Innovation Measurement - Tracking the State of Innovation...

A Report to the Secretary of Commerce by

The Advisory Committee onMeasuring Innovation in the

21st Century Economy

January 2008

Innovation MeasurementTracking the State of Innovation in the American Economy

The Advisory Committee on Measuring Innovationin the 21st Century Economy

http://www.innovationmetrics.gov/

Covers 1 and 4.qxd 1/7/08 1:39 PM




January 2008


BUSINESS MEMBERS

Steve Ballmer CEOMicrosoft Corporation

David L. Bernd CEOSentara Healthcare

James Blanchard Retired Chairman of the Board and CEOSynovus Financial Corp.

George Buckley CEO3M

Art CollinsFormer Chairman and CEOMedtronic

Michael EskewChairman and CEOUPS

Luther Hodges, Jr. PresidentPhoenix Associates, Inc.

John MenzerVice-ChairmanWal-Mart Stores

Samuel J. PalmisanoChairman, President and CEO IBM Corporation

ACADEMIC MEMBERS

Ashish AroraProfessor of Economics and Public PolicyH. John Heinz III School of Public Policy

and ManagementCarnegie Mellon University

Rajesh ChandyJames D. Watkins Professor of MarketingCarlson School of ManagementUniversity of Minnesota

Kathleen B. Cooper Senior FellowTower Center for Political StudiesSouthern Methodist University

Dale W. Jorgenson Professor of EconomicsHarvard University

Donald Siegel Professor and Associate DeanGraduate School of ManagementUniversity of California at Riverside

Innovation Measurement: Tracking the State of Innovation in the American Economy

ADVISORY COMMITTEE MEMBERS

CHAIRCarl Schramm

President and CEOEwing Marion Kauffman Foundation


Acknowledgements

The members of the Advisory Committee would like to acknowledge the contributions of those whohave assisted us in our task.

We are very appreciative of the assistance we have received from U.S. Department of Commercepersonnel. We are particularly grateful for the work of Patricia Buckley, the Executive Director of theAdvisory Committee and a Senior Economic Advisor to Secretary Gutierrez, and E. R. Anderson,the Federal Designated Officer for the Advisory Committee and Deputy Under Secretary forEconomic Affairs.

We have benefited greatly from the guidance of Cynthia A. Glassman, Under Secretary for EconomicAffairs, and appreciate the assistance we have received from her staff at the Department as we haveprepared this report.*

We are also grateful for the support of the Department’s statistical agencies: the Census Bureau andthe Bureau of Economic Analysis.

In addition, we wish to thank all those who submitted comments to us in response to our FederalRegister request. All of the submissions helped us greatly in thinking through our ideas andformulating our recommendations.

* Particular thanks go to: Jane W. Molloy, David N. Beede, Sabrina L. Montes, Joseph V. Kennedy, Cassandra A. Ingram,Beethika S. Khan, Daniel D. Bachman, David K. Henry, Kemble Stokes, Jacque Mason, and Dar Davis.


Transmittal Letterfrom the Committee

January 2008

The Honorable Carlos M. GutierrezSecretary of CommerceU.S. Department of CommerceWashington, DC 20230

Dear Mr. Secretary:

You charged this Committee with developing “new and improved measures of innovation” in three areas: howinnovation occurs in different sectors of the economy, how it is diffused across the economy, and how it affectseconomic growth. As chair of the Advisory Committee on Measuring Innovation in the 21st CenturyEconomy, I am pleased to present a report that is the culmination of nearly a year’s worth of study andconsideration by the members, and that we believe represents the most fundamental changes that can be madeto advance our understanding of innovation.

While we recognize that the American economy is changing in fundamental ways—and that most of thischange relates directly to innovation—our understanding remains incomplete. Indeed, data collection andmeasurement, while seemingly mundane, loom large in understanding these changes. Policymakers, investors,executives, managers, consumers, and researchers require accurate and complete information in order to makeinformed decisions. The centrality of the need to advance innovation measurement cannot be understated.

The difficult work of improving our measurement systems is only just beginning. On behalf of the Committee,I want to thank you for this opportunity, and I look forward to the improved information that will becomeavailable if the Committee’s recommendations are implemented.

Sincerely yours,

Carl J. SchrammChairAdvisory Committee on Measuring Innovation in the 21st Century Economy


Table of Contents

EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

A Message from the Chair: WHY MEASURING INNOVATION MATTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix

Chapter One:THE CURRENT STATE OF INNOVATION MEASUREMENT AND THE ESTABLISHMENT OF THE ADVISORY COMMITTEE . . . . . . . . . . . . . . . . . . . . 1

Chapter Two:GUIDING PRINCIPLES FOR INNOVATION MEASUREMENT . . . . . . . . . . . . . . . . . . . . . . . 5

Chapter Three:WHAT THE GOVERNMENT SHOULD DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Chapter Four:HOW THE BUSINESS COMMUNITY CAN HELP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Chapter Five:WHERE RESEARCH IS NEEDED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

The Advisory Committee on MeasuringInnovation in the 21st Century Economywas established by the Secretary of

Commerce in September 2006 to recommend waysto improve the measurement of innovation in theeconomy. In its report, the Advisory Committeeoutlines its recommendations to the Secretary ofCommerce for steps to be taken by the government,the business community, and government and privatesector researchers to foster and improve themeasurement of innovation in the economy.

The first act of the Advisory Committee was toestablish a definition of innovation that wouldidentify what should be measured. The definitionadopted by the Committee is:

The design, invention, development and/orimplementation of new or altered products,services, processes, systems, organizationalstructures, or business models for the purpose ofcreating new value for customers and financialreturns for the firm.

The definition recognizes that the innovation to bemeasured is more than simply something new; it hasthe added component of adding value for bothcustomers and firms. The definition also recognizesthat innovation measurement needs to extendbeyond simply measuring inputs. While it isimportant to track inputs to innovation – such asresearch and development spending – that is notenough. Outcomes of innovative activity need to betracked and measured to determine fully the impactof innovation on the economy.

As part of its work, the Committee published aFederal Register request for public comments andreceived many helpful responses. These commentswere extremely useful in informing both the guidingprinciples and the recommendations adopted by

the Committee. The Federal Register request andcopies of all the comments received can be accessedon the Advisory Committee’s web site atwww.innovationmetrics.gov.

The following set of principles was developed by theAdvisory Committee to guide its own work:

� Innovation data collection efforts should buildon the way firms assess the effectiveness of theirinnovative activities.

� When developing better ways to quantifyinnovation in the marketplace, considerationshould be given to measuring the impact oflegislation and regulations on innovation.

� Because of the nature of innovation and, inparticular, the collaborative nature of theinnovative process, there needs to be toleranceof qualitative and subjective measures.

� Innovation measurement should not be static.Measurement is an iterative process that needsto be treated less like a ‘project’ and morelike an ongoing ‘dialogue.’ Learning andimprovement are to be gained from each stageof the process.

� Innovation measures should allow for analysisat the establishment, firm, industry, country,international, and, where possible,regional levels.

� A conservative approach should be taken to anynew data collection effort by recognizingtradeoffs between costs and potential benefitsand considering resource and regulatoryconstraints. The implementation of pilotprojects to gauge the costs and benefits of newdata collection efforts is encouraged. To the


i

Executive SummaryMEASURING INNOVATION AND ITS IMPACT ON THE ECONOMY

extent possible, new innovation measuresshould be able to be ‘back-tested’; that is, ifapplied to historical data, the measures shouldproduce the expected innovation relationship.

The principles guided the work of the AdvisoryCommittee and should also apply to implementationof the Committee’s recommendations.

The bulk of the Advisory Committee’s work wasdevoted to developing recommendations to theSecretary of Commerce for actions to improveinnovation measurement. The recommendationsendorsed by the Committee are summarized brieflybelow and appear in a more comprehensive list at theend of this summary. All of the recommendations aredescribed in more detail in the body of the report.

WHAT THE GOVERNMENTSHOULD DO

To achieve the long-term goal of measuring theimpact of innovation on the economy, the AdvisoryCommittee recommends that the government createa coordinated emphasis on innovation measurement.The effort will require structural refinements ofexisting government data sets, the collection of newand better data, improved linkages among statisticalagency data sets, and expanded datasharing/synchronization authority.

In particular, the Advisory Committee recommendsthat the government:

� Create a stronger framework for identifying andmeasuring innovation in the national economy.

� Better leverage existing data among thestatistical agencies to allow for the consistentestimation of the contributions of innovation inthe gross domestic product (GDP) andproductivity accounts and to develop greaterunderstanding of innovation.

� Increase access to data in order to facilitate morerobust innovation research.

� Convene one or more workshops or forumsunder the auspices of the Secretary ofCommerce to discuss innovation drivers,impediments and enablers.

� Continue participation in the internationaldialogue related to measuring and analyzinginnovation and ensure that U.S. effortsare internationally compatible to theextent possible.

� Consider development of a nationalinnovation index when more work has beendone on both data collection and analysis ofinnovation drivers.

� Support funding necessary to implement theabove recommendations.

Most of the recommendations for government actionbuild on existing programs or activities. Theseinclude the U.S. National Income and ProductAccounts (NIPAs) which constitute the officialframework used by the Commerce Department’sBureau of Economic Analysis (BEA) to estimateoutput, income and expenditure, trade, capitalformation, and wealth in the U.S. economy. TheNIPAs were devised to help policymakers deal withthe severe economic fluctuations produced by theGreat Depression and World War II and have beencontinually refined since then. The NIPAs nowprovide policymakers with an unrivaled ability toadjust policy quickly and appropriately in response toshort-term economic fluctuations. The NIPAs werenot originally designed to measure innovation ordelve into the causes of long-term productivitygrowth. Today, however, when policymakers areincreasingly shifting attention from short-termstabilization to long-term economic growth, changesto the NIPAs need to be made to accommodate thisnew focus.

Refining the framework for measuring theperformance of the national economy is an essentialelement in the government’s program to measureinnovation in the national economy and refine overalleconomic measurement. These improvements willtake time as there is considerable preliminary work tobe done not only by BEA but also by the other


ii

statistical agencies upon which BEA relies for data.The four major elements of this program are:integrating industry-level estimates of total factorproductivity1 with the NIPAs; creating asupplemental innovation account; improving servicesector data; and improving the measurement ofintangibles. Work on each of these four elements hasalready begun, although additional funding will berequired to move forward.

