Technology Infrastructure

Cristiano Antonelli,Albert N. Link,Stan Metcalfe

• Editore: Routledge
• Anno: 2013
• ISBN:
  9781317990567
• Edizione: 1
• Argomento: Negocios y empresa
• Categoria: Negocios en general

Technology infrastructure supports the design, deployment and use of both individual technology-based components and the systems of such components that form the knowledge-based economy. As such, it plays a central role in the innovation process and in the promotion of the diffusion of technologies. Thus, it is an important element contributing to the operation of innovation systems and innovation performance in any modern economy.

Technology infrastructure, either in the narrow or broad sense, is not well understood as an element of a sector's technology platform or of a national innovation system. Similarly misunderstood are the processes by which such infrastructure is embodied in standards or diffused through various institutional frameworks. In fact, because of the public and quasi-public good nature of technology infrastructure, firms as well as public-sector agencies under invest in it, thus inhibiting long-term technological advancement and economic growth.

This volume of essays brings together a collection of papers from eminent scholars on all of the various dimensions of technology infrastructure mentioned above. To our knowledge, it is the first such collection of papers and we expect this scholarship to become the foundation for future research in this area.

This book was published as a special issue of Economics of Innovation and New Technology.

MODELING AND MEASURING THE ECONOMIC ROLES OF TECHNOLOGY INFRASTRUCTURE

GREGORY TASSEY

National Institute of Standards and Technology, Gaithersburg, MD, USA

The first part of this paper presents a disaggregated or ‘multi-element’ model of technological change. Such a model allows examination of the roles and impacts of the major elements of technology, each of which is distinguished by a different degree and type of public-good content. This distinction implies unique investment behavior with respect to each element with consequent public-policy implications. The second part draws upon the considerable experience of the US National Institute of Standards and Technology in designing and conducting practical approaches to estimate the economic benefits from public and private investment in these quasi-public-good technology elements.

1 DISAGGREGATING THE KNOWLEDGE PRODUCTION FUNCTION

The typical industrial technology is composed of three elements: the generic technology base (also, technology platform); supporting infratechnologies; and proprietary market applications (innovations and subsequent improvements). The first two have public-good content and therefore, embody infrastructure characteristics. These critical quasi-public technology goods are supplied by a combination of firm-specific assets and sources external to the innovating unit of the firm, such as central corporate research labs, government labs, and increasingly, universities. The fundamental relationships among these elements require a technology production function that captures the interactive nature of the two quasi-public-good elements with each other and with private-sector investments in the third element, proprietary technologies. Most important, each element responds to different sets of investment incentives (Tassey, 2005a).

The failure to disaggregate the technology variable based on the distinctly different character of each element and its associated unique investment incentives has limited economists’ ability to explain R&D investment behavior and the subsequent relationships with economic growth. Both macroeconomic and microeconomic growth models have made technology an endogenous explanatory variable. However, the vast majority of this literature has treated technology and the process that creates it, R&D, as homogeneous entities. Only a few efforts have attempted even a partial disaggregation, and those have been limited to separating scientific research from technology research. In other words, the technology variable remains aggregated.

This failure has also inhibited government technology investment policies by prohibiting assessments of the distinctly different incentives associated with each of these three elements. This policy analysis problem is becoming more severe for several reasons: (1) corporate laboratories have reduced their share of national spending on the quasi-public elements, in particular, early-phase research on new, radical technologies; (2) in many countries, such as the USA, government spending on such research has been erratic and skewed toward a few technologies tied to specific social objectives; and (3) universities in many economies are assuming a larger role in such early-phase technology research, with implications for intellectual property (IP) and research portfolio management.

