Industrial Economics

Antiti What Is Industry? Many people think of industry as the collective large-scale manufacturing of goods in well-organized plants with a high degree of automation and specialization. Although this is a common example of industry, it can also include other commercial activities that provide goods and services such as agriculture, transportation, hospitality, and many others. Industry can be classified into different categories or levels for a better understanding of the different types and for making it easier to study. Although many school textbooks list only three levels, more advanced books classify industry into five levels. The terms for each level originate from Latin words referring to the numbers one to five. Levels of Industry Primary (first): Primary industries are those that extract or produce raw materials from which useful items can be made. Extraction of raw materials includes mining activities, forestry, and fishing. Agriculture is also considered a primary industry as it produces “raw materials” that require further processing for human use. Secondary (second): Secondary industries are those that change raw materials into usable products through processing and manufacturing. Bakeries that make flour into bread and factories that change metals and plastics into vehicles are examples of secondary industries. The term “value added” is sometimes applied to processed and manufactured items since the change from a raw material into a usable product has added value to the item. Tertiary (third): Tertiary industries are those that provide essential services and support to allow other levels of industry to function. Often simply called service industries, this level includes transportation, finance, utilities, education, retail, housing, medical, and other services. Since primary and secondary levels of industry cannot function without these services, they are sometimes referred to as “spin-off” industries. Much of the city of Thompson, for example, is made up of tertiary or service industries to support the primary industry of mining. Quaternary (fourth): Quaternary industries are those for the creation and transfer of information, including research and training. Often called information industries, this level has seen dramatic growth as a result of advancements in technology and electronic display and transmission of information. Quinary (fifth): Quinary industries are those that control the industrial and government decision-making processes. This level includes industry executives and management and bureaucrats and elected officials in government. Policies and laws are made and implemented at this level. BLM 4–1 

Industrial Organization in Context Innovation I Stephen Martin March 2010 c 2010 () Innovation I 03/10 1 / 40 Schumpeter Schumpeter Mark I (The Theory of Economic Development, 1934, p. 66): it is not essential to the matter — though it may happen — that the new combinations should be carried out by the same people who control the productive or commercial process which is to be displaced by the new. On the contrary, new combinations are, as a rule, embodied, as it were, in new …rms which generally do not arise out of the old ones but start producing beside them; . . . in general it is not the owner of stage-coaches who builds railways. If this vision of innovation in a market system is correct, to promote innovation means facilitating the establishment of new, and often relatively small, …rms. c 2010 () Innovation I 03/10 2 / 40 Schumpeter Schumpeter Mark II (Capitalism, Socialism, and Democracy, 1943, p. 82): As soon as we go into details and inquire into the individual items in which progress was most conspicuous, the trail leads not to the doors of those …rms that work under conditions of comparatively free competition but precisely to the doors of the large concerns . . . and a shocking suspicion dawns upon us that big business may have had more to do with creating that standard of life than with keeping it down. This alternative vision of the innovative process leads to the idea that poor static market performance (large …rms that are able to hold price above marginal cost for extended periods of time) may be a small price to pay for good dynamic market performance (a cornucopia of new products, produced ever more e¢ciently). c 2010 () Innovation I 03/10 3 / 40 Schumpeter Large-…rm advantages in R&D? larger …rms are able to spread …xed cost of research over a larger sales base; large …rms may have advantages in …nancial markets; larger …rms may be better able to exploit economies of scale and scope in research, if such economies exist; (the serendipity e¤ect) a large, diversi…ed …rm is more likely to be able to exploit an unexpected discovery. But: large, established …rms may become bureaucratic and resistant to change; familiarity with established products and processes may make management slow to see the advantages to be gained from new products or processes. c 2010 () Innovation I 03/10 4 / 40 Pro…t to be gained from innovation: monopoly Figure: Pro…t to be gained by innovation under monopoly; p = 100 Q; c1 = 50; c2 = 25; Q1 = 25, P1 = 100 25 = 75, Q2 = 37.5, P2 = 62.5. c 2010 () Innovation I 03/10 5 / 40 Pro…t to be gained from innovation: monopoly Compare monopoly pro…t before and after cost-reducing innovation Inverse demand equation p = 100 Q. Constant marginal and average cost, output, and pro…t MC = AC = c1 = 50 ! qm = 25, π 1 = 625 MC = AC = c2 = 25 ! qm = 37.5, π 2 = 1406.5. π2 c 2010 () π 1 = 1406.25 Innovation I 625 = 781.25. 03/10 6 / 40 Pro…t to be gained from innovation: perfect competition Figure: Pro…t to be gained by innovation under perfect competition; p = 100 Q; p1 = p2 = c1 = 50; c2 = 25; q1 = q2 = 50. c 2010 () Innovation I 03/10 7 / 40 Pro…t to be gained from innovation: perfect competition Compare long-run perfect competition pro…t (0) with innovator’s maximum pro…t when fringe …rms can use the old technology. MC = AC = 50, long-run equilibrium price = 50, quantity supplied = 50 units, π 1 = 0. If a single …rm develops a cost-reducing innovation for which it receives a completely e¤ective patent and is able to produce at MC = AC = 25: sell slightly more than 50 units of output at a price slightly less than 50, π 2 = (50 25)(50) = 1250. π2 c 2010 () π 1 = 1250 0 = 1250 > 781.25. Innovation I 03/10 8 / 40 Pro…t to be gained from innovation Perfect competition versus monopoly This is a general result: for a monopolist, the payo¤ to innovation is the di¤erence between the monopoly pro…t it will get with a new product or process and the monopoly pro…t it gets in any case with the existing technology. for an innovating …rm in a competitive industry — a …rm that innovates to escape the constraints of static product market competition — all of the pro…t that ‡ows from successful innovation is a net gain. c 2010 () Innovation I 03/10 9 / 40 Large …rm advantages? (once again) large …rms able to spread …xed cost of research over a larger sales base; large …rms advantages in …nancial markets; large …rms are better able to exploit economies of scale and scope in research, if such economies exist; (the serendipity e¤ect) a large, diversi…ed …rm is more likely to be able to exploit an unexpected discovery. c 2010 () Innovation I 03/10 10 / 40 Concentrated market advantages? leading …rms in concentrated industries will earn economic pro…ts, and so be able to …nance costly R&D e¤orts. with few signi…cant rivals, a large …rm can be con…dent that it would be able to appropriate the pro…ts ‡owing from success — that the innovation would not be imitated by numerous small rivals and the pro…ts competed away. c 2010 () Innovation I 03/10 11 / 40 Market structure & innovation An inverted-[ relation between seller concentration and the rate of technical progress (Villard, 1958, p. 491): [I]ndustries where “competitive oligopoly” prevails are likely to progress most rapidly and that therefore “competitive oligopoly” may well be the best way of organizing industry. The basic point is that progress is likely to be rapid (1) when …rms are large enough or few enough to a¤ord and bene…t from research and (2) when they are under competitive pressure to innovate—utilize the results of research. c 2010 () Innovation I 03/10 12 / 40 Firm structure & innovation Large …rms have advantages over small in raising funds to …nance R&D Small …rms less likely to be locked-in (by internal bureaucracies) to established technologies Arrow (1983) Smaller …rms will tend to specialize more in the research phase and in smaller development processes; larger …rms will devote a much smaller proportion of their research and development budget to the research phase. They will specialize in the larger developments and will buy a considerable fraction of the research basis for their subsequent development of innovations. c 2010 () Innovation I 03/10 13 / 40 Firm size & innovation Caveat investigator: let the researcher beware! There is no unambiguous measure of the degree of innovation or technological progress. Measures that have been used R&D inputs: spending on R&D; employment of scientists & engineers R&D output: patents (an elastic yardstick); productivity growth (a residual approach — that part of changes in input-output relations that is not explained by changes in input usage) c 2010 () Innovation I 03/10 14 / 40 Firm size & innovation (Cohen and Klepper, 1996b): empirical studies show that the probability that a …rm does R&D at all rises with …rm size; for most industries, among …rms