knowledge economy measurement: methods, results and insights from the malaysian knowledge content...
TRANSCRIPT
Research Policy 35 (2006) 1522–1537
Knowledge economy measurement: Methods, results and insightsfrom the Malaysian Knowledge Content Study
Philip Shapira a,∗, Jan Youtie b, K. Yogeesvaran c, Zakiah Jaafar c
a School of Public Policy, Georgia Institute of Technology, Atlanta, GA 30332-0345, USAb Georgia Tech Enterprise Innovation Institute, Atlanta, GA 30332-0640, USA
c Economic Planning Unit, Prime Minister’s Department, 62502 Putrajaya, Malaysia
Available online 9 November 2006
Abstract
Building on a conceptual model of knowledge content, we discuss the methodology and results of a project to develop sectoralknowledge content measures in Malaysia. Through a survey of over 1800 Malaysian firms in 18 manufacturing and servicesindustries, levels of knowledge content are assessed by sector. Industries vary in their emphasis on specific knowledge contentcomponents. Positive associations between technological innovation and at least one knowledge content variable are evident across
all but four industries, although generally the results suggest that knowledge-based innovation is modest in Malaysia. Further insightsand opportunities for policy from tracking knowledge content are considered.© 2006 Elsevier B.V. All rights reserved.JEL classification: D83; O21; O32
y
a study of Knowledge Content in Key Economic Sectorsin Malaysia1 and an associated Malaysian Knowledge
Keywords: Knowledge measurement; Innovation; Development polic
1. Introduction
Many countries and regions are seeking to shift theireconomies to become more knowledge-intensive. How-ever, there are no standard methods of describing theextent to which an economy is knowledge-intensive or, inparticular, of measuring levels or changes in the ‘knowl-edge content’ of the various sectors that comprise aneconomic system. We define knowledge content as “thesum of human capabilities, leadership assets and experi-
ence, technology and information capital, collaborativerelationships, intellectual property, information stocks,and capabilities for shared learning and utilization that∗ Corresponding author. Tel.: +1 404 894 7735;fax: +1 404 385 0504.
E-mail addresses: [email protected] (P. Shapira),[email protected] (J. Youtie), [email protected](K. Yogeesvaran), [email protected] (Z. Jaafar).
0048-7333/$ – see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.respol.2006.09.015
can be used to create wealth and foster economic com-petitiveness”. We then describe the results of a study thatsought to operationalize and measure this definition ofknowledge content.
The paper draws on the first attempt by the MalaysianGovernment to assess Malaysia’s knowledge content inkey economic sectors. The assessment was undertaken in
Content (MyKe) Survey. The MyKe survey was con-
1 The study on Knowledge Content in Key Economic Sectors inMalaysia was undertaken during the period 2002––2004 by the Tech-nology Policy Assessment Center (TPAC) at Georgia Institute ofTechnology and Intelligent Information Services Corporation (IISC)in collaboration with the Malaysian Economic Planning Unit. Studysponsorship was provided by the United Nations Development Pro-gramme (UNDP).
ch Polic
dcgMaiktiab
rcdkitookpe
2
fcof(sclS
tTongs2ltek
P
P. Shapira et al. / Resear
ucted in 2002/2003 by the Malaysian Government,2 inollaboration with a team led by researchers from Geor-ia Institute of Technology (USA). The objective of theyKe survey was to assess the knowledge content char-
cteristics and constraints of 18 key private industriesn Malaysia towards enhancing policy making for thenowledge economy. This sectoral perspective is distinc-ive relative to the metrics typically used in national-levelndicator studies of knowledge. All previous attemptst assessing Malaysia’s progress towards a knowledge-ased economy were at the macro-level.
Following sections describe the methodologies andesults of this extensive project to develop knowledgeontent measures at the sectoral level in Malaysia. Weiscuss the development and use of a range of metrics ofnowledge content, and discuss interrelationships andssues in their measurement and usage, particularly inhe context of developing countries. Based on analysesf the MyKe survey, results and conclusions are thenffered. Our aim is to offer insights related to developingnowledge measurement systems that can be helpful toublic policy decision-makers when implementing andvaluating knowledge economy development policies.
. Policy context
In recent decades, Malaysia has transformed itselfrom a country that long depended on agriculturalommodities and mining to an industrializing econ-my where manufacturing and services now accountor 32 percent and 57 percent of GDP, respectively2005 data, Department of Statistics Malaysia, 2005a;ee also Economist, 2004). Today, manufactured goodsontribute more than four-fifths of Malaysia’s exports,ed by exports of electronic products (Department oftatistics Malaysia, 2005b).
However, transitioning to an industrialized produc-ion economy is not the end objective of policymakers.he Malaysian Government has established the goalf developing a knowledge-based economy to advanceational economic growth and competitiveness. Thisoal was highlighted in Malaysia’s Third Outline Per-pective Plan, 2001–2010 (Economic Planning Unit,001). A Knowledge-Based Economy Master Plan was
aunched in 2002 and contains 136 recommendationso accelerate the transformation to a knowledge-basedconomy (ISIS, 2002). However, the foundation for thenowledge-based economy began in the mid-1990s in2 The MyKe Survey was implemented by the Malaysian Economiclanning Unit and the Malaysian Department of Statistics.
y 35 (2006) 1522–1537 1523
the areas of human resource, information and com-munications technology (ICT), science and technology(S&T), research and development (R&D), infostructureand financing (Mani, 2001). Malaysia has also initi-ated efforts to try to ensure that the development of aknowledge-based economy does not result in a knowl-edge divide (Sahibbudin, 2001).
The development of S&T and the promotion of R&Dhave been integrated into the overall national develop-ment plan since the First National Science and Technol-ogy Policy and the Industrial Technology DevelopmentPolicy were drawn up in 1986. In 2003, the MalaysianGovernment launched the Second Science and Tech-nology Policy (STEP2) to further strengthen S&T andto spur greater technological development as well asa more innovation-led growth. Several growth-enablinginitiatives were also introduced to achieve these statedobjectives, chief among them was to grow a strongerbase of knowledge workers to meet the demands of thetechnological age.
To implement these plans, a series of strategicinitiatives in human resource development (HRD)were initiated to intensify the growth and creation ofa critical mass for S&T in Malaysia. These includeincreasing the proportion of students pursuing science,technical and engineering disciplines at high school,university undergraduate, and post-graduate levels;increasing support for science fellowships, trainingscientists and researchers in public research institutesand higher education; strengthening industry-led train-ing programs; and establishing distinguished visitingscientist programs in research institutes and universitiesin collaboration with industry. STEP2 also aims toincrease Malaysia’s national R&D spending to at least1.5 percent of GDP by the year 2010 through sevenstrategic thrusts. These are aimed at strengtheningresearch and technological capacity and capabil-ity; research commercialization; developing humanresources; promoting a culture of science, innovationand techno-entrepreneurship; strengthening institutionalframeworks and S&T management while implementinga more market friendly S&T policy; accelerating thewidespread diffusion and application of technology;and building competence in key emerging technologies.