1.1 The Three Elements of Industrial Technology

The enabling role of generic technologies for the development of market applications (innovations) has been discussed qualitatively (Link and Tassey, 1987; Nelson, 1992; Tassey, 1997, 2007).¹ Dosi (1982, 1988) defines a ‘technology paradigm’ as a ‘pattern’ of solutions to selected techno-economic problems based on highly selected principles derived from the natural sciences. Such ‘highly selected principles’ form a generic technology base from which market applications are drawn. A generic technology provides in essence a ‘proof of concept’ that reduces technical risk sufficiently to enable applied R&D investments to be rationalized.²

Infratechnologies are the other quasi-public technology element. They include research tools (measurement and test methods), scientific and engineering data, the technical basis for interface standards, quality control techniques, etc. Collectively, they constitute a diverse technical infrastructure, various types of which are applied at each stage of economic activity. Infratechnologies often are implemented as industry standards (Tassey, 1997, 2000).³

Both generic technology and infratechnology elements are drawn upon by competing firms to create proprietary technology. However, although attainment of partial property rights is possible, spillovers and other sources of market failure are prominent. In fact, widespread use of generic technologies is desirable from a public-policy perspective because the more firms draw upon a technology platform, the larger the number and variety of innovations produced. When infratechnologies are adopted as the technical basis for standards, uniform as well as widespread use is mandatory. These characteristics result in various degrees of underinvestment across technologies and over each technology’s life cycle. Consequently, every industrialized nation provides funds to leverage generic technology and infratechnology research and subsequent assimilation by domestic industries. Such funding policies constitute recognition of the public-good content, even though identifying and measuring this content remains difficult conceptually and empirically.

1.2 The Multi-element Knowledge Production Function

The microeconomics literature has partially recognized the need for a disaggregated technology framework to address these phenomena but has not progressed beyond a dichotomous model in which technology is separated into scientific and technological stocks of knowledge. In such models, scientific information is appropriately characterized as a pure public good (Nelson, 1959) with external (to the industry) sources of supply. However, in such models, technological knowledge is implicitly assumed to be a purely private good, even while acknowledging the existence of spillovers.⁴

The following disaggregated knowledge production function separately specifies the key public and private technology elements and thereby allows the explicit representation of the critical elements of an industrial technology, specifically generic technologies and infratechnologies. Such an investment-based model of innovation allows assessment of the productivity of private-sector applied R&D, as determined by both private and public-sector expenditures that precede or concurrently support it.⁵

As a point of departure for explicitly separating the proprietary technology element from the quasi-public-good elements, the following generalized model is used:

where Q is a firm’s output of technology-based goods and services. KN represents the non-excludable (and hence public-good) portion of the industry’s generic technology and is assumed equally available to all firms in the industry. X is a set of factors that affect output/performance in addition to the public and private technology elements. ϕ represents the innovation infrastructure of the industry, which consists of a set of infratechnologies and associated standards, as well as other infrastructure elements such as the availability of risk capital, IP laws, technical support for entrepreneurs, etc. This infrastructure affects the efficiency of production and commercialization. S is the science upon which the industry’s generic technology is based. Because, the vast majority of science is developed outside the industry by universities and government research institutes and because major breakthroughs in science occur infrequently, the science base is considered to be externally determined and constant and therefore is entered in the model as a shift parameter.

KE_i is a firm’s stock of excludable (proprietary) knowledge that is used to create new products and services, i.e., innovations. At any point in time, a firm’s proprietary technical knowledge creation is equivalent to the growth in KE_i, represented by

where RE is applied R&D expenditures targeted at developing innovations, λ is a scale parameter, and δ is a firm’s R&D productivity factor.⁶

The productivity factor is represented by

An important point from a policy perspective is the negative sign on KN. It implies a hurdle for investment in innovations, specifically an initial technical-risk barrier that must be overcome before substantial private investment in RE will be forthcoming. The negative sign may seem to be counter intuitive because generic technology does in fact enable the conduct of applied R&D, which in turn produces innovations. However, it is a barrier to applied R&D in the sense that: (1) it must be available for innovation effort to occur; and (2) on average, the greater the potential of a new technology, the greater the required advance in early-phase proof-of-concept research, i.e., the greater the initial barrier to innovative effort posed by the needed investment in the generic technology.