that do R&D, R&D is roughly proportional to …rm size; among …rms that do R&D, R&D productivity — the number of innovations per unit of spending on R&D — falls as …rm size rises. c 2010 () Innovation I 03/10 15 / 40 Firm size & innovation (Cohen and Klepper, 1996, p. 933): . . . larger …rms pursue more marginal R & D projects than smaller …rms because they have a bigger output over which they can apply the results of the projects . . . Consequently, the average project they pursue has a lower return, measured in terms of the number of patents or innovations, than smaller …rms. . . . The greater output over which larger …rms can apply their R & D enables them to pro…t more from R & D than smaller …rms, which leads them to undertake more R & D projects at the margin than smaller …rms. By undertaking more R & D, larger …rms achieve a lower average cost of production and/or higher product quality and hence greater pro…ts than smaller …rms. This explains why larger …rms can prosper despite the lower average productivity of their R & D. c 2010 () Innovation I 03/10 16 / 40 Firm size & innovation (Cohen and Klepper, 1996a) Innovations the use of which is proportional to output levels are more pro…table for large …rms than for small, all else equal Large …rms tend to favor process innovation Small …rms tend to favor product innovations. c 2010 () Innovation I 03/10 17 / 40 Market structure & innovation Scherer (1967): inverted-[ relationship between concentration ratios and employment of technical personnel in 56 U.S. manufacturing industries in 1960: employment of technical personnel rose with the four-…rm seller concentration through values between 50 and 55 per cent, and fell thereafter. Scott (1984) inverted-[ relationship between R&D spending per dollar of sales and seller concentration for 3388 lines of business of 437 U.S. manufacturing …rms; peak at a four-…rm seller concentration ratio of 64 per cent. The inverted-[ relationship disappears when he controls for unobserved …rm and industry e¤ects. Blundell et al. (1999): for 340 UK manufacturing …rms for the years 1972–1982, more concentrated industries had fewer innovations, as did industries facing less import competition. But within an industry, (1999, p. 550) “it was the high market share …rms who tended to commercialize more innovations. . . .” c 2010 () Innovation I 03/10 18 / 40 Rivalry & innovation Schumpeter Mark II: there is a positive impact of market concentration on innovation because the leading …rms in concentrated markets are shielded from rivalry — they earn economic pro…ts that allow them to …nance innovation, and (having few rivals), they will be able to act in con…dence that they will be able to collect whatever pro…t ‡ows from successful innovation. (Baily and Gersbach, 1995): for nine industries in Germany, Japan, and the United States, international rivalry promotes competitivity (1995, p. 308): Vigorous global competition against the best-practice companies not only spurs allocative e¢ciency, it can also force structural change in industries and encourage the adoption of more e¢cient product and process designs. They also …nd (1995, p. 345) that competition by foreign …rms, not bound or blinded by traditional ways of doing things, has a much greater impact on performance than competition by domestic …rms. c 2010 () Innovation I 03/10 19 / 40 Rivalry & innovation Broadberry and Crafts (2001) market structure, trade association activity, and UK innovation in the 1950s. Little evidence that trade association activity a¤ected innovation Negative impact of seller concentration on innovation. (2001, p. 112) On balance, the evidence . . . goes against the claim that market power promotes innovation. c 2010 () Innovation I 03/10 20 / 40 Demand-pull Jacob Schmookler (1962, 1966) In a market system, more resources will be invested in innovations that are expected to be more pro…table, all else equal, and expected pro…t will be positively related to output. (1962, p. 18, footnote omitted) . . . expected pro…ts from invention, the ability to …nance it, the number of potential inventors, and the dissatisfaction which invariably motivates it—are all likely to be positively associated with sales. In a market system it is the lure of pro…t that directs the allocation of resources to innovation (as it directs the allocation of resources to other activities). c 2010 () Innovation I 03/10 21 / 40 Technology-push Exogenous technological advance is a prerequisite to pro…table innovation (Rosenberg, 1974, p. 97): . . . the progress made in techniques of navigation in the sixteenth and seventeenth centuries owed much to the great demand for such techniques in those centuries. . . . But . . . a great potential demand existed in the same period for improvements in