The Advisory Committee also recommends that thestatistical agencies pursue an agenda in support of thedevelopment of linkages between data sets both toimprove data consistency and to provide a richer database for understanding and explaining innovation.Full implementation of such an agenda would requirenew legislation. The statistical agencies are currentlylimited in their authority to share data and that, inturn, affects data consistency.

Linkages between establishment-based data and firm-based data would be particularly useful for trackingand measuring innovation. Most U.S. industrystatistics are estimated using establishment-level dataas the basic ‘building blocks.’ Such statistics havebeen very useful for many purposes, as they combinedata for establishments that do approximately thesame things. However, many firms own or controlmore than one establishment, and thoseestablishments may be located in different geographicareas and may be producing different kinds of goodsor performing different kinds of services, some ofwhich, such as technology licensing transactions, maybe of particular importance to the innovationprocess. Reassembling establishment-level data intofirm-based statistics may lead to better innovationmeasurements. Furthermore, firm- or establishment-level data from one data source can be augmented bymatching them to corresponding data from othersources to obtain a more complete picture ofinnovation. All of these linked data are understoodbest when also matched over time to createlongitudinal data sets. In the longitudinal records,the dynamics of business and innovation beginto emerge.

The Advisory Committee recommends that theSecretary support legislation to enable the statisticalagencies to undertake expanded datasharing/synchronization activities. In particular,amending the law to expand access to IRS data toadditional statistical agencies for the purposes ofreconciling the business lists and designing moreeffective survey business frames would improve ourunderstanding of the U.S. economy.

To encourage more research by non-governmentresearchers, the Advisory Committee recommendsthat the government encourage innovation researchby making public data more transparent through theuse of data-tagging or similar methods of makingdata more user-friendly and by improving access todata through the creation of more public use datafiles. Such efforts are being undertaken currently bysome agencies and the Committee encourages theexpansion of such efforts. In addition, the AdvisoryCommittee also recommends the expansion of non-government researcher access to confidential micro-data, including that on business dynamics, whilemaintaining high standards for confidentiality.

A major issue raised by many of the Committeemembers was the need to examine innovation drivers,impediments, and enablers. Since the issue wasoutside the scope of the Committee’s mandate, it wasnot fully explored. However, given the importance ofthe topic, the Committee calls on the Secretary ofCommerce to convene one or more workshops orforums to examine innovation drivers, impedimentsand enablers.

The Advisory Committee recognizes the importanceof the international dialogue on innovation andrecommends that it be continued and that efforts bemade to ensure that new innovation measures allowfor analysis across countries.

Finally, the Advisory Committee recommendssupport for the additional funding that will benecessary to implement the recommendations.


iii

1 Total factor productivity (TFP) is output per unit of total input(hours worked and use of capital). The growth of TFP in excessof the growth of total inputs is attributable, in part, to innovation.

HOW THE BUSINESSCOMMUNITY CAN HELP

Measuring innovation must be a collaborativeprocess, and there is much that the businesscommunity can do to assist and drive improvementsin innovation measurement. In particular, theAdvisory Committee recommends that the businesscommunity:

� Create, expand and assess firm and industry-level measures of innovation and develop bestpractices for innovation managementand accounting.

� Participate in research activities and, asappropriate, make innovation informationavailable to researchers.

One of the guiding principles endorsed by theAdvisory Committee was that innovation datacollection efforts should build on the way firms assessthe effectiveness of their innovative activities.Individual firms, trade associations and otherorganizations are important partners in developingand testing innovation measures.

WHERE RESEARCH IS NEEDED

While our understanding of innovation has increasedover recent years, much more needs to be learnedabout innovation and its measurement.Government, business, and academic researchersshould undertake research – alone and incollaborative efforts – to understand innovationbetter. In particular, the Committee recommendsexploration of the following research areas:

� Identification and assessment of innovationoutcome measures.

� Identification of gaps in innovation data andhow they might be filled.

� Analysis of relationships between innovationactivities and collaboration, innovationperformance and firm performance.

CONCLUSION

The recommendations, if adopted, will go far insetting this nation on a course toward effectivelymeasuring the impact of innovation on the economy.The work is essential to understanding anddeveloping better policies for innovation. TheCommittee calls on the government, the businesscommunity and researchers to work together toimprove the understanding and measurement ofinnovation.


iv

COMPLETE LIST OFADVISORY COMMITTEERECOMMENDATIONS

WHAT THE GOVERNMENTSHOULD DO

Create a stronger framework for identifying andmeasuring innovation in the national economy.

� Develop annual, industry-level measures of totalfactor productivity by restructuring theNational Income and Product Accounts(NIPAs) to create a more complete andconsistent set of accounts integrated with datagenerated by other statistical agencies to allowfor the consistent estimation of thecontributions of innovation to economicgrowth.

� Create a supplemental innovation account forthe NIPAs in order to expand the categories ofinnovation inputs and allow those inputs to betracked as they flow between industries.

� Improve service sector data and increase surveycoverage to provide the data needed to improveestimates from the integrated GDP/productivityaccounts and supplemental innovation account.

� Improve measurement of intangibles,particularly intellectual property, building onwork currently under way at the NationalScience Foundation. Consider the best way inwhich to collect data on transactions involvingkey intangible assets such as intellectualproperty licensing expenditures and revenues.The Commerce Department should also explorewhether additional identifying informationfrom patent and trademark applicants mightprovide useful data.

Better leverage existing data among thegovernment statistical agencies to allow for theconsistent estimation of the contributions ofinnovation in the GDP and productivity accountsand to develop greater understanding ofinnovation.

� Develop linkages within and between existingdata; for example, develop linkages betweenestablishment-based data sets and firm-baseddata sets to provide both greater consistency inestimations and to provide researchers a broaderrange of innovation data.

� Develop more robust classification methods; forexample, classify firms on the basis of bothdomestic and international activities.

� Seek expanded interagency data sharing/synchronization legislative authority in order toimprove the quality of innovation measureswhile balancing data needs with confidentialityprotection.

Increase access to data in order to facilitate morerobust innovation research.

� Increase the transparency of and access to publicdata by fostering the use of data tagging andsimilar processes.

� Create more public use files in order toencourage more non-governmental research.

� Expand non-government researcher access toconfidential micro-data, including that onbusiness dynamics, while maintaining highstandards for confidentiality.

Convene one or more workshops or forums underthe auspices of the Secretary of Commerce todiscuss innovation drivers, impedimentsand enablers.

Continue participation in the internationaldialogue related to measuring and analyzinginnovation and ensure that U.S. efforts areinternationally compatible to the extent possible.

Consider development of a national innovationindex when more work has been done on bothdata collection and analysis of innovation drivers.

Support funding necessary to implement theabove recommendations.


v

HOW THE BUSINESS COMMUNITYCAN HELP

Create, expand and assess firm and industry-levelmeasures of innovation and develop best practicesfor innovation measurement.

� Institute firm-level measures of innovation totest the correlation of particular measures withknown innovation and to measure innovationin the firm and its impact on firm performance.Possible measures might be based on marketshare or on innovation intensity (e.g., the shareof firm revenue attributable to recentlyintroduced products and services).

� Develop and implement best practices ininnovation management and accounting.

Participate in research activities and, asappropriate, make innovation informationavailable to researchers.

� Participate in collaborative projects as a meansof assembling a broad range of data related toinnovation.

� File public reports in XBRL (ExtensibleBusiness Reporting Language), a data-taggingformat, when it is an option.

WHERE RESEARCH IS NEEDED

Much more needs to be learned about innovationand its measurement. Government, business, andacademic researchers should undertake research –alone and in collaborative efforts – to explore thefollowing research areas:

Identification and assessment of innovationoutcome measures.

� Assessment of the effectiveness of measuresbased on market share as innovation measuresand the feasibility, cost, and burden ofdeveloping such measures.

� Assessment of the feasibility, cost, and burden ofdeveloping measures of innovation intensity,including a review of other countries’ experiencein this area, and consideration of a pilot project.

� Analysis of the qualitative and quantitativeimpacts of specific innovation drivers,impediments and enablers on innovationoutcomes.

Identification of gaps in innovation data and howthey might be filled.

� Identification of new data that would be usefulin measuring innovation.

� Assessment of the feasibility, cost, andburden of collecting data on intellectualproperty transactions.

� Identification of ways to overcome gaps andshortcomings in historical empirical measures ofintangible investments.

Analysis of relationships between innovationactivities and collaboration, innovation perform-ance and firm performance.

� Analysis of the relationship between innovationand occupational employment at firms, usingfirm-level micro-data.

� Evaluation of whether firms with highinnovation intensities perform better thanotherwise similar firms with low intensities.

� Assessment of the effect of collaboration oninnovation outcomes and identification of thekey elements of successful collaborativeactivities.

� Assessment and analysis of cross-nationalinnovation activities of firms.

� Analysis of publicly filed financial and otherdata on firms, particularly as the data becomemore user-friendly, to identify innovativepractices and firms.


vi

� Description and explanation of businessdynamics and, to the extent possible, analysisof their relationship to innovation usinglongitudinal business databases.

� Exploration of the use of different sources ofavailable data to determine whether there arecorrelations between innovative performance offirms in different regulatory environments.


vii

In one of his final poems, W.B. Yeats observed,“Measurement began our might.” Consideringthe role innovation has played in the story of

America’s economic expansion, we might say thatinnovation is our might. As the world’s economiesflourish, it will become increasingly easy for ourfellow citizens and people across the globe alike toforget that this prosperous economic system we allshare—what I have referred to as “entrepreneurialcapitalism,”2—is largely an American invention. Tomake this observation, however, is not to write afinal chapter. Rather, reflecting on this greatachievement—sustained long-term growth and thepolitical stability it has engendered—prompts at leastfour key questions that are central to America’s nexteconomic chapter.

