation’, Jour-nal of Industrial Economics 46 (2), 183–206.

Jaffe, A., M. Trajtenberg, and R. Henderson, 1993, ‘Geograph-ical Localization of Knowledge Spillovers as Evidencedby Patent Citations’, Quarterly Journal of Economics 63,577–598.

Kamien, M.I. and N.L. Schwartz, 1982, Market Structure andInnovation, Cambridge, UK: Cambridge University Press.

Kaufer, E., 1989, The Economics of the Patent System, Chur,UK: Harwood Academic Publishers.

Kerr, C., 1994, Troubled Times in American Higher Education,Albany: State University of New York Press.

Leyden, D.P. and A.N. Link, 1999, ‘Federal Laboratories asResearch Partnerships’, International Journal of IndustrialOrganization 17, 575–592.

Lichtenberg, F.R. and D. Siegel, 1991 ‘The Impact of R&DInvestment of Productivity—New Evidence Using LinkedR&D-LRD Data’, Economic Inquiry 29 (2), 203–228.

Link, A.N., 1981, ‘Basic Research and Productivity Increase inManufacturing’, American Economic Review 71 (5), 1111–1112.

Link, A.N., 1983, ‘Market Structure and Voluntary ProductStandards’, Applied Economics 15, 393–401.

Link, A.N., 1990, ‘Firm Size, University-Based Research, andthe Returns to R&D’, Small Business Economics 2, 25–32.

Link, A.N., 1995, A Generosity of Spirit: The Early Historyof Research Triangle Park, Research Triangle Park, NC:Research Triangle Foundation of North Carolina.

Link, A.N., 1996a. Evaluating Public Sector Research and Devel-opment, New York: Praeger Publishers.

Link, A.N., 1996b, ‘On the Classification of Industrial R&D’,Research Policy 25, 397–401.

Link, A.N., 1999, ‘Public/Private Partnerships in the UnitedStates’, Industry and Innovation 6, 191–217.

Link, A.N., 2000, ‘An Assessment of the SBIR Fast TrackProgram in Southeastern United States’, in C. Wessner,(ed.), The Small Business Innovation Research Program:An Assessment of Fast Track, Washington, DC: NationalAcademy of Sciences.

Link, A.N. and J. Rees, 1990, ‘Firm Size, University BasedResearch and the Returns to R&D’, Small Business Eco-nomics 2, 25–32.

Link, A.N. and J.T. Scott, 1998, Public Accountability: Evalu-ating Technology-Based Institutions, Norwell, MA: KluwerAcademic Publishers.

Link, A.N. and J.T. Scott, 2000, ‘Evaluating Federal ResearchPrograms’, mimeographed.

Link, A.N. and J.T. Scott, 2001, ‘Public/Private Partnerships:Stimulating Competition in a Dynamic Economy’, Interna-tional Journal of Industrial Organization 19 (5), 763–794.

Link, A.N. and G. Tassey, 1987, Strategies for Technology-basedCompetition: Meeting the New Global Challenge, Lexington,MA: Lexington Books.

Mansfield, E., 1968, Industrial Research and TechnologicalChange, New York: W.W. Norton.


Mansfield, E., 1971, Technological Change, New York: W.W.Norton.

Mansfield, E., 1980, ‘Basic Research and Productivity Increasein Manufacturing’, American Economic Review 70 (5), 863–873.

Mansfield, E., J. Rapoport, A. Romeo, S. Wagner, andG. Beardsley, 1977, ‘Social and Private Rates of Returnfrom Industrial Innovations’, Quarterly Journal of Eco-nomics 91 (2), 221–240.

Morgan, R.P., 1998, ‘University Research Contributions toIndustry: The Faculty View’, in P. Blain and R. Frosch(eds.), Trends in Industrial Innovation: Industry Perspectives& Policy Implications, Research Triangle Park: Sigma Xi,The Scientific Research Society, pp. 163–170.

Morrison, C. and D. Siegel, 1997, ‘External Capital Factorsand Increasing Returns in U.S. Manufacturing’, Review ofEconomics and Statistics 79 (4), 647–654.

Mowery, D.C. and D.J. Teece, 1996, ‘Strategic Alliances andIndustrial Research’, in R. Rosenbloom and W. Spencer(eds.), Engines of Innovation: U.S. Industrial Research atthe End of an Era, Boston: Harvard Business School Press,pp. 111–129.

Mueller, D.C., 1967, ‘The Firm Decision Process: An Econo-metric Investigation’, Journal of Political Economy 81,58–87.

National Resources Committee, 1938, Research—A NationalResource, Washington, DC: Science Committee of theNational Resources Committee.