To enhance private sector involvement and commit-ments in R&D activities, three grant schemes have beenestablished by the Malaysian Government. The IndustryResearch and Development Grant Scheme provides risk-
sharing between the Government and the private sector inR&D activities; the Multimedia Super Corridor (MSC)Research and Development Grant Scheme promotes thedevelopment of R&D clusters among MSC-status com-
ch Polic
1524 P. Shapira et al. / Researpanies with at least 30 percent Malaysian equity; andthe Demonstrator Application Grant Scheme encouragesthe diffusion of ICT into the community. Additionally, toaccelerate the rate of commercialization of R&D findingswithin the public sector, the Government has introducedan action plan to: strengthen the institutional and supportstructure for R&D efforts; stimulate R&D commercial-ization; encourage the development and adoption of keyemerging technologies; generate greater sensitivity tothe commercial potential of R&D activities and therebyboost the total number of innovations; enhance thecommercialization expertise of research institutions andnational centers of excellences; strengthen funding, andpolicies to enhance R&D commercialization; intensifycollaboration between industry, the Government, univer-sities, research institutions and their counterparts abroad;and create a large pool of highly skilled and knowledge-centric workers capable of meeting new challenges.
With these and other policies to promote a moreknowledge-intensive and innovative economy that theMalaysian Government has introduced, a new challengearises: how can progress towards a knowledge economybe assessed? In particular, what measures can be madeof the current level of knowledge in the economy at thesectoral level and how can changes in these measures beassessed in future years?
Answering such questions goes far beyond the simpleequation of knowledge with input indicators, such asR&D spending as a percentage of GDP or the proportionof scientists and engineers in the workforce. It alsorequires going beyond the national macro-level science,technology and knowledge indices that internationalbodies (for example, UNDP, 2001; Commission ofthe European Communities, 2004) and a number ofcountries have implemented to date. Indeed, Malaysia,through its Economic Planning Unit, developed aknowledge-based development index in its ThirdOutline Perspective Plan (Economic Planning Unit,2001). This comprised a series of comparative nationalperformance measures in education and training (e.g.higher education enrollment), R&D and technology(e.g. scientists and engineers in R&D), and computerand information infrastructures (e.g. computers or cellphones per 1000 people). This index placed the US,Japan, Sweden, Finland, and Norway in the top 5countries, with Malaysia ranking 17th, below Singaporeand above Thailand and China.
However, notwithstanding the utility of such indices
in other contexts, for the purposes of building a systemthat can be used for detailed knowledge measurementand assessment at the sectoral level, the further devel-opment of knowledge concepts and models is required.y 35 (2006) 1522–1537
This is what the MyKe study sought to do. But, beforedescribing the study and its results, the next section firstdiscusses some of the key ideas and trends in knowl-edge measurement. This is helpful in understanding thelogic of the model that the MyKe study developed andits points of similarities and departures from work doneto date.
3. Knowledge and its measurement
3.1. Interest in the knowledge economy
It is evident that in current policy discussions abouteconomic development, and national and regional com-petitiveness there is heightened attention to the role ofknowledge and its relationships to innovation and eco-nomic performance (see, for example, OECD, 1996).Of course, knowledge has always been understood tocontribute to economic growth. From Adam Smith andKarl Marx to Alfred Marshall and Joseph Schumpeter,economic thinkers have highlighted the importance ofknowledge-dependent factors – such as skill, the orga-nization of production, the development of technologyand innovation – in the growth of productivity and eco-nomic development. But, in recent times, the significanceattributed to knowledge in economic development hasmarkedly increased; indeed, arguably the growth in thescale, scope, and character of knowledge has itself trans-formed the ways in which successful businesses andeconomies operate. Today, much attention is paid to anew global ‘knowledge economy’ where information,skill and know-how is critical not only to business com-petitiveness but also to broader processes of regionaland national economic development (see, for exam-ple, Nonaka and Takeuchi, 1995; Stewart, 1997; Cooke,2002). In advanced economies as well as developingcountries, intangible knowledge inputs appear to be evermore important in the output of goods and services (rela-tive to the conventional tangible economic inputs of land,labor, and capital).
Additionally, recognition has grown in scientific dis-course, business, and policymaking of the importance ofimplicit, tacit, or embodied knowledge as well as of theaccumulation of formal or explicit knowledge (Polanyi,1967; Leonard and Sensiper, 1998; Senker, 1995).Knowledge acquisition and learning through experi-ence, interaction, heterogeneity, and network exchangehas grown in competitive importance relative to for-
mally produced, discipline-based, scientific knowledge(OECD, 2001). Moreover, the ways in which companies,industries, and institutions manage knowledge, applyinformation technology, and develop systems to enhance
ch Polic
ci
oasultaarkaieikiberckbtpcpk
3n
ufiltsmlmefinoeoats
P. Shapira et al. / Resear
apability and competence have surfaced as key factorsn business and economic performance.
For developing economies, these shifts in the rolef knowledge in the economy present both challengesnd opportunities. Traditional industrial developmenttrategies that rely on commodity production of man-factures or routine assembly or services clearly haveimits in terms of growth—unless ways can be foundo leverage knowledge and use it to enhance productsnd production processes to add value to customersnd end-users. All economies now need to developesources and capabilities to acquire and apply newnowledge. Education and training, as well as researchnd development, assume increasing importance. Yet,t is also critical to build the internal capabilities ofnterprises and to embed industrial sectors and regionsn networked relationships that facilitate externalnowledge economies. In dealing with such challenges,ndustrializing countries have follower advantages ineing able to benchmark experience elsewhere. How-ver, the rapid evolution and globalization of knowledgeeduces the effectiveness of this approach unless it is alsooupled with initiatives to identify and pursue distinctivenowledge development strategies. In determining thealance of challenges and opportunities facing indus-rializing countries in the new knowledge economy,ublic policy assumes an influential role, with the addedhallenge that governments themselves must adopt newublic management approaches, including improvednowledge-based systems, to guide policy development.