The first is, what is the relationship betweeninnovation and economic growth? Even the mostcasual of observers must notice processes that yieldnot only improvements in the goods and services weencounter in our daily lives but also radically newinventions, unimaginable but a few years or monthsearlier. This is the hallmark of daily economic life.From continuous improvements in computers andservice delivery to the evolution of methods to makethe human gene respond to medical intervention,mankind has come to expect that next year we willenjoy an economy that produces yet deeper meaningto the simple expectation that things will be “better,faster, cheaper.” This is the axiom of innovation,and the vocabulary of growth. If we consider thestate of economic knowledge, we can look back on atremendous accumulated body of insight. Ourdedication to understanding economic behavior andthe workings of markets has led to what should beappreciated as marvelously practical knowledge. Asa result of many years of research, mankind is able tooperate within a global milieu where central bankingsystems, efficient capital markets, and highly

specialized labor markets have produced expandingwelfare for more and more people.

But to look at the formal statistical record of theUnited States economy, the vocabulary of innovationis muted if not silent. To appreciate this lacuna, it isimportant to understand that our system ofmeasurement was conceived largely in response to theGreat Depression. Since then, our statistical agencieshave been remarkably competent at maintaining andrefining public data series. However, the frameworkof our data infrastructure does not support or reflectthe nature of today’s dynamic economy—aneconomy that has undergone profound changes overthe past seventy years. Whatever innovation is (andthe concept, as we shall see throughout this report,presents significant definitional problems), its forcein the economy is captured only indirectly.

In an era where capital and labor seem abundant forat least the foreseeable future, we have come to seeinnovation as the most important avenue to growth.

While obviously important, the nexus betweeninnovation and growth is one of the least understoodareas of economic life. One reason is that we havenot paid sufficient attention to the question ofgrowth itself and what drives economic growth.Prior to the industrial revolution, growth was anunknown dimension of human experience. It isestimated that for millennia, mankind experiencedno growth; centuries came and went with liveschanging little. Even more importantly, global wealthremained stagnant—few people became richer andfew poorer. If there appeared to be an expansion ofwealth in one locale, it usually entailed armedconflict in which static goods or precious metals werewrested away from another family, city, or nation.Since about 1820, America’s particular form ofentrepreneurial capitalism has presented the mostconsistent expansion of wealth experienced in history– a twenty-fold increase in living standards. And, it


ix

A Message from the ChairWHY MEASURING INNOVATION MATTERS

2 Schramm, C. J. 2006. The Entrepreneurial Imperative.Harper Collins.

was the wealth and opportunity in the United Statesthat caused nearly all nations to set aside centrallyplanned socialist economic orders in favor of thedynamic and unpredictable phenomenon of oneform or another of capitalism.3

As suggested, however, economic growth remainssomething of a mystery. The chart that followspresents an illustration of the relationship betweeneconomic growth and economic knowledge. Thetriangle, known for Professor Arnold Harberger, whofirst observed the limits of economic research,circumscribes most of what has occupied economistsover the decades. Studies in monopoly theory, labor,public finance, industrial organization, etc., have ashared goal, namely, advancing the optimizedefficiency of the economy so as to capture a marginalgain in growth. But, the story of the growth thatcounts, the tremendous shift over time of the nation’stotal production, is largely unaccounted for in suchoptimization theories. Decades ago, JosephSchumpeter, the first economist to devote himself toentrepreneurship and innovation, wrote: “Thishistoric and irreversible change in the way of doingthings we call ‘innovation’ and we define: innovationsare changes in production functions which cannot bedecomposed into infinitesimal steps. Add as manymail-coaches as you please; you will never get arailroad by so doing.” Many economists havedocumented that the most important story ofeconomics is that of how innovation came to be andhow it is the hinge that opens the door toeconomic growth.4

The second question that must be addressed inthinking about America’s economic future is, what isinnovation? In previous ages, invention was thoughtto be the key to explaining growth in the American

economy. We thought of ourselves as a nationmarked by our propensity to invent new things andnew processes. Over time, our history suggested tous that this was in our character. Frederick JacksonTurner, one of our most distinguished historians,developed the frontier hypothesis that suggested thatwhile Americans thought that exploring and movingwest was our defining frontier, this was really only ametaphor for the creative pioneering that Americansfeel is one of our distinctive shared attributes.

Today we increasingly think of innovation, ratherthan invention, as a word of our times. We believeour propensity to innovate is central to our characterand our economy. But just what is innovation, andhow is it different from other phenomena that havebeen with us since our founding? Obviously, as wethink of innovation in the context of this Committee,it is not an ephemeral national characteristic orpersonality attribute—it is a defining part of oureconomy. It is the edge where the “new” comes intobeing and is transformed into a concrete reality thatproduces benefits. So, one of the first acts of theCommittee was to develop a definition of innovationthat tied “new” to “impact.” The Committee decidedthat innovation is “the design, invention,development and/or implementation of new oraltered products, services, processes, systems,organizational structures, or business models for thepurpose of creating new value for customers in a waythat improves the financial returns for the firm.”


x

3 Baumol, W. R. Litan, and C. Schramm. 2007. Good Capitalism,Bad Capitalism and the Economics of Growth and Prosperity:Yale, 2007.4 The seminal work on this subject is found in Robert Solow, “AContribution to the Theory of Economic Growth.” QuarterlyJournal of Economics 70:65-94 (1956); and Edward F. Denison,“Sources of Economic Growth in the United States and theAlternatives before Us.” (New York: Committee for EconomicDevelopment, 1962). For a more recent treatment, see Baumol,Landes, & Mokyr, Entrepreneurship and Economic History.Princeton (forthcoming).

Considering how to define innovation prompts thethird question, which is the focus of this report—thatis, how do we measure innovation? For those whoworry about questions related to expanding humanwelfare through economic growth, it may be the mostimportant practical inquiry of our times. This reportsuggests the complexity of this task. It describes theissues surrounding the definition of innovation.And, as might be expected, it points to what amountto proxy measures. As we deliberated, it became clearthat sufficient research does not exist to guide ustoward a single measure. Perhaps this will always bethe case. The very nature of innovation suggests thatit will never yield to a tidy and static metric. Anenormously complex economy produces millions ofways in which “better, faster, cheaper” comes intoplay every single day.

In this regard, the report serves as a pointer,describing the first steps of what will necessarily be avery long, maybe never-ending journey. TheCommittee recommendations start with a suggestionthat better measures of the growth in “total factorproductivity” – the change in productivity left overafter taking account of the growth of capital andlabor – are required to begin to zero in on thecontribution of innovation. In reaching thisconclusion we have benefited from the particularexpertise of Professor Dale Jorgenson, one of ourfellow members. But, total factor productivity is, aswe’ve noted, a surrogate approach. Such a measureonly accounts for a portion, albeit a large portion, ofthe change in the economy’s performance that mightbe said to be innovation. There are other measuresthat add more texture to the task but make anyapproach to an inclusive measure that muchmore difficult.

In fact, after extensive consideration, the Committeehas concluded that as much as an aggregate measureof the entire economy’s innovation might bedesirable, recommending an “innovation index,” asort of all-in-one measure of innovation, would beunwise given the current state of research oninnovation and economic growth. A single indexwould be hard to construct and harder still to defend.Innovation being innovation, it would hardly beestablished before it would have to be changed.Moreover, the economy does not innovate evenly:

there will always be some firms and sectors thatinnovate at a greater pace than others. Creating asingle index that treats innovation as a singlephenomenon might lend itself to policy distortions.It would be used immediately in discussions of whatpolicy steps might be appropriate to stimulateinnovation, and an error in the construction of sucha unitary index could play through to disastrousconsequences.

In the absence of a single indicator, the Committeeproposes that the Department of Commerce takesteps toward improving, integrating, and expandingon its current data collection efforts in the next fewyears, while research is undertaken to better informthe task of measuring innovation. Our statisticalsystem captures two of the major inputs that arelinked to the innovation process: research anddevelopment spending and the number of engineers,scientists, and technicians employed. In addition,data are collected on some other categories ofinvestments, such as expenditures on informationtechnology equipment, which is certainly a factor inexpanding innovation. We also measure the numberof innovations that are protected by newly-issuedpatents each year. But, in many firms and industries,significant amounts of investments in innovation aremade outside of these categories and go consistentlyunmeasured or unconnected by the current statisticalsystem. Indeed, even for the variables we measure,such as research and development, the Committeerecognizes the need to update the way we quantifythese measurements, the frequency of ourmeasurement, and our coverage of younger firms andemerging industries. Further, we must developmore finely calibrated measures of how firms investin the inputs that become innovation, especiallyhuman capital.

In the absence of common public measures of thescope of innovation, many surrogates have sprung upin the private sector. Numerous interest groups,industry associations, and think tanks produceindices of innovation that should be considered aspart of the expanding mosaic of data sources. In fact,one of the innovations resulting from this reportmight be that the government absorbs several of themost statistically valid private sector data series intoits own overall measures of innovation.


xi

The final question is the most important one of all.What can we do to drive the American economy to astate of continuous and sustained growth by makingour country more innovative? Making innovation anexplicit concern of public policy requires that we canmeasure how innovative the economy is, as well asdetermine whether the “state” of innovation isimproving or decaying. It also requires that weunderstand the pathways of innovation lest wedevelop the wrong policy interventions, either toquicken innovation or to shift its course. While it isdifficult to say what policy should be, some things areunderstood and should inform policy considerations.

For example, innovation is meaningless if notconsidered in the larger context of growth. Indeed,growth should be the touchstone of policy, and in thecase of innovation, it is important to understand itsrole. Innovation propels growth by pushing existingcompanies to real cost reductions and new firmstoward growth. The United States remains uniquein its ability to grow companies. Every year roughlythirty-one firms achieve “top line” revenue in excessof $1 billion.5 Each one of the firms that experiencesthis kind of growth takes as its basis some innovationin the nature of its products or in the nature of itsapproach to marketing.

Entrepreneurial activity, both in start-ups and inexisting large-scale companies, is critical to the firm’ssuccess. Every entrepreneur is a party to innovation;in fact it could be said that entrepreneurs are to oureconomy as the marines are to our armed forces.Calculated risk taking, with a goal that is often onlydefined as “let’s win,” is the mindset of entrepreneurs,both those working on their own and those whoshoulder the path-breaking and often risky task ofchanging the culture and identity of mature firms.To make the economy more congenial to innovationis to construct an economic culture that appreciateswhat entrepreneurs do.

