Because this paper will be of interest for policy makers and other analysts who are not necessarily economists, we believe it may be useful to have some kind of terminological discipline, and describe the most important technical terms employed in this paper.

Beneficial effects of Innovation

The concept of innovation comprises both technological innovation (creation of novel products or new production processes ¹³ by means of R&D effort or other forms of creative effort) as well as organizational (or managerial) innovation. Novel products are sold to others – either to other enterprises or to consumers. Normally, if a novel product is a producer good, it should serve to improve either the efficiency or the quality of output in the buying sectors, while if it is a consumer good, it should presumably enhance the quality of life directly. A new process is a technical improvement in one’s own production methods and can have productivity effects by reducing input prices or consumer good prices. Organizational innovations are changes to the business’ strategies, structures or routines which aim to improve the performance of the business.

Some innovations may have only a direct impact on growth (the creation of digital games), while others may also have indirect repercussions through productivity-enhancing effects (new technologies that reduce input prices) or through efficiency-enhancing effects (new software labor-saving products). These beneficial effects can happen either within the sector where the innovation occurs or across sectors or at the consumer level. Furthermore, the beneficial effects of an innovation can be either transmitted through the market mechanism, and thereby, the beneficial effect is paid for, or can percolate through the economy without full compensation due, for example, to reverse engineering.

It is worthwhile mentioning here that an R&D project ends when the innovation has been materialized, that is, when the work is no longer experimental and pre-production begins. Even when a research endeavor turns out to be unsuccessful (and by definition, there is no innovation), an R&D project may yield valuable information, and the returns to this kind of knowledge can often not be excluded. For example, a drug that fails to obtain the FDA approval may leave fruitful insights for new drugs, but the unapproved drug cannot be considered as an innovation from the economic viewpoint.

Splitting Innovation Spillovers

In the most recent line of empirical research technological change is also considered as the result of the existence of both intentional investments in R&D and R&D spillovers. Estimates of innovation spillovers startedwith Griliches (1958), and today we have at our disposal a number of papers where estimations of the magnitudes of R&D spillovers can be found. These include Adams (1990), and Bernstein and Nadiri (1988), (1989). The measurements of spillovers suggest that there are substantial intra- and interindustry R&D spillovers. Notwithstanding, the econometric evidence is plagued with unresolved questions concerning the existing estimates ¹⁴.

It is generally agreed that the term ‘R&D spillovers’ refers to the fact that firms undertaking R&D are unable to appropriate all of the benefits from their R&D activities. This standard characterization of the notion of spillovers may be termed weak (or catchall) definition of R&D spillovers, because it implies that the spillover effects occur through two quite distinct channels. One is “knowledge spillovers”, which refer to the effect of R&D performed in one firm in improving technology in a second firm without compensation for the former. The other refers to inputs purchased by one (‘beneficiary’) firm that embody efficiency-enhancing attributes and quality improvements, and these beneficial effects are not fully appropriated by the selling (or ‘source’) firm.

It should be emphasized that the weak definition of R&D spillovers encompasses two different mechanisms of transmission: while knowledge spillovers are not transmitted through the market mechanism, the inputs purchased by the beneficiary firms are obviously transacted in the market place. Or, to put it differently, the weak definition encapsulates both technological externalities (knowledge spillovers) and pecuniary externalities (efficiency gains through the acquisition of novel products) ¹⁵.

It should also be emphasized that the second channel postulates that the source firm is, at least to some extent, “unfairly” treated by the market mechanism because it provides benefits to other firms without getting the corresponding piece of the action. This prompts the question: why are source firms willing to do a “favor” to beneficiary firms? The answer may well be that source firms introduce new inputs to gain competitive advantage and are totally satisfied with the premium price received for their novel products. In other words, in a free market economy transactions presumably occur because both the beneficiary firm and the source firm are willing to exchange the items in order to maximize their profits. In brief, no one is doing a favor to anyone.

The foregoing leads naturally to a strong definition of the notion in question, namely: R&D spillovers are knowledge spillovers originated by firms undertaking R&D activities. Clearly, the strong definition constitutes a proper subset of the weak definition. The partition between the weak and the strong definition leaves us with a ‘residual’ of innovation spillovers, namely beneficial spillovers originated either by efficiency-enhancing novel products or by reductions in the input prices. This residual category is useful to visualize one of the simplifying assumptions underlying the ER taxonomy: Table 3 focuses only on innovation spillovers that constitute pecuniary externalities stemming from efficiency-enhancing novel products.

For the sake of completeness, we finally make contact with the nomenclature used in Kettle et al. (2000, esp. pp. 485-486): the ‘weak’ and the ‘residual’ definitions of innovation spillovers correspond to ‘pure knowledge spillovers’ and ‘rent spillovers’, respectively.

References

Adams, James D. (1990): “Fundamental Stocks of Knowledge and Productivity Growth”, Journal of Political Economy, 98, pp.673-702.