.2. Knowledge measurement at the enterprise andational levels
To date, efforts to measure knowledge have beenndertaken at one of two levels: first, at the individualrm level; and, second, at the national macro-economic
evel. Inevitably, because knowledge has informal andacit aspects (as well as formal or codified forms), alluch measurements involve proxies and indirect esti-ates where direct measurement is not possible. Firm-
evel measurement arises out of business initiatives toanage knowledge and measure intangible assets. These
fforts are operationalized at the micro or individualrm level and use a combination of accounting andon-financial indicators to measure stocks of intellectualr knowledge capital and flows of changes in knowl-dge stocks (Stewart, 1997). The knowledge capacity
f firms is proxied by means of instruments like bal-nced scorecards, intangible assets monitoring, intellec-ual capital accounts, and stylized models of knowledgepillovers (Sveiby, 1997, 1998; Lev, 2001; Boudreau,y 35 (2006) 1522–1537 1525
2002). In addition to knowledge stocks and flows, knowl-edge enablers are measured as a way of identifyingpractices with the potential to change or maintain knowl-edge stocks and flows. These may include leadership,strategy, organizational partnerships, or talent (Kermally,2002). Other recent studies have focused on document-ing the adoption of knowledge management practicesby enterprises. For a sample of Canadian firms, Earl(2002) examined the use of 23 business practices associ-ated with knowledge management in areas of leadership,knowledge capture, training, policies, communication,and incentives. Similar studies have been undertaken inDenmark, France, and Germany (OECD, 2004).
At the macro-level, efforts have been undertaken inrecent decades to build economic models that can capturethe generation of ideas and their association with wealthin the production function. Conceptually, the genericproduction function relates total product to combinationof labor, capital, and other inputs. The deficiency of thebasic Cobb–Douglas function in handling new innova-tions and endogenous technical change has since resultedin many refinements, dating back to the seminal work ofSolow (1957) and Abramowitz (1956). The resulting lit-erature, termed ‘growth accounting’, attempts to disag-gregate the residual in the standard production functionby employing increasingly sophisticated econometricmethods. Knowledge is seen as embodied in techni-cal change (Solow, 1957; Abramowitz, 1956). A morerecent development is the ‘knowledge production func-tion’ which postulates the generation of new knowledgeto be dependent on R&D capital, labor and other inputs.Various measures of ‘new knowledge’, including cita-tion weighted patents as well as new product announce-ments have been used in these econometric models (seeGriliches, 1990, 1992, for discussions of related efforts).
As noted previously, there have also been a vari-ety of attempts by international bodies and countriesto develop indices of science, technology or knowl-edge standing (see also discussion in Grupp and Mogee,2004). For example, the UNDP’s Technology Achieve-ment Index (UNDP, 2001) is a comparative nationalmacro-composite of indicators for technology creation(e.g. patents per capita), diffusion of new innovations(e.g. internet hosts per capita), diffusion of old inno-vations (e.g. telephones per capita), and human skills(e.g. mean years of schooling for people over 15 years).Similarly, the 2004 European Innovation Scoreboardemploys 20 indicators, comprised of these four groups
– human resources; the creation of new knowledge; thetransmission and application of knowledge; and innova-tion, finance, output, and markets – to develop compos-ite indices of innovation performance for EU member
ch Polic
1526 P. Shapira et al. / Researstates, the US, and Japan (Commission of the EuropeanCommunities, 2004). The ‘creation of knowledge’ indi-cators include public and business R&D/GDP and hightech patents/population. The ‘transmission and appli-cation of knowledge indicators’ include proportion ofsmall and mid-size enterprises (SMEs) that report mak-ing innovations or collaborating in innovations, innova-tion expenditures/sales, and non-technical innovationsby SMEs. The UNDP, EU, and other similar indices drawon already available data reported at the national level;national comparisons of standing are then made.
The National Innovation Systems (NIS) literatureintegrates both the management and economic traditionsto measure the interaction of institutions in affectingknowledge generation, sharing, utilization, and effects ata national level (Lundvall, 1992; Pavitt and Patel, 1994).Many recent innovation studies, including those basedon national and regional innovation surveys in Europe,North America, and elsewhere, draw on this body of lit-erature. The Oslo Manual is the OECD’s guidelines forcollecting technological innovation data (OECD, 1997).The manual identifies two major types of innovation:technological product innovation (which includes tech-nologically new products and technologically improvedproducts) and technological process innovation (new orsignificantly improved production methods). It also con-siders the category of ‘non-technological innovations’which can include organizational and management inno-vations (OECD, 1997, p. 117). Paralleling the develop-ment of the first Oslo Manual, countries of the EuropeanCommunity have developed a Community InnovationSurvey (CIS) process, through which survey question-naires that probe innovation activities and results areadministered to companies in participating Europeannations. The first CIS was administered in 1992; withmodifications, further CIS rounds were pursued in 1996and 2001 (European Communities, 2004). The Oslomanual and the European CIS have served as modelsfor innovation surveys conducted by statistical agenciesaround the world, including in Canada, Japan, and SouthAfrica. However, the Oslo approach places emphasis on‘formal’ research inputs (e.g. R&D budgets) and out-puts (e.g. publications and patents); while CIS has beenoriented to technology-based innovation (Salazar andHolbrook, 2003). Less attention has been paid to ques-tions of organizational innovation or to innovation intraditional manufacturing sectors and services (wherea variety of knowledge sources and relationships are
deployed to promote advancement). This has promoteddiscussion not only about how to broaden the Oslo andCIS concepts to reflect the contemporary realities andcomplexities of knowledge development and its associa-y 35 (2006) 1522–1537
tion with innovation, but also to develop measures whichcan be applied to developing countries where research,technological progress, and innovation do not necessar-ily follow the same paths as in developed economies(Jaramillo et al., 2001).
3.3. Design and methodological implications
The management and business literature offersinsights into knowledge determinants at the micro-level, but also methodological implications. While theknowledge accounting measures and indicators providesnapshots of the knowledge capabilities of individualfirms, they do not orient very well to aggregation. Also,many of the current business-oriented knowledge mea-sures ill-fit small organizations and do not provide abasis for inter-firm comparisons. They do not take sec-toral heterogeneity into account and lack the analyticrigor of the econometric, causal models. On the otherhand, economic models are restrictive in their usageof production–function approaches. Issues like time-lagand an overemphasis on research inputs, specificallyR&D inputs make them less applicable in their currentforms to developing country situations and service sec-tors. Besides, there is no agreement on the right proxy of‘knowledge’ to be used. The development of indices tomeasure knowledge is interesting, but such indices aregenerally available only at the national level and morefundamentally they tend to be ‘data-driven’ (using thatdata which is available across countries) rather than ‘con-ceptually driven’ (e.g. being based on a model of knowl-edge acquisition and use and relationships to innovationand economic performance). Innovation surveys remedysome of the problems prevalent in the business and eco-nomic approaches by recognizing that the relationshipsbetween knowledge and its effects are multi-faceted anddiverse. However, for our purposes, we need a greateremphasis on knowledge processes (which we can thenattempt to link to innovation outcomes).