As we consider these observations, it is clear that oureconomic ecosystem can be helpful to innovation andits ability to flourish. Policy should first focus onputting in place improvements to our measurement

of innovation and then focus on what we knowalready as key factors for innovation. Withoutdoubt, we must have people who are skilled andtrained to see the opportunities where the “better,faster, cheaper” can be realized. This means in realterms that our schools, community colleges, anduniversity-level institutions must prepare more of ouryoung citizens to be better able to advance technicalinsights such that innovative products and servicescontinue to flow into our economy.

If we are to grow, we need more and more peopleready to take entrepreneurial risk. And, we needthem flowing from institutions attuned to producingparticularly creative people for the new economy.Recent economic and psychological research hasconfirmed what scientists and entrepreneurs haveknown for decades: innovative breakthroughsfrequently come at the estuary region where differentfields, not necessarily related, intersect. This meanswe need much more cross-disciplinary training wherethe edge between fields can be developed for theinnovations that lie within.

In conclusion, the work of the Committee hasbrought together experts from across government,industry, and academia to consider changes that canbe made today to improve our understanding of theAmerican economy and its innovative capacity inthe future. Measuring innovation is central tounderstanding the economy as it evolves andresponds to growing world competition. Indeed, it isbetter to travel an illuminated path toward futureeconomic progress than to stumble in an unlitdirection. Improvements to our measurement ofinnovation will help to ensure continued economicstrength.


xii

5 Thomas, David G. 2006. Blueprint to a Billion: 7 Essentials toAchieve Exponential Growth. John Wiley & Sons

Innovation propels economic growth. Yet themeasurement of innovation, in this country andelsewhere, remains rudimentary. To understand

better the dynamics of economic growth and,hopefully, avoid harmful policies and enactfacilitative policies, we must design improvedmeasures of innovation. With this objective in mind,the Secretary of Commerce, Carlos M. Gutierrez,established the Advisory Committee on MeasuringInnovation in the 21st Century Economy.

In the popular mind, innovation is often equatedwith invention. In some cases, this is correct –Thomas Edison stands out as a prominent example –but invention is simply one type of innovation.Innovation can be an astounding breakthrough (aswas often the case with Edison), a mundane shift inprocess, or a subtle change in culture. Yet each typehelps generate a higher yield on resources, boostingeconomic growth. Examples provided by membersof the Advisory Committee at its first meetingillustrate this variety.

� David Bernd, CEO of Sentara Healthcare, anintegrated health care provider in Virginia,spoke of getting a cold call from a small start-upcompany. The caller told him about aninnovative method of remotely monitoringintensive care beds from a centralized location.Sentara ultimately became the first hospital touse the system, and now remotely monitors101 intensive care beds in five institutions. Thehope was that the new system would improvethe quality of care. At the end of two years,Sentara found that mortality had been reducedby 15 percent, ICU length of stay was reducedby 16 percent, variable costs went down25 percent and retention of registered nurses inthe covered units went up 20 percent.

� James Blanchard, former CEO of Synovus,spoke of the creation of his company’s

subsidiary Total System Services, now one of theworld’s largest processors of credit cards. Itstarted when his small Georgia bank, one of thefirst to issue a credit card, bought somecomputers, wrote a program from scratch andautomated the entire operation. Ten years later,BankAmericard automated and Synovus becamethe first bank to use the process. In 1974,almost as a joke, the company offered to processcards for another bank. The other bankaccepted the offer and that started a process thatcontinues today. In 1983 Synovus spun off theoperation, capitalized it, and did an initialpublic offering (IPO). Today it is listed on theNew York Stock Exchange and processes over400 million credit card accounts all over theworld.

� John Menzer, Vice Chair of Wal-Mart, spoke ofhis company’s process-based gains in the areas ofenvironmental sustainability, prescription drugsand supply chain efficiencies. He considersWal-Mart’s process innovation to be a uniquemarriage of corporate culture, imagination andtechnology. He spoke of setting big stretch goalsfor the company and associates—a 25 percentmore efficient trucking fleet in three years anddouble in ten years, a 20 percent reduction inenergy use in new stores in four years, and a25 percent reduction in solid waste in threeyears. The company has no specific plans inplace to meet the goals but uses the goals toinspire associates to help the company get there,create a business environment that thrives onnew ideas and change, and challenge theirsupplier base to set individual goals. And henoted that all Wal-Mart’s innovation is gearedtoward a better consumer experience and betterreturns for the company.

� Michael Eskew, Chairman and CEO of UPS,spoke about how UPS transformed itself from a


1

Chapter OneThe Current State of Innovation Measurement and the Establishment of the

Advisory Committee

local delivery company, to a ground company,to an air company, to an international company,to a supply chain company and to what is nowlargely a technology company. UPS currentlyuses four innovation processes. For innovationrelated to its core business it relies both on aninternal marketing committee and on externalsources, primarily customers. For ‘non-core’innovation, UPS uses both an internal strategygroup and external groups including universitiesand pre-IPO venture companies funded by aUPS strategic enterprise fund.

These examples underscore the challenges inattempting to measure innovation in the economy.

The State of InnovationMeasurement

Innovation measurement is in its infancy – both hereand around the world. Some data related toinnovation are collected by the U.S. governmentstatistical agencies; however, the data are incompleteand miss substantial sources of innovation in theeconomy. Some innovation data are tracked byprivate sector organizations and firms, but theseefforts are also limited. Recently, however,government agencies, businesses and tradeassociations have been devoting more resourcestoward the development of such data.

Examples of some of the U.S. statistical agencyprograms that either currently collect innovation-related data or could be vehicles for enhancedinnovation data collection and analysis throughlinkages, new data or new analysis includethe following:

� The Commerce Department’s Bureau ofEconomic Analysis produces the U. S. NationalIncome and Product Accounts (NIPAs). Themajor agencies providing data for the NIPAsinclude: (1) the Census Bureau (data frombusiness and population censuses and surveys);(2) the Bureau of Labor Statistics (employmentstatistics, wage and salary data, productivitystatistics, and most underlying price

information); (3) the Internal Revenue Service(tax-based data on individuals and businesses);and (4) the Federal Reserve Board (incomestatements and balance sheets for majorfinancial and non-financial sectors).

� The Bureau of Economic Analysis, with supportfrom the National Science Foundation (NSF),has developed a Research and DevelopmentSatellite Account that estimates the effect ofinvestment in research and development onU.S. economic growth. These experimentalestimates of the effect of intangible assets onthe U.S. GDP show the size of the impact ofresearch and development (R&D) on U.S.economic growth. BEA also collects data oninternational technology licensing expenditures.

� The Census Bureau has data on some measuresof innovative activities, such as the diffusion ofinnovation, human and organizational capital,entrepreneurship and other worker and firmcharacteristics. It is working on improvementsin some areas of importance to innovation,particularly related to the service sector.Detailed information has long been collectedon manufacturing activities; only recently hasemphasis been put on non-manufacturingsectors. The Census Bureau, on behalf of theNational Science Foundation (NSF), conductsan annual firm-level survey of industrial R&Dthat requests data on firms’ R&D investments.Census has also created an IntegratedLongitudinal Business Database (ILBD)that includes businesses with and withoutemployees, permitting tracking of the growthof start-ups and other aspects of businessdynamics.

� The Securities and Exchange Commission(SEC) collects extensive data from publiccompanies. The SEC has embarked on a data-tagging project that will permit analysis of suchfinancial data across public firms.

These examples represent just some of the sourcesof innovation-related data available. While theycurrently provide data on only some pieces of theinnovation puzzle, they provide a trove of possibilities


2

for the development of innovation data inthe future.

Dedicated innovation surveys such as the EuropeanCommunity’s Community Innovation Survey areused by statistical organizations in European Unionand some other countries including Australia andCanada. Relatively new, and only tested amongmanufacturing firms in the U.S., the surveys collectinformation on different varieties of innovations,including ‘new to the firm, new to the industry andnew to the world.’ They also collect extensiveinformation on innovation expenditures (e.g., capitalinvestment, training and marketing costs) and costsof protecting innovation (e.g., patent and copyrightcosts). However, such surveys are very costly andhave encountered both definitional and responserate problems.

The Establishment of theAdvisory Committee

Secretary Gutierrez formed the Advisory Committeeto enlist the help of leading business representativesand academics in formulating a plan for improvinginnovation measurement in this country. One of thefirst steps taken by the Committee was the adoptionof a definition.

For the purposes of its work, the AdvisoryCommittee defined innovation as:

The design, invention, development and/orimplementation of new or altered products,services, processes, systems, organizationalstructures, or business models for the purpose ofcreating new value for customers and financialreturns for the firm.

Early last year, the Advisory Committee published aFederal Register request for public comments oninnovation measurement. The purpose of thenotice was to get broad input into the work ofthe Committee.

In the notice, the Committee asked for commentson improved or new innovation measures. The

Committee sought comments in four categories:(1) improvement of the underlying architecture ofthe U.S. system of national accounts to facilitatedevelopment of an improved and more detailedmeasure of innovation and productivity;(2) identification of appropriate economy-wide andsector-specific statistical series or other indicators;(3) identification of firm-specific data items thatcould enable comparisons and aggregation; and(4) identification of specific ‘holes’ in the currentdata collection system that limit our ability tomeasure innovation.

The Committee received a broad array of thoughtfuland substantive responses that covered the full rangeof categories for which comments were sought. All ofthe comments were considered during theCommittee’s deliberations.

In determining the path on which to direct itsrecommendations, the Advisory Committeedetermined that its recommendations would buildon existing and ongoing work to the extent possible.Where appropriate, new approaches or new datawould be recommended but, in the interests of costand delivery time, the primary focus would be onimproving the scope and robustness of existing work.

The Committee recognized that, at least at this time,there is only limited knowledge about innovationdrivers, impediments and enablers. And there wasawareness of the fact that many businesses do notcurrently collect all the data that might be desired inan ideal world.

Because it is also clear that innovation measurementis still in its infancy, the Committee chose torecommend differing approaches to innovationmeasurement. There is no single innovation measurethat can be recommended at this time, but the dataand analysis to be generated by the improvedinnovation measures recommended by theCommittee should lead to greater understanding ofthe process of innovation and prove valuable topolicymakers in the future.

The next chapters detail the principles andrecommendations endorsed by the AdvisoryCommittee.