the healing arts generally, but that no such improvements were forthcoming. The essential explanation is that the state of mathematics and astronomy a¤orded a useful and reliable knowledge base for navigational improvements, whereas medicine at that time had no such base. c 2010 () Innovation I 03/10 22 / 40 Testing Distinguish fundamental innovations, the result of exogenous technological leaps, and incremental innovations, which perfect and apply fundamental innovations, and which are more likely to respond to the lure of pro…tability. (Scherer, 1982a) Patenting activity by 443 large U.S. …rms responded immediately to demand ‡uctuations, patenting responded more to di¤erences in sales in areas like chemicals and electronics, less in sectors using traditional technologies. c 2010 () Innovation I 03/10 23 / 40 R&D Spillovers Arrow (1962, p. 615): no amount of legal protection can make a thoroughly appropriable commodity of something so intangible as information. The very use of the information in any productive way is bound to reveal it, at least in part. Mobility of personnel among …rms provides a way of spreading information. Legally imposed property rights can provide only a partial barrier, since there are obviously enormous di¢culties in de…ning in any sharp way an item of information and di¤erentiating it from similar sounding items. c 2010 () Innovation I 03/10 24 / 40 R&D Spillovers Spillovers may have bene…cial e¤ects, from a social point of view 1956 U.S. antitrust consent decree required AT&T to openly license patent-controlled technology at reasonable rates: facilitated the development of UNIX AT&T required cross-licenses of patents from …rms to which it granted licenses under the consent decrees, and this pattern of cross-licenses contributed to rapid entry into and growth of the U.S. semiconductor chip industry. c 2010 () Innovation I 03/10 25 / 40 R&D Input Spillovers When the R&D e¤orts of one …rm help rivals reach their own research goals The pharmaceutical industry: is characterized by high rates of publication in the open scienti…c literature, and many of the scientists . . . stressed the importance of keeping in touch with the science conducted both within the public sector and by their competitors. Nearly all of them had a quite accurate idea of the nature of the research being conducted by their competitors, and they often described the ways in which their rivals’ discoveries had been instrumental in shaping their own research. E¤ective R&D e¤ort in high science-content sectors like pharmaceuticals requires that researchers keep abreast of the knowledge frontier in their …eld. The interactions this requires reveal what they are doing to researchers working in other places, just as they learn what other researchers are doing. c 2010 () Innovation I 03/10 26 / 40 R&D Output Spillovers When …rst-discoverers are not able to collect all of the economic pro…t generated by their innovation. An innovator would not be able to appropriate the consumers’ surplus generated by the new product or process without engaging in price discrimination (charge each consumer his or her personal maximum price for each unit purchased) — the private return to investment in innovation will, in general, be less than the social return. For 48 US new product innovations, Mans…eld et al. (1981) report that 60 percent of successful patented innovations were imitated within four years of introduction. For a sample of 100 US manufacturing …rms, Mans…eld (1985) reports survey evidence indicating that rivals have information about R&D decisions in 12–18 months, and information about new products or processes in 12 months or less. c 2010 () Innovation I 03/10 27 / 40 R&D Output Spillovers Such leakages occur (Mans…eld, 1985, p. 221) because input suppliers and customers are important channels (since they pass on a great deal of relevant information), patent applications are scrutinized very carefully, and reverse engineering is carried out. In still other industries, the di¤usion process is accelerated by the fact that …rms do not go to great lengths to keep such information secret, partly because they believe it would be futile in any event. c 2010 () Innovation I 03/10 28 / 40 R&D Spillovers Neither R&D input spillovers nor R&D output spillovers are unambiguously bad, from a social point of view. R&D input spillovers increase the productivity of such R&D e¤orts as do take place, and make it more likely that some …rm will discover (which is what is important from a social point of view). R&D output spillovers improve static product market performance after discovery takes place. But both types of spillovers reduce the incentives of …rms to invest in innovation. c 2010 () Innovation I 03/10 29 / 40 Appropriability and