We thus seek to advance the current state-of-the-artby developing an approach that focuses at the ‘meso’level of industrial sectors. In contrast to the ‘macro’national level or the ‘micro’ company level, this industrysector orientation provided the opportunity for a studyapproach that could offer far more Malaysian industrydetail and comparison than previous national measure-ment efforts, but with a level of generalizability greaterthan that offered by a sample of corporate case stud-
ies. We also seek to conceptualize and operationalizemeasures of the constituent elements of knowledge andto understand relationships to a variety of outcomes,including different types of innovation.
ch Polic
4a
tams“eostcfttawinptath
P. Shapira et al. / Resear
. Knowledge inputs and innovation outcomes:n organizing framework
The central objective of our study is to develop andest a conceptual framework that measures knowledgend its relationships to innovation and business perfor-ance. For the framework (and drawing on our synthe-
is of the literature) we define knowledge content asthe sum of human capabilities, leadership assets andxperience, technology and information capital, collab-rative relationships, intellectual property, informationtocks, and capabilities for shared learning and utiliza-ion that can be used to create wealth and foster economicompetitiveness”. We then sought to operationalize theramework (and this definition). As a design parame-er, we recognized that the framework operationaliza-ion, including measures of knowledge inputs, linkages,nd outcomes, should be general and flexible in thathile emphasizing industrial sector knowledge action,
t should transition well to firm level measurements andational level (policy) considerations. Our framework isresented in Fig. 1. It will be seen that we attempt to dis-
inguish knowledge ‘enablers’ from knowledge ‘actions’nd ‘outcomes’. It also allows hypothesizing the direc-ion and nature of interaction between various factors andence serves as a guide to causal/predictive modeling.Fig. 1. Conceptualization of knowledge content
y 35 (2006) 1522–1537 1527
This section explains the components of the frameworkand our underlying logic.
4.1. Knowledge components
Knowledge enabling components can be viewed asinput stock variables. These inputs along with exter-nal factors like overall economic environment, marketand industry structure feed into knowledge processingand outcome processes. External factors also influencehow the application of new knowledge by means of newproducts, processes, and organizational changes trans-lates into enhanced business performance for the firms.In our model (as shown in Fig. 1), we further decomposeknowledge stock and action variables as follows.
4.1.1. Knowledge enablers or stocksKnowledge enablers can be distinguished along
three dimensions. Human capabilities measure how anorganization’s people are suited to participate actively inknowledge activities (e.g. literacy, creativity, familiaritywith pertinent IT, requisite skills). Knowledge leader-
ship gauges the extent to which management championsknowledge-driven efforts (e.g. top-level managerscommitted to knowledge-based competitiveness).Technology/infostructures assesses the use of advancedcomponents and knowledge outcomes.
ch Polic
1528 P. Shapira et al. / Researtechnology systems to transmit, receive or applyknowledge. For example, it measures the availabilityof sufficient computing and networking to carry outknowledge oriented work suitable to the organization’smission and current capabilities. Knowledge environ-ment refers to external aspects (sectoral knowledgeactors, policies, sector structure and dynamics, culturalorientation) that influence actions at the enterprise level.
4.1.2. Knowledge processing (actions/flows)Knowledge processing highlights the significance of
generating knowledge and putting that knowledge touse—a progression which is built on, and contributes toknowledge capabilities. Knowledge generation refers tothe creation or production of new knowledge. This couldbe through R&D, process learning, or other mechanisms.Knowledge acquisition includes internal, but especiallyexternal, information compilation. For instance, theestablishment gains access to external databases and con-sultant expertise. It also addresses the degree to which theestablishment (or sector) makes information and knowl-edge available to its would-be users. Users could includethose within the organization and those outside (e.g.customers). Elements include centralized access, organi-zation, searchability, and user-friendliness. Knowledgesharing concerns the extent to which knowledge is shared(or transferred), both via electronic information mecha-nisms and by people (e.g. by teaming). Organizationalstructures and mechanisms to encourage sharing areposited as vital (e.g. incentives). Knowledge utilizationdirects attention to the ways and extent to which knowl-edge is brought to bear in establishment practices (e.g.design, marketing) and decision making. Diverse fields– such as management science and operations research,policy analysis and research evaluation, and statistics andpatent analysis – lament the degree to which the knowl-edge they generate is underutilized.
4.2. Knowledge outcomes
Effective change is the target outcome of knowl-edge efforts, and two major sets of outcomes can beclarified. Firstly, innovation reflects the application ofknowledge and creativity toward new products, pro-cesses, services, and organizations. This category can beregarded as including ‘top-line’ measures of knowledge-driven outcomes. Economic performance directs atten-tion to those aspects of economic performance related
to knowledge enhancement. Core measures may includeincreased sales or profit, gains in market share, improvedvalue-added, or changes in wage levels. This categorycan be regarded as including ‘bottom-line’ measuresy 35 (2006) 1522–1537
of knowledge-driven outcomes. However, we must notethat knowledge-based economic performance is clearlynot independent of general economic influences andother business strategies.
Nevertheless, these links may be indirect and may besubject to time lags. Also, while the framework empha-sizes an industrial orientation, it notes that many of thefactors relevant at the firm and sector level are influencedby public policy. For example, the public sector is themajor player in most aspects of education and training.Government is a major creator of certain key informa-tion and the compiler in some cases. For instance, inMalaysia, the Ministry of Agriculture proactively worksto engage various farmers and other agricultural enter-prises in using information they provide. This involveswebsite access, provision of training seminars, extensionservices, and so forth. Knowledge incentives can make adifference in adoption of knowledge economy objectivesand speeding investments and commitments to attainthem. A powerful innovation has been the creation of theMultimedia Super Corridor and attendant organizations,incentives, and laws to promote knowledge economyindustry and education. Our framework hence seeks toinclude measures that provide policy-rich information onpublic sector contributions to sector knowledge content.
5. Methodology
The MyKe study (2002–2004) operationalized theknowledge measurement model described in the pre-vious section using three exploratory data collectionapproaches: (1) direct measurement through a nation-wide knowledge content survey of more than 1800 man-ufacturing and services firms in Malaysia linked withannual statistical surveys; (2) data mining of knowledge-intensive outputs (research publications, patents, trade-marks) in Malaysia; and (3) Malaysian industry expertinterviews.