3

The Advisory Committee’s recommendationsare intended to improve the measurement ofinnovation and its impact on the U.S.

economy. In particular, the goal is to develop betterestimates of the resources devoted to and valuegenerated by innovation by focusing not only onmeasuring innovation activities and inputs but alsoon innovation results and outputs.

Measuring innovation has not been a major goal ofeconomic data collection by U.S. statistical agencies.Innovation measurement to date has been largelypiecemeal, incomplete, and accidental (i.e., relyingon data sets not designed for the purpose ofinnovation measurement).

Improving our understanding of how much is spenton innovation and how much we benefit frominnovation is crucial to answering several keyquestions: Are we as a society spending too much ortoo little on innovation? Are the places in whichfocus is being given to innovation having results?And what, if anything, can we do to improve ourinnovative activities in the future?

The Committee’s recommendations call on thegovernment, the business community and researchers(in both government and the private sector) to worktogether to further both the understanding and themeasurement of innovation in the economy.

The Advisory Committee recommends that the workbe guided by the following principles.

� Innovation data collection efforts should buildon the way firms assess the effectiveness of theirinnovative activities.

Data collection should be informed by what firms areactually doing rather than being based solely ontheory. Also, to the extent possible, the burden onfirms should be minimized.

� When developing better ways to quantifyinnovation in the marketplace, considerationshould be given to measuring the impact oflegislation and regulations on innovation.

Some regulatory policies (e.g., certain tax andeducation policies) may explicitly supportinnovation. Other policies may have the unintendedconsequence of inhibiting innovation (e.g., overlyrestrictive caps on immigration and some Sarbanes-Oxley rules). Improved data on innovation areessential for assessing the impact that regulatorypolicies have on innovation. Better firm-specific andeconomy-wide data will help reveal the impact – bothpositive and negative – of regulatory policies oninnovation and the environment that producesinnovation.

� Because of the nature of innovation and, inparticular, the collaborative nature of theinnovative process, there needs to be toleranceof qualitative and subjective measures.

Not all measures of innovation may be quantifiable;and progress in developing better quantification ofsome dimensions of innovation may depend onimproved qualitative measures. For example,measuring the resources invested in and the outcomesof collaboration may be very important but also verydifficult, especially if such partnerships are informalor if the benefits are subject to spillovers (i.e., aredifficult to capture fully in a contractualarrangement).

� Innovation measurement should not be static.Measurement is an iterative process that needsto be treated less like a ‘project’ and morelike an ongoing ‘dialogue.’ Learning andimprovement are to be gained from each stageof the process.


5

Chapter TwoGUIDING PRINCIPLES FOR INNOVATION MEASUREMENT

As new innovation data are collected, they should berefined and continually re-evaluated for their cost-effectiveness and ability to push out the frontiers ofknowledge about innovation and its impact on theeconomy. The government needs the aid of theresearcher community in order to do so in a timelymanner. And while timeliness is important inunderstanding innovation, different measures ofinnovation may capture relatively quick-returninnovation while other measures may captureinnovations which require longer periods of time tocause measurable economic change.

� Innovation measures should allow for analysisat the establishment, firm, industry, country,international and, where possible,regional levels.

Improved data on innovation should permitindustry- and sector-specific analysis, recognizingthat innovation manifests itself differently indifferent parts of the economy. In particular,international comparisons would help explain whydifferent countries are experiencing different

economic growth rates. When compared withdifferent national policy mixes, it may be possible toachieve a better understanding of the impact publicpolicies have on growth and innovation.

� A conservative approach should be taken to anynew data collection efforts by recognizingtradeoffs between costs and potential benefitsand considering resource and regulatoryconstraints. The implementation of pilotprojects to gauge the costs and benefits of newdata collection efforts is encouraged. To theextent possible, new innovation measures shouldbe able to be ‘back-tested’; that is, if applied tohistorical data, the measures should produce theexpected innovation relationship.

The costs of new data collection include both directprogram costs and the cost burden imposed onpotential survey respondents. Many conceptual ideasfor new data collection need to be tested on diverseindustry samples of large and small or young and oldfirms before new surveys (or substantial changes toexisting surveys) are implemented.


6

To measure innovation and its economicimpact better, the government must makestructural refinements to existing

government data sets, improve linkages among them,collect new and better data, and expand datasharing/synchronization.

The Advisory Committee’s recommendations forgovernment action and the elements within eachrecommendation are discussed in this chapter.

1. THE GOVERNMENT SHOULD CREATE ASTRONGER FRAMEWORK FOR IDENTIFY-ING AND MEASURING INNOVATION IN THENATIONAL ECONOMY.

Refining the framework for measuring theperformance of the national economy is an essentialelement in the government’s program to measureinnovation. These improvements will take time, asthere is considerable preliminary work to be done.The four major elements of this program include:refining the National Income and Product Accounts(NIPAs) to permit estimation of industry-levelmeasures of total factor productivity; creating aninnovation supplementary account; improvingservice sector data; and improving the measurementof intangibles (particularly intellectual property andimproved measures of technology licensing activity).Each of these elements is discussed in more detailbelow.

� Develop annual, industry-level measures of totalfactor productivity by restructuring the NIPAsto create a more complete and consistent set ofaccounts integrated with data generated byother statistical agencies to allow for theconsistent estimation of the contribution ofinnovation to economic growth.

The NIPAs constitute the official framework used bythe Commerce Department’s Bureau of Economic

Analysis (BEA) to estimate output, income andexpenditure, trade, capital formation, and wealth inthe U.S. economy. The NIPAs were devised to helppolicymakers deal with the severe economicfluctuations produced by the Great Depression andWorld War II. The NIPAs were not originallydesigned to measure innovation or delve into thecauses of long-term productivity growth. Today,however, when policymakers are increasingly shiftingattention from short-term stabilization to long-termeconomic growth, changes need to be made tomeasure innovation in this context.

Economic growth depends – in large measure – onproductivity growth. Innovation is a majordeterminant of productivity growth. Withoutinnovation, output generally can only grow byincreasing inputs – through a combination ofexpanding the labor force and increasing theutilization of capital (such as machines, buildings,and skills, using existing technologies). Theimplementation by entrepreneurs, managers, andemployees of innovations – in the form of newproducts or services, processes, organizationalstructures, or business models – enables outputgrowth to exceed the growth of inputs. Measuringthat excess growth – known as Total FactorProductivity (TFP) growth – for the entire privatesector economy as well as by industry is the primarygoal of this proposal. A longer-term goal, discussedlater in this report, is to parse TFP growth estimatesinto amounts attributable to expenditures by firmson scientific R&D and information technologies(using available data), and other measurableinvestments by firms in innovation (using data theCommittee proposes to be collected), and amountsdue to other factors.

The Committee recommends an ambitious programto eliminate gaps and inconsistencies within theNIPAs and between data produced by other statisticalagencies and those produced by BEA. The proposed


7

Chapter ThreeWHAT GOVERNMENT SHOULD DO

new ‘architecture’ for the NIPAs would consist of aset of income statements, balance sheets, flow offunds statements, and productivity estimates for theentire economy and by sector that are more accurateand internally consistent. The new architecture willmake the NIPAs much more relevant to today’stechnology-driven and globalizing economy and willfacilitate the publication of much more detailed andreliable estimates of innovation’s contribution toproductivity growth.

The most salient (for purposes of this Committee)goal of the proposal is to integrate BLS’s current TFP(also known as ‘multifactor productivity’) growthestimation program with the NIPAs and extend theircoverage to all industries. Doing so would requirecompletion of BEA’s plan to improve its industry-level statistics, including fully integrating its input-output tables with the NIPAs and incorporating newdata collected by the Census Bureau on the outputand intermediate inputs of services industries, asdescribed later in this report. Greater coordinationbetween BEA and BLS will be required in other areasas well, including increased data sharing between thetwo agencies; acceleration of the development of thecapital services accounts; and expansion of efforts toincorporate quality adjustments and identify newproducts and services for prompt inclusion in BLS’sConsumer, Wholesale, and International Price Indexprograms. In addition, BEA will have to work withthe Federal Reserve Board (FRB) to develop anintegrated wealth account.

The new NIPA architecture will be flexible enough toaccommodate the inclusion of information from newsupplemental accounts and to permit the easyincorporation of these accounts into the proposedNIPA system. Data developed for capital stocks andservice flows for intangible assets (such as technologythat passes among firms through licensingagreements and other forms of intellectual property)can be used to parse TFP growth estimates.

The TFP growth estimates proposed by theCommittee would be consistent with estimatescurrently published by the European Union, Canada,Japan, and South Korea and with estimates plannedby other countries, including Brazil, China, India,and Russia. The availability of consistently estimated

measures of TFP will allow for rigorous comparisonsof the sources of growth across countries and overtime. Such information would be valuable topolicymakers trying to gauge the effectiveness ofnational innovation policies.

� Create a supplemental innovation account forthe National Income and Product Accounts(NIPAs) in order to expand the categories ofinnovation inputs and allow those inputs to betracked as they flow between industries.

Supplemental accounts (sometimes referred to assatellite accounts) provide additional detail about apart of the economy using the same structure as theNational Income and Product Accounts withoutbeing integrated with the NIPAs. Establishing a newsupplemental account that is conceptually consistentwith the NIPAs but based on data that are not astime-tested as those used in the core NIPAs helpsfacilitate the development of the NIPAs. Forexample, in 2006, BEA inaugurated a supplementalaccount for investments in research and developmentactivities conducted by scientists and engineers;ultimately, BEA proposes to incorporate the R&Dsupplemental account into the NIPAs.

Developing a broad supplemental innovationaccount for intangible assets – including estimates ofintangible asset capital stocks and service flows –would introduce a broader range of capital inputsinto the NIPA product and capital accounts. Suchinputs have hitherto been classified as expenses ratherthan as accumulated capital. Data on intangible assetcapital stocks and service flows could be used tofurther refine estimates of the impacts of innovationon economic growth.

The Committee calls for the development of a newNIPA supplemental innovation account that wouldinclude intangible assets, such as investments byfirms in research and development (conducted byscientists and engineers as well as workers in otheroccupations); human capital; patents and tradesecrets; copyrights, trademarks and brands; and otherforms of intellectual property. The development of asupplemental innovation account depends on thecollection of additional data on intangibleinnovation assets.