National Science and Technology Council, 1996, Assessing Fun-damental Science, Washington, DC: National Science andTechnology Council.

National Science Board, 2000, Science and EngineeringIndicators—2000, Washington, DC: National Science Foun-dation.

National Science Foundation, 1998, Graduate Studies and Post-doctorates in Science and Engineering: 1996, NSF 98–301,Washington, DC: National Science Foundation.

Nelson, R.R. and E.S. Phelps, 1966. ‘Investment in Humans,Technological Diffusion, and Economic Growth’, AmericanEconomic Review 56, 69–75.

Nightingale, P., 1998, ‘A Cognitive Model of Innovation’,Research Policy 27, 689–709.

Office of the President, 1990, U.S. Technology Policy,Washington, DC: Executive Office of the President.

Office of Technology Policy, 1996, Effective Partnering: AReport to Congress on Federal Technology Partnerships,Washington, DC: U.S. Department of Commerce.

Olson, K., 1998, ‘Total Science and Engineering Gradu-ate Enrollment Falls for Fourth Consecutive Year’, DataBrief, Arlington, VA: National Science Foundation, Decem-ber 17.

Pakes, A., 1985, ‘On Patents, R&D, and the Stock MarketRate of Return’, Journal of Political Economy 93 (2), 390–409.

Pakes, A., 1986, ‘Patents as Options: Some Estimates of theValue of Holding European Patent Stocks’, Econometrica54 (4), 755–784.

Pakes, A. and Z. Griliches, 1980, ‘Patents and R&D at theFirm Level: A First Report’, Economics Letters 5, 377–381.

Pakes, A. and Z. Griliches, 1984, ‘Patents and R&D at theFirm Level: A First Look’, in Z. Griliches (ed.), R&D,Patents, and Productivity, Chicago: University of ChicagoPress, pp. 55–72.

Pakes, A. and M. Schankerman, 1984, ‘The Rate of Obsoles-cence of Patents, Research Gestation Lags, and the PrivateRate of Return to Research Outcomes’, in Zvi Griliches,(ed.), R&D, Patents, and Productivity, National Bureau ofEconomic Research for the University of Chicago Press,Chicago: University of Chicago Press, pp. 73–88.

Paul, C.J. Morrison and D.S. Siegel, 2001, ‘The Impacts ofTechnology, Trade, and Outsourcing on Employment andLabor Composition’, Scandinavian Journal of Economics103 (2), 241–264.

Prevezer, M., 1997, ‘The Dynamics of Industrial Clustering inBiotechnology,’ Small Business Economics 9, 255–271.

Reis, P. and D.H. Thurgood, 1991, Summary Report 1991: Doc-toral Recipients from United States Universities, WashingtonDC: National Academy Press.

Roessner, J.D. and A. Bean, 1991, ‘How Industry Interactswith Federal Laboratories’, Research and Technology Man-agement 21, 22–26.

Rogers, J. and B. Bozeman, 1997, ‘Basic Research and Tech-nology Transfer in Federal Laboratories’, Journal of Tech-nology Transfer 22, 37–48.

Romeo, A., 1975, ‘Interindustry and Interfirm Differences inthe Rate of Diffusion of an Innovation’, Review of Eco-nomics and Statistics 57, 311–319.

Romer, P.M., 1986, ‘Increasing Returns and Long-RunGrowth,’ Journal of Political Economy 94 (5), 1002–1037.

Romer, P.M., 1994, ‘The Origins of Endogenous Growth,’ Jour-nal of Economic Perspectives 8, 3–22.

Romer, P.M., 2000, ‘Should the Government Subsidize Supplyor Demand in the Market of Scientists and Engineers?’,NBER Working Paper No. 7723.

Saxenian, A., 1990, ‘Regional Networks and the Resurgence ofSilicon Valley’, California Management Review 33 89–111.

Schankerman, M. and A. Pakes, 1985, ‘The Rate ofObsolescence and the Distribution of Patent Values:Some Evidence from European Patent Renewals’, RevueEconomique 36 (5), 917–941.

Schankerman, M. and A. Pakes, 1986, ‘Estimates of the Valueof Patent Rights in European Countries During the Post-1950 Period’, Economic Journal 96, 1052–1076.

Scherer, F.M., 1965, ‘Firm Size, Market Structure, Opportunityand the Output of Patented Inventions’, American Eco-nomic Review 55, 1097–1125.

Scherer, F.M., 1983, ‘The Propensity to Patent’, InternationalJournal of Industrial Organization 1, 107–128.

Schmookler, J., 1966, Invention and Economic Growth, Cam-bridge, MA: Harvard University Press.