Of these three methods, the principal source of datafor the present paper is the nationwide survey (MalaysianKnowledge Content Survey, 2002). This survey targetedprivate firms in 18 manufacturing and services industriesin Malaysia (Table 1). Firms in targeted sectors wereselected from the Malaysian Department of Statisticsmaster database of all private manufacturing and ser-vice establishments in Malaysia using a random stratifiedapproach (based on sector and size class) to representthe composition of the 18 sectors. Firms with 20 or more
employees in manufacturing industries and 10 or moreemployees in services industries were targeted. Smalland medium-size firms were defined to include thosewith 20–99 employees in manufacturing and 10–99 in
P. Shapira et al. / Research Polic
Table 1Industries targeted in the Malaysian Knowledge Content Study
Sector Abbreviation
Manufacturing1. Food processing Food2. Chemicals, petroleum, and pharmaceuticals Chem3. Rubber and plastic products Rubb4. Wood-based products Wood5. Fabricated metals Fab6. Automotive Auto7. Transport equipment Tranm8. Textile, wearing apparel, and footwear Text9. Electrical and electronics EE10. Machinery and instruments Mach
Services11. Tertiary education services Educ12. Transportation services (ports, airports,
shipping)Trans
13. Finance services (Head Offices) Fin14. Selected tourism services Tour15. Telecommunications and courier services
(Head Offices)Tele
s1vmoATsr(ntlDwascp
e1e
St(
16. Health services Hlth17. Information technology services IT18. Selected business services Bus
ervices. Large firms were defined to include those with00 or more employees in both manufacturing and ser-ices. The survey was administered in 2003 to senioranagers of selected firms by the Malaysian Department
f Statistics (DOS) in conjunction with the Malaysiannnual Survey of Manufacturing and Services Sectors.3
his ensured a consistent survey approach and a highurvey response rate. The survey goal was to obtainesponses from 1669 firms, based on individual cellsector and size) sampling targets. Assumptions abouton-response and firms that might be below the sizehreshold or out-of-business at the time of the surveyed to an over-sampling approach targeting 2207 firms.OS obtained completed responses from 1819 firms,hich was in excess of the original goal and representsgross response rate of 82.4 percent based on the over-
ample frame. Proportionately, the responses achievedlosely matched the original sector and size targets. Therofile of responding firms is provided in Table 2.
Some 52 percent of those surveyed were single-
stablishment firms, 37 percent were subsidiaries and1 percent were parent or holding companies. How-ver, there were marked industry differences, with some3 The questionnaire used for the Malaysian Knowledge Contenturvey 2002 may be downloaded from the section “Survey Ques-
ionnaire” of the Department of Statistics Malaysia Web Portalhttp://www.statistics.gov.my).
y 35 (2006) 1522–1537 1529
industries having a higher number of subsidiaries orholding companies, while others were mostly singleestablishment firms. In terms of employment size, 70percent of the firms were small or medium sized. How-ever, the presence of larger firms was higher than aver-age in the automotive industry (47 percent), electricaland electronics industry (43 percent), the chemicals,petroleum, and pharmaceuticals industry (43 percent)and the finance services industry (49 percent).
The MyKe survey design drew on the formulation ofthe conceptual framework. A series of variables that hadpromise in measuring levels, changes, and trends withinthe 10 component factors of the knowledge contentmeasurement model were identified. In total, some 45variables with about 140 prospective metrics were iden-tified. From this array, it was fully expected that some ofthese variables and metrics would have greater explana-tory power than others (with perhaps a few not reallyproving that useful). To allocate variables to measureaggregate knowledge content in each of the 18 indus-tries of the study, three methodologies were explored:an a priori allocation of variable; factor analysis; and ahigh-powered variable selection based on factor analy-sis. All three methods produced very consistent resultsin terms of their measurement of knowledge content ineach of the 18 study industries. The third method (using21 high powered variables) was found to offer an opti-mal and efficient method to measure knowledge contentby the model components and by the 18 study industries.Table 3 lists and briefly describes the high-powered vari-ables used in the measurements and analyses describedin the balance of the paper (Table 3).
6. Sectoral results
Based on the MyKe survey, we found that all theindustries under review had built a foundation in knowl-edge competencies and capabilities, and embarked onsome form of knowledge acquisition, generation andprocessing activities. Fig. 2 illustrates measures of thevarious industries against the 21 variables that constitutethe knowledge enabling and knowledge action compo-nents, which together comprise aggregate knowledgecontent. Overall, the IT, chemical, telecommunications,education, and finance industries have the highest knowl-edge content, while the textiles, transport services, andwood-based industries have the lowest. All industriesdisplay significant differences in their accomplishments
but, as differences by establishment performance withinindividual industries demonstrated, many firms have thepotential to improve considerably. Key findings are dis-cussed below.
1530P.Shapira
etal./Research
Policy35
(2006)1522–1537
Table 2Profile of firms responding to the Malaysian Knowledge Content Survey
Sector Total sample Type of firms Size of firmsa,b Location of HQa,c Exportersa,d
Single Parent Subsidiary Total SMEs Large Total Domestic Foreign Total Yes No Total
Manufacturing 1118 567 109 442 1118 761 353 1114 358 191 549 658 272 930Food processing 143 65 8 70 143 94 49 143 69 9 78 51 55 106Chemicals, petroleum, and pharmaceuticals 101 36 9 56 101 58 43 101 32 33 65 69 16 85Rubber and plastic products 151 83 15 53 151 105 46 151 46 22 68 109 26 135Wood-based products 132 86 132 34 132 105 27 132 45 2 47 82 32 114Fabricated metals 152 83 15 54 152 112 39 151 43 25 68 92 39 131Automotive 76 30 9 37 76 40 35 75 35 11 46 37 15 52Transport equipment 49 31 5 13 49 40 8 48 12 4 16 18 28 46Textile, wearing apparel, and footwear 111 63 16 32 111 78 33 111 36 12 48 60 29 89Electrical and electronics 145 58 13 74 145 83 62 145 28 59 87 102 22 124Machinery and instruments 58 32 7 19 58 46 11 57 12 14 26 38 10 48
Services 701 380 89 232 701 515 184 699 274 41 311 66 311 377Tertiary education services 84 47 9 28 84 63 21 84 34 3 37 5 40 45Transportation services (ports, airports, shipping) 119 66 21 32 119 84 33 117 45 5 50 12 61 73Finance services 55 10 6 39 55 28 27 55 38 6 44 11 16 27Selected tourism services 105 61 3 41 105 77 28 105 39 4 43 5 52 57Telecommunications and courier services 39 20 5 14 39 24 15 39 16 3 19 8 6 14Health services 89 59 11 19 89 75 14 89 26 2 28 5 52 57Information technology services 69 27 14 28 69 51 18 69 29 13 42 13 27 40Selected business services 141 90 20 31 141 113 28 141 47 5 52 7 57 64
Total number of firms 1819 947 198 674 1819 1276 537 1813 632 232 864 724 S83 1307
Source: Malaysia Knowledge Content Survey, 2003.a Total number of firms did not sum up to total size of sample due to non-response.b Small- and medium-sized enterprises (SMEs) are defined as firms with less than 100 employees.c A firm is considered domestic-oriented if its headquarters is located in Malaysia and is foreign if otherwise.d A firm is considered an exporter if it exported its output in 2002.