8

� Improve service sector data and increase surveycoverage to provide the data needed to improveestimates from the integrated GDP/productivityaccounts and supplemental innovation account.

It is widely believed that much innovation in recentdecades has taken place in the service sector, and thisbelief is bolstered by evidence (albeit based onincomplete and highly aggregated data) that servicesector productivity has increased after a long periodof stagnation. However, collection of annual data onthe service sector has lagged that of other sectors (inparticular, manufacturing), and improved servicesector data collection is a necessary step toward bettermeasurement of innovation and its impacts in theservice sector.

The Census Bureau has long covered retail andwholesale trade in its annual surveys. Of the rest ofthe service sector, which currently accounts for 55percent of GDP, the Census Bureau’s annual surveycovers only 30 percent of GDP. Put another way, theCensus Bureau does not collect annual data on one-quarter of the nation’s economy.

The Committee recognizes that the funds needed toimplement the Census Bureau’s proposal to extendannual survey coverage to all remaining service sectorsby FY2010 and to improve the data on serviceindustry inputs have not been appropriated. Thecurrent lack of funding will impede some activitiesnecessary for improved innovation measurement. Forexample, BEA’s efforts to develop more detailed dataon the input/output matrices to track innovationacross industries hinges on the Census Bureaucollecting more comprehensive service sector data.

The Committee also calls for the development of amore nuanced classification system with finergranularity of data on different types of economicactivity in the service sector. In particular, theCommittee calls for improved accounting fortransfers of intangible assets (such as intellectualproperty licensing and assignments) and thedevelopment of more detailed and quality-adjustedprice indices for services.

Since a number of service sector industries (e.g.,software, consulting, and intellectual property

licensing and transfers) are considered particularlyinnovative, better data on the service sector willimprove our understanding of innovation processes.Improved services price indices will improveestimates of inflation-adjusted inputs purchased fromthe service sector inputs, including those purchasedfrom highly innovative industries, and will enhancethe quality of estimates of productivity growthattributable to innovation.

� Improve measurement of intangibles,particularly intellectual property, building onwork curently under way at the NationalScience Foundation. Consider the best way inwhich to collect data on transactions involvingkey intangible assets such as intellectualproperty licensing expenditures and revenues.The Commerce Department should also explorewhether additional identifying informationfrom patent and trademark applicants mightprovide useful data.

Statistical agencies largely focus on collecting data onfirms’ investments in physical capital. Little is knownabout firms’ investments in intangible assets beyonddata on expenditures on scientific and engineeringresearch and development collected by the CensusBureau for the National Science Foundation (NSF)and data on own-account software developmentcollected by the Census Bureau. Data oninvestments by firms in ‘research and development’activities other than scientific and engineering R&Dare scarce, as are data on investments inorganizational and human capital and marketing andbrand equity.

Data on patent and trademark applications andassignments are publicly available, but it is difficultfor researchers to match these data to data onapplicants and assignees. Better matching wouldhelp researchers develop ways to put a value onintellectual property.

Data are similarly scarce on the income earned onintellectual property. The Bureau of EconomicAnalysis collects data on cross-border royaltypayments for intangible assets, but no data arecurrently collected on purely domestic royaltyrevenues and payments.


9

The lack of information on intangible assets’depreciation rates and how to adjust the value ofinvestments in intangibles for price inflation makes itdifficult to estimate the wealth that is created byfirms’ investments in innovation. Sound estimates ofintangible capital stocks are needed to refineestimates of innovation’s contribution to productivitygrowth. However, intangible asset valuation isdifficult because of the relative scarcity of directlyobservable prices, and the data collection andmethodological challenges are daunting.

The Committee recommends improvements inintangibles measurement by expanding datacollection on intangible investments beyondscientific and engineering R&D and own-accountsoftware development to include all resourcesdevoted to innovation. Such improvements might beachieved by expanding NSF’s R&D survey toencompass a broader range of innovation investmentactivities. Past experience with collecting such data atBLS and in other countries (notably Canada) couldbe used to inform the development of a survey in theUnited States.

The U.S. Patent and Trademark Office should collectadditional identifying information from patentand trademark applicants to facilitate matchingapplication and assignment data to census andother data sets. Such information could includeEmployer Identification Numbers (with appropriateconfidentiality safeguards), CUSIP numbers, DUNSnumbers, or other widely used firm identifiers.Collecting such additional information shouldimpose little additional burden on patent andtrademark applicants and would aid the developmentof patent quality indices using stock marketvaluations and other data.

The Committee further recommends that BEA andthe Census Bureau collaborate to collect domestic (aswell as cross-border) data on revenues from andexpenses associated with licensing and transferringtechnology (patents and trade secrets), copyrights,franchise fees, and trademarks.

Finally, the Committee calls for the development ofimproved estimates of depreciation rates of intangibleassets and specific price indices for intangible assets.

2. THE GOVERNMENT SHOULD BETTERLEVERAGE EXISTING DATA AMONG THEGOVERNMENT STATISTICAL AGENCIES TOALLOW FOR THE CONSISTENT ESTIMA-TION OF THE CONTRIBUTIONS OF INNO-VATION IN THE GDP AND PRODUCTIVITYACCOUNTS AND TO DEVELOP GREATERUNDERSTANDING OF INNOVATION.

Much could be learned about innovation from datathat have already been collected if the statisticalagencies were to develop new linkages. Much of thiswork would require new legislative authority.

� Develop linkages within and between existingdata; for example, develop linkages betweenestablishment-based data sets and firm-baseddata sets to provide both greater consistency inestimations and to provide researchers a broaderrange of innovation data.

Most U.S. industry statistics are estimated usingestablishment-level data as the basic ‘building blocks.’An establishment is a commercial or non-profitentity at a single physical location that producesgoods or performs services, for example, a store or amanufacturing plant. Statistical agencies gather datafrom establishments located within the United Statesoperating within the same industry and publishdescriptive statistics for each industry based on suchdata. Such statistics have been very useful for manypurposes, as they combine data for establishmentsthat do approximately the same things.

However, many firms own or control more than oneestablishment, and those establishments may belocated in different geographic areas and may beproducing different kinds of goods or performingdifferent kinds of services, some of which may beimportant to innovation. Statistics based on firm ratherthan establishment data can be useful in othercontexts, especially in innovation measurement. Bydisassembling the establishment building blocks used toconstruct industry statistics and reassembling them intofirm-based statistics, better innovation measurementsmay be obtained. Furthermore, firm or establishmentdata from one data source can be augmented bymatching to corresponding data from other sources toobtain a more complete picture of innovation.


10

� Develop more robust classification methods;for example, classify firms on the basis of bothdomestic and international activities.

Even in those cases where data are collected andclassified at the firm level, classification is based onlyon the activities of domestic establishments.Consider a firm that designs and markets shoes underits own brand name in the United States butmanufactures them abroad. Based exclusively on thedata collected from establishments located in theUnited States, the firm would be classified as awholesaler, but such a firm would likely be verydifferent in terms of innovation activities from a firmthat acted solely as a merchant wholesaler (i.e.,buying shoes from manufacturers and reselling themto retailers without any design or manufacturingactivities). Classification on the basis of bothdomestic and international activities would improveinnovation measurement.

� Seek expanded interagency data sharing/synchronization legislative authority in orderto improve the quality of innovation measureswhile balancing data needs withconfidentiality protection.

While the Committee has given much considerationto new or expanded surveys of firms, much remainsto be learned from data that have already beencollected. However, because of a host of laws andother restrictions affecting firm micro-data (that is,establishment- or firm-specific data records), agenciesare often limited in their abilities to share data withother agencies.

Within the federal government, limitations placedupon the use of IRS data affect the accuracy ofbusiness registers used by federal agencies to carryout business surveys. Partially because of theseinaccuracies, differences arise in how differentagencies measure economic activity. Additionally,existing topical surveys have been hampered by lackof access to a good business frame from which todraw a sample.

The Committee recommends that Congress enactlegislation to enable and encourage statistical agenciesto expand data sharing activities subject to strict

confidentiality requirements. Specifically, theCommittee recommends that access to IRS data beexpanded to additional statistical agencies for thepurposes of reconciling the business lists anddesigning effective special surveys. The grant of suchauthority would greatly enhance data quality andconsistency and improve innovation measurement.

3. THE GOVERNMENT SHOULD INCREASEACCESS TO DATA IN ORDER TO FACILITATEMORE ROBUST INNOVATION RESEARCH.

� Increase the transparency of and access to publicdata by fostering the use of data tagging andsimilar processes.

The government collects a vast store of non-confidential data from firms. Analysis of such databy researchers might provide considerable newinformation on innovation and innovation-relatedactivities. For example, analysis of financial data andreports filed with the Securities and ExchangeCommission (SEC) by public companies mightprovide insights into what differentiates innovativefirms. But, until recently, such data have been lockedinto a format consisting of fixed blocks of text,making it difficult for researchers to process andanalyze the data.

Government agencies have begun a process ofmaking public data more user-friendly; for example,the SEC, the Federal Deposit Insurance Corporationand other banking regulators, and several Europeangovernments are using or testing XBRL (ExtensibleBusiness Reporting Language), a process that ‘tags’data (i.e., attaches a standardized, searchable label toeach data item filed by firms). Tagging enables‘intelligent’ computerized recognition of each pieceof information in a document, thereby facilitatingaccurate, automated, and flexible selection, analysis,storage, exchange, and presentation of data at highspeed and reduced cost. The SEC data-taggingproject is still in its pilot stage but some firms arereporting tagged data.

Expanded use of data tagging of public informationcould increase transparency of public informationand provide a wealth of information for researchersand other members of the public.


11

� Create more public use files in order toencourage more non-governmental research.

Statistical agencies should consider creating morepublic use files that would be accessible to non-government researchers. While confidentialityrequirements may impose some limits to the depth ofpublic use files, greater public access through web-based public use files could increase innovationresearch since the data would be far more widelyavailable. This is not considered a substitute butrather a complement to greater accessibility ofconfidential micro-data to the research community.

� Expand non-government researcher access toconfidential micro-data, including that onbusiness dynamics, while maintaining highstandards for confidentiality.