Aghion, P. and P. Howitt (1992): “A Model of Growth through Creative Destruction”, Econometrica, March, 60, pp. 323-351.

Bernstein, Jeffrey, and M. Ishaq Nadiri (1989): “Research and Development and Intra- Industry Spillovers: An Empirical Application of dynamic Duality”, Review of Economic Studies, 56, pp. 249-269.

Bernstein, Jeffrey, and M. Ishaq Nadiri (1988): “Interindustry R&D Spillovers, rates of Return, and Production in High- Tech Industries”, American Economic Review, 78, pp. 429-434.

Carroll, Peter, Eduardo Pol, and Paul L. Robertson (2000): “Classification of Industries by Level of Technology: An Appraisal with Some Implications”, Prometheus (forthcoming).

Griliches, Zvi (1958): “Research Cost and Social Returns: Hybrid Corn and Related Innovations”, Journal of Political Economy, 66, pp. 419-431.

Grossman, Gene M. and Elhanan Helpman (1991):Innovation and Growth in the Global Economy. Cambridge, MA: MIT Press.

Kettle, Tor J., Jarle Moen, and Zvi Griliches (2000): “Do Subsidies to Commercial R&D Reduce Market Failures? Microeconometric Evaluation Studies”, Research Policy, April, 29, pp.471-495.

OECD (1996), Science, Technology and Industry Outlook: Part V – Special Theme: The Knowledge-based Economy. Directorate for Science, Technology and Industry, March, pp.1-42.

OECD (1999), OECD Science, Technology and Industry Scoreboard 1999. Benchmarking Knowledge-Based Economies, pp. 1-178.

Pavitt, Keith (1984): “Sectoral Patterns of Technical Change: Towards a Taxonomy and a Theory”, Research Policy, 13, pp. 343-373.

Quesnay’s Tableau Economique (1972), Kuezynski, M. and R.L. Meek, eds., London, Macmillan.

Romer, Paul M. (1990a): “Are Nonconvexities Important for Understanding Growth?”, American Economic Review, May, 80, pp. 97-103.

Romer, Paul M. (1990b): “Endogenous Technological Change”, Journal of Political Economy, October, 98 (5), pp. 71-102.

Schumpeter, Joseph, A. (1961): History of Economic Analysis. London, Allen & Unwin.

Scitovsky, Tibor (1954): “Two Concepts of External Economies”, Journal of Political Economy, April, pp. 143-151.

Stern, Nicholas (1991): “The Determinants of Growth”, Economic Journal, 101, January, pp. 122-133.

Young, Allyn (1928): “Increasing Returns and Economic Progress”, Economic Journal, 38, December, pp.527-542.

Endnotes

1. In this paper we follow the standard practice of using the terms ‘sector’ and ‘industry’ synonymously.

2. According to Schumpeter (1961, esp. pp. 241-243), the Tableau Economique can be thought of as a pictorial description of a system of simultaneous equations, and thereby, a precursor of the input-output system and modern general equilibrium analysis.

3. Although the latest OECD classification of industries consists of four groups of industries, (High tech, Medium-high tech, Medium-low tech, and Low-tech industries), it is customary to refer this as the OECD’s “dichotomy” of high-tech/low-tech industries.

4. For the sake of definiteness, NGT is identified here with the line of formal reasoning inaugurated by Romer (1990b), and the contributions of Grossman and Helpman (1991), Aghion and Howitt (1992), and others.

5. Compound document that blends information from an internet browser with information from a Microsoft application like Excel spreadsheet.

6. New cost-reducing technology expected to be introduced into aluminum plants in a year or two.

7. New technology expected to double the savings of the inert anodes technology and to be introduced into aluminum plants in four or five years.

8. The relationships between knowledge spillovers, R&D spillovers and externalities is not free of subtleties, and they are discussed in a brief appendix at the end of this paper.

9. The allocation of the symbols V and X that appears in Table 2 is for illustrative purposes only.

10. In passing, we note that the use of language about the “old” and the “new” economy is confusing because it opens the possibility of a play on words. In fact, a new economy is characterized by the coexistence of old (mature) and new (enabling) sectors, and therefore, an integral part of a “new” economy is still the “old” economy.

11. The fundamentals for technological change are: a stable and predictable political environment; credible macro and microeconomic policies; secure property rights; an appropriate endowment of human capital; suitable technology distribution power (i.e. science and technology institutions that address the industry needs); and government support for innovation.

12. We are designing a large scale field study to quantify the enabling linkages underlying the ER taxonomy.

13. In many cases the creation of a novel product is linked to new manufacturing processes, that is, process development turns out to be an integral part of product development.

14. Kettle et al. (2000) contains a detailed discussion of the difficulties associated with measurement of R&D spillovers.

15. The partition of external economies into technological externalities (external influences in the technological possibilities of a firm) and pecuniary externalities (external economies operating through the market mechanism) is due to Scitovsky (1954).