P. Shapira et al. / Research Policy 35 (2006) 1522–1537 1531
Table 3High-powered variables used to measure knowledge content
Knowledge content component High-powered variable
E1: Human capabilities Percentage of firms with employees received training or education in 2002Mean percentage of workers with university degree
E2: Knowledge leadership Percentage of firms with a written knowledge strategyPercentage of firms with a committee leading knowledge developmentPercentage of firms certified to ISO
E3: Technology and infostructure Percentage of firms with PCs installedMean number of PC in firms per employeePercentage of firms with e-commerce
E4: Knowledge environment Percentage of firms with association membershipPercentage of firms aware of the knowledge planPercentage of firms participating in projects
K1: Knowledge generation Percentage of firms that filed a copyrightPercentage of firms that filed a patent applicationPercentage of firms engaged in R&D in 2002
K2: Information gathering Percentage of firms using formal procedures to assess employee satisfactionPercentage of firms using formal procedures to assess customer satisfaction
K3: Knowledge sharing Percentage of firms using project teamsPercentage of firms sharing knowledge electronicallyPercentage of firms sharing knowledge with other companies
K4: Knowledge utilization Percentage of firms often using experiential knowledge to perform jobsPercentage of firms using external knowledge to make improvements
O1: Innovation outcomes Percentage of firms which introduced product or process innovationPercentage of firms which introduced other managerial innovations
O2: K-driven economic performance Mean percentage of establishment’s sales in 2002 from technology new to the firmMean percentage of establishment’s sales exportedPercentage of firms selling c
Note: Based on factor analysis of data from Malaysian Knowledge Content S
Fig. 2. Aggregate knowledge content measures, by industry(Malaysia). Source: Analysis of Malaysian Knowledge Content Sur-vey, 2003. N = 1819.
onsulting services
urvey, 2003.
6.1. Knowledge enablers
All industries did fairly well in technology and infos-tructure (E3) and in knowledge environment (E4). Asexpected, education, finance, and IT industries did espe-cially well in technology and infostructure (Fig. 3). How-ever, the levels of achievement in the other enablingcomponents (E1 and E2) were less encouraging. It wasespecially discouraging with regard to human capabil-ities. The low cross-sectoral average suggests that the‘soft’ factors needed to sustain a knowledge-based econ-omy require much greater attention despite the measuresthat have already been implemented to encourage skillsdevelopment and training.
6.2. Knowledge actions
All industries performed well in knowledge utiliza-tion (K4), with almost all of them reaching or exceed-ing the cross-sectoral average. However, performance in
1532 P. Shapira et al. / Research Policy 35 (2006) 1522–1537
ource:
Fig. 3. Cross-industrial analysis by knowledge content components. Sthe other knowledge process components (K1, K2, andK3) was uneven with a significant number of industriesunderperforming the cross-sectoral average. In addition,absolute levels of attainment were generally low with
Analysis of Malaysian Knowledge Content Survey, 2003. N = 1819.
respect to knowledge generation (K1). This is substan-tiated by other studies which indicated that generally,Malaysian manufacturing firms are good adopters andadapters of technology rather than innovators.
P. Shapira et al. / Research Policy 35 (2006) 1522–1537 1533
rce: An
6
eaNiuodt
6
shofiskoSstiomi0fii
7
c
ing, electrical and electronic manufacturing, transporta-tion services, and health services. A fifth sector (businessservices) showed a positive association between value-
4 Two notions of innovation capability were measured. Specifically,firms that reported having developed technologically new or signifi-cantly improved products or processes or both are termed technologyinnovators. Firms that reported significant improvements in products,processes, internal management, organizational methods, marketingconcepts or business strategies are simply classified as innovators. Thisclassification allows us to take a broader view of innovation both as atechnical, as well as organizational/commercial endeavor and allowsus to study the nature of differential impacts due to the two. In theanalysis of the survey, firms that reported the introduction of TNSIproducts or processes in the 2000–2002 time period were coded “1”
Fig. 4. Cross-industrial analysis of knowledge-driven outputs. Sou
.3. Knowledge-driven outcomes
Firms in Malaysia, particularly in the IT, finance, andducation industries have been active in product, processnd managerial innovations (O1), as shown in Fig. 4.evertheless, these innovations have yet to be translated
nto higher output as measured by the proportion of prod-ct sales deriving from innovations, and the proportionf exports to total sales (O2). It is possible that the studyid not allow for an adequate time lag between innova-ion and economic performance.
.4. Knowledge gaps
The MyKe study also enables inter-firm and intra-ector knowledge gaps analyses. In general, large firmsave higher measures of knowledge content than smallernes. The median knowledge content measure for largerms is 10.3 compared to 7.3 for small- and medium-ized enterprises (SMEs). In other words, the mediannowledge content measure for SMEs was about 0.7f that of large firms. However, the top 5 percent ofMEs have profiles of knowledge and technology usageimilar to the top 5 percent of large firms, suggestinghe current emergence of a small cluster of knowledge-ntensive SMEs in Malaysia. By ownership, foreign-wned firms scored a median sector knowledge contenteasure of 10.8 compared to 8.1 for domestic firms, giv-
ng a domestic–foreign knowledge content gap ratio of.8. This is attributed to larger, mostly foreign-ownedrms undertaking R&D and providing employee train-
ng more readily.
. Knowledge content and innovation
We also examined how elements of the knowledgeontent measurement model impacted innovation per-
alysis of Malaysian Knowledge Content Survey, 2003. N = 1819.
formance. Defining technology innovation as the intro-duction of technologically new or significantly improvedproducts, goods, services, or processes (TNSI), ouranalysis of the MyKe survey finds that slightly morethan 40 percent of the firms introduced technology inno-vation during the period 2000–2002. The percentage oftechnology innovators ranged from a low of 18 percent(wood-based sector) to a high of 72 percent (IT services).A broader definition of innovation that included any tech-nical organizational improvement had less variability.4
Using standard logistic regression modeling, positiveassociations between technological innovation and atleast one knowledge content variable were evidentacross all but five industries (See Table 4). Linkage withexternal knowledge sources was particularly importantin explaining variations in technology innovation. Fiveindustries had significant positive associations betweentechnology innovation and value-added per employee:fabricated metals production, automotive manufactur-
on the technology innovation variable and a negative response regis-tered a coding of “0”. Similarly firms that reported any innovation asper our more-inclusive definition were coded “1” on the innovationvariable and negative responses were coded as “0”. These serve as thebinary dependent variables in the set of models that were estimated.