The Committee recommends that the statisticalagencies expand research access to confidentialmicro-data while maintaining high standards forpreserving confidentiality. Outside researcher accessto business data, particularly business micro-data,collected by the Federal agencies remains extremelylimited. Many scholars are interested in – and withproper confidentiality training and safeguards couldbe engaged in – research that would further theunderstanding of different aspects of innovation.While concerns about confidentiality are importantand should be at the fore of any discussion onresearcher access, current limitations on access areimpeding understanding of the dynamic aspects ofthe economy. Federal agencies do not have adequatefunding for in-house staff to exploit fully theresearch potential of the data. In addition, advancesin remote access technologies and the success to dateof specific federal agencies in offering outsideresearcher access make this an opportune time toexpand outside researcher access to businessmicro-data. Academic researchers, subject toconfidentiality provisions, could facilitate linkagesbetween confidential micro-data and other existingdata such as data collected by the Securities andExchange Commission, thus providing a richer

picture of innovation in the economy. Someagencies have made significant progress indemonstrating the feasibility of expanded accesswhile maintaining strict confidentiality.

4. THE GOVERNMENT SHOULD CONVENEONE OR MORE WORKSHOPS OR FORUMSUNDER THE AUSPICES OF THE SECRETARYOF COMMERCE TO DISCUSS INNOVATIONDRIVERS, IMPEDIMENTS AND ENABLERS.

Many of the members suggested that the developmentof more effective innovation policies requiresundertaking more work to get better identificationand examination of innovation drivers, impedimentsand enablers. Insofar as the Advisory Committee’scharge was limited to innovation measurement, theseissues have not been directly addressed by theCommittee. Nonetheless, given the importance ofthe issue in the view of most of the Committeemembers, the Committee is recommending that theSecretary of Commerce convene one or moreworkshops or forums devoted to examininginnovation drivers, impediments and enablers.

Members of the Committee and public commentersidentified a broad variety of drivers and enablers ofinnovation. These included tax policy (R&D taxcredits); education policy (promoting science andmath); the availability of venture capital; the businessclimate; ease of entry into and exit from markets;education and skills; research and developmentexpenditures; and specific firm culture andenvironment. However, in many instances, thecorrelation between the identified driver or enablerand innovation outcomes is not well understood orconvincingly demonstrated empirically.

Impediments to innovation that have been suggestedinclude some immigration policies, trade restrictions,Sarbanes-Oxley rules (particularly Section 404), andother government regulations. As with drivers andenablers, there are many assertions made but theactual impact of such policies on innovation activityis much less clear.


12

5. THE GOVERNMENT SHOULD CONTINUEPARTICIPATION IN THE INTERNATIONALDIALOGUE RELATED TO MEASURING ANDANALYZING INNOVATION AND ENSURETHAT U.S. EFFORTS ARE INTERNATIONALLYCOMPATIBLE TO THE EXTENT POSSIBLE.

The Advisory Committee is aware of the work thathas been and is being done on innovationmeasurement in the international community. TheAdvisory Committee supports the internationaldialogue on innovation and, indeed, one of theprinciples adopted by the Committee calls forinnovation measures to allow for analysis atinternational levels.

The United States is actively engaged in internationaldialogue on innovation and innovation measurement.The National Science Foundation works closely withthe Organisation for Economic Co-operation andDevelopment’s Working Party of National Experts onScience and Technology Indicators (NESTI).NESTI’s mission – to “ensure the continuedimprovement of the methodologies for the collectionof internationally comparable data for measuring theinput, output, diffusion and impact of science,technology and innovation (including linkages toeconomic growth)” – is similar to the mission of thisAdvisory Committee. The Financial AccountingStandards Board (FASB) and the InternationalAccounting Standards Board (IASB) are workingtogether to converge two sets of standards, whichcurrently differ on the circumstances under whichresearch and development costs can be expensedversus capitalized as intangible assets. The Committeesupports such efforts.

6. THE GOVERNMENT SHOULD CONSIDERDEVELOPMENT OF A NATIONAL INNO-VATION INDEX WHEN MORE WORK HASBEEN DONE ON BOTH DATA COLLECTIONAND ANALYSIS OF INNOVATION DRIVERS.

An index of innovation or innovation scorecard ordashboard could give some sense of how the U.S.fares on a combination of different measures ofinnovation. However, even among the members whosupported the idea of constructing an index, therewas uncertainty as to what measures should be

aggregated and how the components should beweighted, if done as a scorecard, and the purpose towhich the index would be put. Additionally, whilesome innovations have immediate and significanteconomic impact, many innovations appear to have alonger lag before affecting output. As such, a carefulconsideration of how to weight the long-term andshort-term impacts of innovations is necessary priorto implementing an index.

Consideration would need to be given as to the typesof measures that would be reported and whether ornot data currently are collected on these measures;whether this should be all public data or a mix ofpublic and private information; and whether or not itshould be done immediately or be an end productfollowing the development of new innovationmeasures. Such an index could be a useful tool forinternational comparisons if there were comparabilitywith measures used in other countries. Thedevelopment of an innovation index could be usefulin the future when more data are available and moreis known about specific innovation drivers.

7. THE GOVERNMENT SHOULD SUPPORTFUNDING NECESSARY TO IMPLEMENT THEABOVE RECOMMENDATIONS.

Many of the above recommendations will requirenew funding. The Advisory Committee recommendsthat the government support the additional fundingnecessary.


13

Innovation is a collaborative process and there ismuch that the business community – individualfirms and business organizations – can do to

both assist and drive improvements in innovationmeasurement. One of the guiding principlesendorsed by the Advisory Committee was thatinnovation data collection efforts should build on theway firms assess the effectiveness of their innovativeactivities. Insights from individual firms and otherorganizations can be integrated with those fromgovernment efforts to assess innovation activities andperformance more comprehensively.

1. THE ADVISORY COMMITTEE RECOM-MENDS THAT THE BUSINESS COMMUNITYCREATE, EXPAND AND ASSESS FIRM ANDINDUSTRY-LEVEL MEASURES OF INNOVA-TION AND DEVELOP BEST PRACTICESFOR INNOVATION MEASUREMENT.

To provide information on firm and industry-levelinnovation measures and assess their effectiveness, theAdvisory Committee calls on the businesscommunity to:

� Institute firm-level measures of innovation totest the correlation of particular measures withknown innovation and to measure innovationin the firm and its impact on firm performance.Possible measures might be based on marketshare or on innovation intensity.

Two possible market share measures of innovationwere discussed by the Committee. The rationale forlooking at a measures based on market share is theassumption that products and services gain marketshare because they provide better value by beinglower priced and/or better satisfying customers’needs. One variation focuses specifically on the shareof a firm’s new product sales in markets in which thefirm is gaining market share. The second variationwould measure the extent to which a firm’s overall

share is growing in a growing market. Individualfirms might wish to develop their own market sharemeasures to measure their own innovationperformance. In addition, trade associations or otherorganizations might develop such measures forparticular industries and aggregate the results intoindustry-level measures.

A major issue in designing survey questions based onmarket share is how to define ‘market’ in a way thatis sufficiently clear to survey respondents to yield ameaningful innovation measure, since a growingmarket share even in a growing market is notnecessarily the result of innovative activity.

Measures of innovation intensity (i.e., the share offirm revenue attributable to recently introducedproducts and services) may be another useful way totrack the impact of innovation. Again, individualfirms could track such a measure for their own use ortrade associations or other organizations coulddevelop surveys to collect data on total sales revenuesand those attributable to products and servicesintroduced to the market in the preceding threeyears. The data would be used to calculateinnovation intensities for each firm in the sample andperhaps could be aggregated to arrive at publishable,industry- and economy-level measures. Researcherscould also analyze the resulting firm-level data tobetter understand the characteristics andperformance of firms with different innovationintensities.

Firms with high innovation intensities wouldpresumably be among the most successful indesigning, inventing, developing, and implementingnew or altered products and services. However, thereare formidable interpretive challenges. A ‘new’product or service may be ‘new’ only to the firm andit could be difficult to differentiate truly innovativeactivity. ‘New to market’ in manufacturing may onlycapture product innovations; while ‘new to firm’ in


15

Chapter FourHOW THE BUSINESS COMMUNITY CAN HELP

service industries may be an indirect indicator ofinnovations in processes, systems, organizationalstructures, or business models. At least initially, suchmeasures may be most useful for individual firms todevelop and test the usefulness of their owninnovation intensities.

The Organization for Economic Cooperation andDevelopment (OECD) has designed surveyquestions to collect data for the development ofmeasures of innovation intensity. Several countrieshave implemented such surveys (e.g., the EUCommunity Innovation Survey), and the U.S.business community could build on that experience.

The Committee calls on the business community,particularly individual firms and trade associations,in cooperation with the research community, toexperiment with collecting data on market share-based and innovation intensity measures, analyze therelationship between such measures and firmperformance and report on the findings. Insightsfrom such analyses could be integrated with thosefrom government and academia to provide acomprehensive view of innovation in the economy.

� Develop and implement best practices ininnovation management and accounting.

The development of best practices in innovationmanagement and accounting appears to be in itsinfancy. The Committee calls on the businesscommunity, particularly trade associations, to collectinformation about current innovation managementand accounting practices, and calls on universitiesto develop curricula and research programs ininnovation management and accounting. Thegovernment can help academic research programs in

this area by supporting such efforts through agenciessuch as the National Science Foundation.

2. THE ADVISORY COMMITTEE RECOM-MENDS THAT THE BUSINESS COMMUNITYPARTICIPATE IN RESEARCH ACTIVITIESAND, AS APPROPRIATE, MAKE INNOVA-TION INFORMATION AVAILABLE TORESEARCHERS.

Sharing knowledge about innovation andparticipating in the development of it is a way thebusiness community can help drive the way towardbetter innovation measurement. In particular, theAdvisory Committee recommends that firms andassociations:

� Participate in collaborative projects as a meansof assembling a broad range of data relatedto innovation.

Private sector initiatives to collect and analyze data oninnovation activities and outcomes can complementand help inform government data collection andpublication efforts as well as help identify anddisseminate best practices in innovationmanagement.

� File public reports in XBRL format when it isan option.

As noted elsewhere in the report, firms couldfacilitate measurement and analysis of innovationactivities and outcomes by reporting financial andother information in XBRL format. The Committeecalls on firms to report information in XBRL formatwhenever possible.