Schwalbach, J. and K.F. Zimmermann, 1991, ‘A Poisson Modelof Patenting and Firm Structure in Germany’, in Z. Acsand D.B. Audretsch (eds.), in Innovation and TechnologicalChange: An International Comparison, Ann Arbor: Univer-sity of Michigan Press, pp. 109–120.

Siegel, D., 1994, ‘Errors in Output Deflators Revisited: UnitValues and the PPI’, Economic Inquiry, 32 (1), 11–32.


Siegel, D., 1997, ‘The Impact of Computers on Manufactur-ing Productivity Growth: A Multiple-Indicators, Multiple-Causes Approach’, Review of Economics and Statistics 79(1), 68–78.

Siegel, D.S., 1999, Skill-Biased Technological Change: Evidencefrom A Firm-Level Survey, Kalamazoo, MI: W.E. UpjohnInstitute Press.

Siegel, D. and Z. Griliches, 1992, ‘Purchased Services, Out-sourcing, Computers, and Productivity in Manufacturing’,in Output Measurement in the Service Sector, Chicago: Uni-versity of Chicago Press, pp. 429–458.

Siegel, D., J.G. Thursby, M.C. Thursby, and A.A. Ziedonis,2001, ‘Organizational Issues in University-Industry Tech-nology Transfer: An Overview of the Symposium Issue’,Journal of Technology Transfer 26 (1–2), 5–11.

Siegel, D., D. Waldman, and A.N. Link, 1999, ‘Assessing theImpact of Organizational Practices on the Productivity ofUniversity Technology Transfer Offices: An ExploratoryStudy’, NBER Working Paper No. 7256 and Research Pol-icy, in press.

Siegel, D., D. Waldman, L. Atwater, and A.N. Link, 2000,‘Transferring Knowledge from Academicians to Prac-titioners: Qualitative Evidence from University Tech-nology Transfer Offices’, mimeograph, University ofNottingham.

Siegel, D., P. Westhead, and M. Wright, 2002, ‘Assessing theImpact of Science Parks on the Research Productivity ofFirms: Evidence from the U.K.’, Small Business Economics,in press.

Sobel, D., 1995, Longitude, New York: Penguin Books.Solow, R.M., 1957, ‘Technical Change and the Aggregate Pro-

duction Function’, Review of Economics and Statistics, 39,312–320.

Steelman, J.R., 1947, Science and Public Policy: A Program forthe Nation. Washington, DC: The U.S. Presidents ScientificResearch Board, August 27.

Stephan, P., 1996, ‘The Economics of Science’, Journal of Eco-nomic Literature 34, 1199–1235.

Stephan, P. and S.G. Levin, 2002, ‘Exceptional Contribu-tions to U.S. Science by the Foreign-Born and Foreign-Educated’, Population Policy and Research Review, inpress.

Tassey, G., 1999, ‘R&D Trends in the U.S. Economy: Strate-gies and Policy Implications’, Gaithersburg, MD: NationalInstitute of Standards and Technology.

Tassey, G., 1997, The Economics of R&D Policy, New York:Quorum Books.

Tassey, G., 2000, ‘Standardization in Technology-Based Mar-kets’, Research Policy 29, 587–602.

Terleckyj, N.E., 1974, Effects of R&D on the ProductivityGrowth of Industries: An Explanatory Study, Washington,DC: National Planning Association.

Tibbetts, R., 1999, ‘The Small Business Innovation ResearchProgram and NSF SBIR Commercialization Results’,mimeograph.

Tirole, J., 1988, The Theory of Industrial Organization, Cam-bridge, MA: The MIT Press.

Trajtenberg, M., 1990a, ‘A Penny for Your Quotes: Patent Cita-tions and the Value of Innovations’, Rand Journal of Eco-nomics 21 (1), 172–187.

Trajtenberg, M., 1990b, Economic Analysis of Product Innova-tion: The Case of CT Scanners, Cambridge, MA: HarvardUniversity Press.

UNESCO, 1968, National Science Policies of the U.S.A.: Ori-gins, Development and Present Status, Paris: Unesco.

U.S. Congress, Office of Technology Assessment, 1996, TheEffectiveness of Research and Experimentation Tax Credits,Washington, DC: Office of Technology Assessment.

U.S. Department of Commerce, 1990, Emerging Technolo-gies: A Survey of Technical and Economic Opportunities,Washington, DC: Technology Administration.

U.S. General Accounting Office (GAO), 1999, Managingfor Results: Opportunities for Continued Improvements inAgencies’ Performance Plans, Washington, DC: GeneralAccounting Office.

The Economics of Science and...

Documents

Transcript of The Economics of Science and...