1534 P. Shapira et al. / Research Policy 35 (2006) 1522–1537
Table 4Logistic regressions of innovation with knowledge content variables, by industry
Industry Knowledge content variable Coefficient Model goodness-of-fit
Food E-commerce 2.77 R2 = 0.63PCs installed 2.39 86% correctly predictedPatent application 1.82Employee training 1.35Knowledge strategy 1.33External knowledge sources 0.73Managers with 10 or more years of experience −0.01
Chemicals Cooperation with industry, university and government 4.15 R2 = 0.71Patent application 2.43 89% correctly predictedExternal knowledge sources 1.64Employee satisfaction 1.61Managers with 10 or more years of experience −0.03
Rubber R&D 1.06 R2 = 0.31External knowledge sources 0.31 72% correctly predicted
Wood-based Project teams 3.20 R2 = 0.52Electronic knowledge sharing 2.93 85% correctly predictedExternal knowledge sources 0.63
Fabricated metals Employees with science or technical degrees 9.91 R2 = 0.63External knowledge sources 1.09 84% correctly predictedKnowledge development committee 1.74R&D 1.56Managers with 10 or more years of experience −0.02
Automotive PC per employee 8.51 R2 = 0.67External knowledge sources 2.51 88% correctly predictedManagers with 10 or more years of experience −0.03
Textiles, apparels etc. PC per employee 10.10 R2 = 0.59E-commerce 1.72 84% correctly predictedElectronic knowledge sharing 1.49External knowledge sources 0.47
Electrical and electronics Electronic knowledge sharing 1.85 R2 = 0.53Knowledge sharing with other companies 1.69 78% correctly predictedR&D 1.55Knowledge strategy 1.49External knowledge sources 0.61
Education Employees with science or technical degrees 7.61 R2 = 0.79PCs installed 7.46 93% correctly predictedR&D 5.49Focus on quality and customization 3.63Customer satisfaction 3.41Knowledge development committee 2.22Project teams 2.09External knowledge sources 0.66
Transport services Off-the-job training 1.45 R2 = 0.56Electronic knowledge sharing 1.37 83% correctly predictedExternal knowledge sources 0.48
Finance services Customer satisfaction 5.34 R2 = 0.66ISO certification 4.47 85% correctly predictedKnowledge sharing with other companies 4.37Focus on quality and customization 3.24PC per employee 2.02External knowledge sources 1.29
Tourism services Electronic knowledge sharing 1.05 R2 = 0.37External knowledge sources 0.28 76% correctly predicted
P. Shapira et al. / Research Policy 35 (2006) 1522–1537 1535
Table 4 (Continued )
Industry Knowledge content variable Coefficient Model goodness-of-fit
Health services Project teams 3.29 R2 = 0.70Focus on quality and customization 3.25 87% correctly predictedKnowledge development committee 2.80Managers with 10 or more years of experience −0.02
Transport equipmentMachineryTelecommunications Not significantInformation technologyBusiness services
N dge Conv r signifi2 ificance
a(bppttmcTovmi
8
fecseoikc
ulgl
im
ote: Logistic regression analysis of data from the Malaysian Knowleariable and is defined as the introduction of technologically new o000–2002 time period. Results presented are based on statistical sign
dded per employee and a broader measure of innovationwhich takes into account organizational, marketing, andusiness strategy innovations as well as new products orrocesses). Technology innovation was significantly andositively associated with exporting in only three indus-ries: fabricated metals production, electrical and elec-ronic manufacturing, and machinery and instruments
anufacturing industries. Profitability was not signifi-antly related to innovation except for business services.hese results alone do not definitively confirm or ruleut relationships between innovation and outcomeariables. Nonetheless, they suggest that opportunitiesay exist to enhance returns from knowledge-based
nnovation in many Malaysian industries.
. Conclusions
The knowledge content model presents a frameworkor tracking the extent to which various forms of knowl-dge processes and practices are diffused throughoutompanies in key sectors of Malaysia’s economy. Ituggests that four enabling factors and the four knowl-dge actions are leading indicators of knowledge-drivenutcomes. In fact, it has found that near-term changesn these factors did not uniformly predict changes innowledge-driven outcomes, but did so selectively inertain industries.
The paper summarizes the methodological approach
sed to measure content based on the model and high-ights key findings.5 Industries with high levels of aggre-ate knowledge content are identified. Industries withower levels of aggregate knowledge content are also5 For further information about the Study on Knowledge Contentn Key Economic Sectors in Malaysia, including links to additional
aterials and publications, see http://www.cherry.gatech.edu/myke.
tent Survey, 2003. N = 1819. Technology innovation is the dependentcantly improved products, goods, services, or processes during theof association of at least 10 percent.
noted. However, we find that industries vary by spe-cific knowledge content components, reflecting differ-ences in industrial characteristics and establishmentbusiness strategies. For example, petroleum processingand finance were industries that focused on developinghuman capabilities through training and education, whileIT services firms were more likely to hire employees withprior high capability levels (as measured by employeesholding university degrees.) On the whole, IT serviceswere most likely to be a leading sector across the rangeof knowledge content measures. At the same time, thestudy found that every sector in the Malaysian economyhas a group of leading firms whose adoption of knowl-edge practices and technology significantly exceeds thatof more typical firms. Therefore, raising as many compa-nies as possible closer to current best-practice levels ofknowledge content and use within Malaysia could signif-icantly strengthen competitiveness, particularly in thoseindustries with relatively wider discrepancies betweentop and median users. Specific policy measures thatcould close knowledge gaps include improved technol-ogy transfer and industrial extension services. Such poli-cies would be most effective in the context of generalimprovements in the framework and policies for edu-cation and human resource development and attentionto strengthening innovation culture, opportunities andincentives in Malaysia.
The approach pursued in the paper is exploratory,but it does confirm that it is possible to develop mea-sures of knowledge content that can be tracked on anongoing basis to inform policy making and to assessprogress in certain knowledge economy goals. The high-powered variables that we identified provide the basis
for future monitoring in Malaysia (and perhaps else-where), although we recognize that other variables mayneed to be introduced, particularly to account for ongo-ing changes in the character of knowledge development.
ch Polic
1536 P. Shapira et al. / ResearOur study also confirms the value of pursuing a sector-level approach which is sufficiently fine-grained to alsoidentify trends by different establishment types so as toprovide the basis for detailed analyses on which to assessnational policies and to develop targeted sector measuresand strategies.