16

Much more needs to be learned aboutmeasuring innovation and its impact.Knowledge about innovation would be

greatly enhanced if government, academic and privatesector researchers– alone and in collaborative efforts –were to undertake research in the following areas:

1. IDENTIFICATION AND ASSESSMENT OFINNOVATION OUTCOME MEASURES.

The definition for innovation adopted by theAdvisory Committee focuses on innovationoutcomes. Much of the innovation measurementcurrently focuses solely on inputs to innovation.More work needs to be done to define appropriateoutcome measures and analyze their utility andeffectiveness. Particular research areas include:

� Assessment of the effectiveness of measures basedon market share as innovation measures andthe feasibility, cost, and burden of developingsuch measures.

The Advisory Committee has called on the businesscommunity to develop and test measures based onmarket share. Researchers, too, need to examinethese measures. The Committee calls on researchersto assess the feasibility, cost, and effectiveness ofcollecting data to develop measures based on marketshare, and, where appropriate, to sponsor orparticipate in pilot projects to test such measures.

� Assessment of the feasibility, cost, and burden ofdeveloping measures of innovation intensity,including a review of other countries’ experiencein this area and consideration of a pilot project.

High innovation intensity (the share of revenuesearned by recently introduced products and services)is also considered to be a possible indicator ofsuccessful innovation. The Committee calls forresearch to assess the feasibility, cost, and effectiveness

of collecting data to estimate innovation intensitymeasures, and, where appropriate, to sponsor orparticipate in pilot projects to test such measures.

� Analysis of the qualitative and quantitativeimpacts of specific drivers, impediments andenablers on innovation outcomes.

Work needs to be done to determine the actualimpacts of particular drivers, enablers andimpediments – both internal to firms (such as firmculture) and externally imposed (such as accountingrules and immigration policies). In many instances,the connection between an identified driver, enableror impediment and innovation outcomes is neitherwell understood nor convincingly demonstratedempirically.

2. IDENTIFICATION OF GAPS IN INNOVA-TION DATA AND HOW THEY MIGHTBE FILLED.

There are many gaps in the collection of data thatmight explain innovation better. The AdvisoryCommittee urges work in the following areas:

� Identification of new data that would be usefulin measuring innovation.

Simply implementing the Committee’s proposals fordata collection is unlikely to answer all questionsabout innovation. The process of implementing theproposals will raise new questions and stimulate newideas for measuring various facets of innovation. TheCommittee urges that researchers continually probethe data generated by this project to discover ways toimprove the data or propose new kinds of data thatwould help fulfill the Committee’s mission, andidentify data the Committee has proposed collectingthat should be abandoned as not cost-effective withrespect to the Committee’s mission.


17

Chapter FiveWHERE RESEARCH IS NEEDED

� Assessment of the feasibility, cost and burdenof collecting data on intellectual propertytransactions.

The Committee has proposed the collection of morecomplete data on firms’ total expenditures on andrevenues from intellectual property transactions(including both domestic as well as cross-bordertransactions). In addition, the Committee urgesresearchers to assess the feasibility, cost and burden ofcollecting such data not just at the firm level but alsoat the transaction level by contracting with arepresentative sample of buyers and sellers in theacademic, non-profit, and private sectors to compileperiodic data on transactions involving technologylicensing. It may be appropriate to carry out apilot study to assess the nature and extent of thedata available.

Measuring this “innovations market” is vital formeasuring innovation in a knowledge economy.Insofar as innovation results in the creation ofintangible knowledge assets, which yield a flow ofbenefits over time, measuring transactions inintangible knowledge assets is a good way to measurethe value of these assets. Further, transactions datawould complement the proposed new NIPAarchitecture by providing more accurate measures ofpurchases of the services of intangible inputs. Giventhat such purchases have grown very substantially inthe last two decades, ignoring them risks the veryobjective that the new NIPA architecture is toaccomplish. Finally, transactions data wouldcomplement the proposal to improve measures ofservices industries; many services firms play a crucialrole in generating and diffusing innovations throughthe economy.

� Identification of ways to overcome gaps andshortcomings in historical measures ofintangible investments.

The proposals on measuring intangible assetsdiscussed elsewhere in this report focus mainly oncollecting data on new investments in intangibles orestimates of the market value of intangible capitalstocks. Absent good data on the market value ofintangible capital stocks, however, it may benecessary to develop historical cost estimates in order

to account for the contribution of intangible capitalstocks to productivity growth. The Committeeencourages researchers to develop methods forestimating past investments in intangibles.

3. ANALYSIS OF RELATIONSHIPS BETWEENINNOVATION ACTIVITIES AND COLLAB-ORATION, INNOVATION PERFORMANCE,AND FIRM PERFORMANCE.

A major goal of innovation measurement is todevelop estimates of the rate of return to investmentsin innovation activities.

A methodological challenge is to untangle thedirection of causality between activities andinnovation performance. Suppose researchers findthat increasing innovation activity X (e.g., spendingon a broad set of innovation activities, including butnot limited to R&D) is associated with an increase infirm innovation performance Y (e.g., number of newtrademarks). And suppose researchers also foundthat increasing firm innovation performance Y isassociated with an increase in firm financialperformance measure Z (e.g., sales revenue). Doesthis mean that X causes Z? Or do growing firmssimply tend to spend more on innovation activities?

Another methodological and data challenge is dealingwith the fact that much innovation is not donestrictly within a firm’s boundaries – firms collaboratewith other firms and with universities; firms purchaseand sell intellectual property rights; firms outsourcesome innovation activities – so getting the wholepicture of innovation is key to estimating the rate ofreturn to innovation activities. And, indeed,innovative activity happens in an internationalsetting where many of these interactions crossnational borders.

� Analysis of the relationship between innovationand occupational employment using firm-levelmicro-data.

Some observers have suggested that the United Statesfaces shortages of workers in scientific, engineering,and technical (SET) occupations that will harm ourability to compete with other countries in the globalmarketplace. Others have suggested that firms in the


18

United States can leverage their SET employees byoutsourcing some scientific R&D and otherinnovation-related work or by increasingimmigration of SET workers. Still others havesuggested that competitiveness depends on a broaderrange of occupations that includes other kinds ofcreative workers (e.g., artists and designers) making itharder to say definitively that the United States is‘falling behind’ other economies when it comes to thesupply of innovation workers.

A possible research area using the data generated byimplementing the Committee’s proposals is to explorethe relationship between occupational employment –of SET workers and more broadly defined ‘innovationworkers’ – and various measures of innovation outputssuch as intellectual property (patents, copyrights,trademarks), intellectual property licensing revenueand expenditure flows, innovation intensity, andmeasures of firm performance.

� Evaluation of whether firms with highinnovation intensities perform better thanotherwise similar firms with low intensities.

This is perhaps the most direct measure of how theintroduction of new products and services isassociated with firm performance. It provides a keylink that helps explain how investments ininnovation may yield better firm performance. Itdoes so by bridging the gap between investment andfinancial results to get at how the introduction of newproducts has an impact on firm profitability.

� Assessment of the effect of collaboration oninnovation outcomes and identification of thekey elements of successful collaborative activities.

Collaboration is believed to be a key element infostering innovation yet little research has been doneto confirm this belief. Work needs to be done bothto develop methods to analyze collaboration andassess its actual impact.

� Assessment and analysis of cross-nationalinnovation activities of firms.

In today’s globally connected world, many U.S.multinationals conduct a significant part of their

innovation activities outside the United States.Conversely, many non-U.S. firms conduct asignificant part of their innovation activities in theUnited States. These cross-national innovationactivities undoubtedly affect the U.S. economy.The statistical agencies are currently working to linkdata on domestic and international researchand development expenses by U.S. and foreignmultinational firms. Such efforts should beencouraged and expanded to permit examination ofthe output of cross-national innovation activities.

� Analysis of publicly filed financial and otherdata on firms, particularly as the data becomemore user-friendly, to identify innovativepractices and firms.

Publicly filed financial data could be used to examinecharacteristics of firms considered innovative and toconsider whether these innovative firms have uniquepractices or characteristics.

� Description and explanation of businessdynamics and, to the extent possible, analysisof their relationship to innovation usinglongitudinal business databases.

Studying the entry, growth, and exit ofestablishments and firms could provide importantinformation on the characteristics of successful andunsuccessful firms. Much innovation occurs in thesmallest firms, where high exit and entry rates arecommon. Evidence for several industries suggeststhat entering firms are, on average, more productivethan exiting firms but the role played by innovationis not known.

The Census Bureau is developing a longitudinalbusiness database from which it is planning topublish data on establishment and firm births anddeaths, job creation and destruction by firm size, age,and industrial sector, and other measures of businessdynamics over the past three decades. The resultingdata will help researchers distinguish betweenthe effects of short-term economic fluctuationsand long-term effects (such as innovation) onbusiness dynamics.


19

Analysis of longitudinal business dynamics data isbelieved to be able to provide insight into thepatterns of innovative activity in the economy andthe innovative and entrepreneurial environment invarious time periods. New data sets such as theKauffman Firm Survey show potential forunderstanding how high-impact firms evolve, butlonger business panel surveys and additionalgovernment support are needed to exploit theirfull potential.

The Committee encourages the Census Bureau tocontinue its efforts to publish reports on businessdynamics and explore the possibility of developingpublic use micro-data on business dynamics. TheCommittee also encourages researchers to uselongitudinal business databases to help describe andexplain business dynamics and, to the extent possible,analyze its relationship to innovation.

� Exploration of the use of different sources ofavailable data to determine whether there arecorrelations between innovative performanceof firms and the existence of a new regulatoryenvironment.

Regulatory policies are believed to impact innovation– both positively and negatively – but there is little orno empirical evidence of their impact.

CONCLUSION

The Advisory Committee’s recommendations, ifadopted, will go far in setting this nation on a coursetoward effectively measuring the impact ofinnovation on the economy. The work is essentialand the Committee calls on the government, thebusiness community and researchers to adopt therecommendations.


20




January 2008


The Advisory Committee on Measuring Innovationin the 21st Century Economy

http://www.innovationmetrics.gov/

Covers 1 and 4.qxd 1/7/08 1:39 PM Page 1

Innovation Measurement - Tracking the State of Innovation...

Documents

Transcript of Innovation Measurement - Tracking the State of Innovation...