η_j is an efficiency parameter that represents the portion of an industry’s technical infrastructure that supports knowledge production. This infrastructure is the collective effect of an industry’s (or supply chain’s) infratechnologies and associated standards that affect R&D efficiency. For example, the development and characterization of biomarkers and the ability to detect and interpret them in the human body greatly increases the productivity of biotechnology R&D. Similarly, the ability to accurately image biological activity and transmit the results for analysis also increases R&D efficiency. In general, the availability of such techniques increases potential economic benefits from inventive activity and thereby provides incentives to create proprietary technical knowledge. Such technical infrastructure only changes occasionally (i.e., slower than proprietary technologies). Moreover, because of their large public-good content, they often become industry standards, which themselves are only changed periodically. They, therefore, can be considered constant relative to the firm’s R&D investment aimed at invention and then innovation (RE in Eq. (3)).⁷ Thus, η_j is assumed to be a process constant over a technology life cycle in industry j.

The above model implies that industries based on radically new technologies require larger initial generic technology research expenditures. They will therefore experience lower rates of technical knowledge production for a given level of private R&D expenditures for some time. This phenomenon helps explain the S-shaped growth curve that characterizes the typical technology life cycle. In particular, a ‘risk spike’ is created by the need for investment early in the R&D cycle in a technology platform (generic technology) that enables subsequent innovation; that is, its existence blocks private investment in innovation early in the life cycle (Tassey, 2005a, b, 2007).⁸

In this early phase of the technology life cycle when the generic technology is immature, initial attempts at innovation through applied R&D typically fail miserably.

FIGURE 1 Risk reduction in the R&D cycle.

The exponential function in Eq. (3) is, in effect, a measure of the risk faced by investors at different points in the R&D cycle. When the targeted technological advance is large, as is the case for a radically new technology, the risk is also high that expenditures for developing innovations (through expanding KE) will fail. That is, the hump or risk spike in Figure 1 will be larger than for investment in new but less advanced technologies (for example, a next-generation generic technology, as opposed to one based on new scientific principles).

In all cases, investment in expanding the generic technology base is required to overcome the risk spike, RS, and allow private investments in KE to proceed. Such a risk profile explains why rates of innovation based on emerging technologies can languish for years, even decades. However, once the risk spike is overcome, private investment in R&D can reduce private risk, RP, to levels that permit commercialization.⁹

The extreme case is no generic technology (KN = 0). Under this condition, applied R&D has very low productivity and will likely not be attempted. Growth in the stock of technical knowledge and hence the rate of innovation is then determined by δ = η. This case could be called the ‘natural rate of innovation’ because it is driven solely by the general economic environment included in η. Such inventions fall into ‘Pasteur’s quadrant’; that is, inventions that occur through trial-and-error or ‘inspiration’ processes.¹⁰ This source of invention is increasingly rare for today’s science-driven and compl...

Indice dei contenuti

• Cover
• Half Title
• Title Page
• Copyright Page
• Contents
• Technology Infrastructure: Introduction to the Special Issue
• 1 Modeling and Measuring the Economic Roles of Technology Infrastructure
• 2 Public Technology Infrastructure, R&D Sourcing, and Research Joint Ventures
• 3 Public–Private Partnership to Develop Technology Infrastructure: a Case Study of the Economic Returns of DNA Diagnostics
• 4 Research Networks as Infrastructure for Knowledge Diffusion in European Regions
• 5 Intelligent Machine Technology and Productivity Growth
• 6 To Admit or Not To Admit: the Question of Research Park Size
• 7 Innovation Platforms and the Governance of Knowledge: Evidence from Italy and the UK
• 8 Assessing the Relative Performance of University Technology Transfer in the US and UK: a Stochastic Distance Function Approach
• 9 Placing Innovation: an Approach to Identifying Emergent Technological Activity
• 10 Barriers to the Diffusion of Nanotechnology
• Index