Acknowledgements
We gratefully acknowledge the contributions of col-leagues to various parts of the Study on KnowledgeContent in Key Economic Sectors in Malaysia (MyKe)on which this paper draws, including Alan Porter, NilsNewman, Jue Wang, Deepak Hegde, David Roessner,and staff at the Economic Planning Unit of Malaysiaand the Malaysian Department of Statistics. The sup-port of the United Nations Development Program forthe MyKe study is also much appreciated. An ear-lier version of this paper was presented at the 2005Triple Helix 5 Conference, Panel on New Indicators forthe Knowledge Economy, Turin, Italy, May 2005. Theauthors also benefited from anonymous review com-ments and editing suggestions made by Erin Lamos.It should be noted that the judgments expressed inthis paper are those of the authors and should not beattributed to their organizations or the sponsors of theresearch.
References
Abramowitz, M., 1956. Resource and output trends in the United Statessince 1870. American Economic Review 46, 5–23.
Boudreau, J.W., 2002. Strategic knowledge measurement and man-agement. Working Paper 02-17. Center for Advanced HumanResources, School of Industrial and Labor Relations, Cornell Uni-versity, Ithaca, New York.
Commission of the European Communities, 2004. European Inno-vation Scoreboard 2004: Comparative Analysis of Innova-tion Performance. Enterprise and Industry Directorate Gen-eral, Innovation Policy Development Unit, Brussels, Belgium(http://trendchart.cordis.lu/).
Cooke, P., 2002. Knowledge Economies. Routledge, London, UK.Department of Statistics Malaysia, 2005a. Gross Domestic Product at
Constant Prices, First Quarter 2005. Putrajaya, Malaysia.Department of Statistics Malaysia, 2005b. Key Statistics (External
Trade and Major Exports), 2004 (provision). Putrajaya, Malaysia.Earl, L., 2002. Are We Managing Our Knowledge? Results from the
Pilot Knowledge Management Practices Survey, 2001. Cat. No.88F0006XIE02006. Statistics Canada, Ottawa.
Economic Planning Unit, 2001. The Third Outline Perspective Plan2001–2010. Prime Minister’s Department, Putrajaya, Malaysia.
Economist Intelligence Unit, April 2004. Country Profile: Malaysia—Economic Structure (http://www.economist.com/countries/Malaysia/).
European Communities, 2004. Innovation in Europe: Results for theEU, Iceland, and Norway. European Commission and Eurostat,
y 35 (2006) 1522–1537
Office for Official Publications of the European Communities, Lux-embourg.
Griliches, Z., 1990. Patent statistics as economic indicators: a survey.Journal of Economic Literature 284, 1661–1707.
Griliches, Z., 1992. The search for R&D spillovers. Scandinavian Jour-nal of Economics 94, 29–47.
Grupp, H., Mogee, M.E., 2004. Indicators for national science andtechnology policy: how robust are composite indicators. ResearchPolicy 33, 1373–1384.
ISIS, 2002. Knowledge-Based Economy Master Plan. Institute ofStrategic and International Studies, Kuala Lumpur, Malaysia.
Jaramillo, H., Lugones, G., Salazar, M., 2001. Standardisation ofIndicators of Technological Innovation in Latin American andCaribbean Countries (Bogota Manual). Iberoamerican Networkof Science and Technology Indicators. Organisation of AmericanStates.
Kermally, S., 2002. Effective Knowledge Management: A Best Prac-tice Blueprint. John Wiley and Sons, Chichester, UK.
Leonard, D., Sensiper, S., 1998. The role of tacit knowledge ingroup innovation. California Management Review 40 (3), 112–132.
Lev, B., 2001. Intangibles: Management, Measurement, and Reporting.Brookings Institute, Washington, DC.
Lundvall, B.-A. (Ed.), 1992. National Systems of Innovation: Towardsa Theory of Innovation and Interactive Learning. Pinter Press.
Mani, S., 2001. Government Innovation and Technology Policy: anInternational Comparative Analysis. Edward Elgar, Cheltenham,UK (Chapter 5: The Malaysian Experience).
Nonaka, I., Takeuchi, H., 1995. The Knowledge-Creating Company.Oxford University Press, Oxford, UK.
OECD, 1996. The Knowledge-Based Economy. Organisationfor Economic Cooperation and Development, Paris, France(OECD/GD(96)102).
OECD, 1997. The Measurement of Scientific and Technological Activ-ities. Proposed Guidelines for Collecting and Interpreting Techno-logical Innovation Data (Oslo Manual). European Commission andEurostat.
OECD, 2001. Competencies for the knowledge economy. In: Educa-tion Policy Analysis. Centre for Educational Research and Inno-vation, Organisation for Economic Cooperation and Development,Paris, France (Chapter 4).
OECD, 2004. Measuring Knowledge Management in the BusinessSector: First Steps. OECD Publication 962003021E1. Organisa-tion for Economic Cooperation and Development and Statistics,Canada.
Polanyi, M., 1967. The Tacit Dimension. Routledge and Keegan Paul,London.
Pavitt, K., Patel, P., 1994. National innovation systems: why they areimportant, and how they might be measured and compared. Eco-nomics of Innovation and New Technology 31, 77–95.
Sahibbudin, S., 2001. Technology deployment challenges in Malaysia:minimising the digital divide. In: Abdulai, D.N. (Ed.), Malaysia andthe K-Economy. Pelanduk Publications, Subang Jaya, Malaysia.
Salazar, M., Holbrook, J.A.D., 2003. A debate on innovation surveys.In: Paper Presented at the Conference in Honor of Keith Pavitt:What Do We Know About Innovation?, Science and TechnologyPolicy Research Unit (SPRU), Sussex University, UK.
Senker, J., 1995. Tacit knowledge and models of innovation. Industrialand Corporate Change 4 (2), 425–447.
Solow, R.M., 1957. Technical change and the aggregate productionfunction. Review of Economics and Statistics 39, 214–231.
Stewart, T., 1997. Intellectual Capital. Doubleday, New York.
ch Polic
S
S
UNDP, 2001. Human Development Report 2001: Making New Tech-
P. Shapira et al. / Resear
veiby, K.E., 1997. New Organizational Wealth: Managing and Mea-
suring Knowledge-based Assets. Berrett-Koehler Publishers, SanFrancisco.veiby, K.E., 1998. Measuring Intangibles and IntellectualCapital—An Emerging First Standard (http://www.sveiby.com/articles/EmergingStandard.html).
y 35 (2006) 1522–1537 1537
nologies Work for Human Development. United Nations Develop-ment Programme, Oxford University Press, New York, USA, andOxford, UK.