absorptive capacity Arrow argued that knowledge ‡ows freely throughout the economy. A contrary view is that one must work to acquire and maintain the ability to absorb knowledge (Cohen and Levinthal, 1989, pp. 569–70) we argue that while R&D obviously generates innovations, it also develops the …rm’s ability to identify, assimilate, and exploit knowledge from the environment—what we call a …rm’s ‘learning’ or ‘absorptive’ capacity. While encompassing a …rm’s ability to imitate new process or product innovations, absorptive capacity also includes the …rm’s ability to exploit outside knowledge of a more intermediate sort, such as basic research …ndings that provide the basis for subsequent applied research and development. c 2010 () Innovation I 03/10 30 / 40 Appropriability and absorptive capacity Tacit knowledge Vonortas (1994, p. 415): technological knowledge involves a combination of poorly-de…ned, and often incomplete, know-how and a set of highly codi…ed information which is hard to acquire and utilize e¤ectively. Where knowledge is tacit, …rms need to maintain their own stock of knowledge and technical ability to absorb knowledge generated elsewhere in the economy. c 2010 () Innovation I 03/10 31 / 40 Appropriability and absorptive capacity Tacit knowledge Angelmar (1987, pp 73–74) studies business responses to a question “whether it bene…ts to a signi…cant degree from patents, trade secrets, or other proprietary methods of production or operation” to measure industry appropriability conditions. For 160 business units that were parts of large …rms in 1978, R&D spending per dollar of sales was consistently higher, the greater appropriability. there was a positive impact of seller concentration on R&D spending in industries with low appropriability and low customer switching costs, a negative impact of seller concentration on R&D spending in industries with high appropriability and high customer switching costs. c 2010 () Innovation I 03/10 32 / 40 Appropriability and absorptive capacity Tacit knowledge Where other market conditions do not favor private investment in innovation, high seller concentration has the e¤ects envisaged by Schumpeter Mark II. Where market conditions favor private investment in innovation even by small …rms, the inclination of large …rms to pursue the quiet life makes Schumpeter Mark I a better explanation for market structure-dynamic market performance relationships. c 2010 () Innovation I 03/10 33 / 40 Uncertainty The creative process is inherently uncertain. How a particular approach will turn out cannot be foreseen; in the words of one drug-industry researcher (quoted in Tapon and Cadsby, 1996, pp. 389–90): I think that rational drug design is obviously very admirable. It’s more than a great idea, it’s a move in the right direction. It applies as much rationality to your programs as possible. But, you’re not going to be able to predict 100% . . . of the outcome. You’re always going to have things that happen that nobody really foresaw and you look back in hindsight and say that there is no way that we could have predicted that outcome. . . There is a certain amount of good luck involved . . . you have to have the breaks; if you don’t have the breaks in drug development you may have great di¢culty in getting any compound. c 2010 () Innovation I 03/10 34 / 40 Uncertainty Not only is it uncertain, ex ante, how a particular research project may turn out, but it may also be uncertain, ex post, whether results obtained have interesting commercial possibilities. The market for genuinely new products may not be immediately recognized. For example, the fundamental innovation embodied in the now ubiquitous post-it R sticker was essentially an adhesive substance that was not terribly adhesive, developed in 1968. The product was …rst introduced in 1980, 12 years later. Uncertainty, too, reduces the incentives of …rms to invest in R&D. c 2010 () Innovation I 03/10 35 / 40 Basic vs. applied R&D (National Science Foundation, 2001) Basic Research: directed toward increases in knowledge or understanding of the fundamental aspects of phenomena and of observable facts, without speci…c application or commercial objectives Applied Research: research directed toward gaining knowledge or understanding necessary for determining the means by which a recognized and speci…c need or commercial objective may be met. Development: the systematic use of the knowledge or understanding gained from research directed toward the production of useful materials, devices, systems or methods, including design and development of prototypes and processes (excludes quality control, routine product testing, and production). c 2010 () Innovation I 03/10 36 / 40 Basic vs. applied R&D Basic Research: carried out mostly in universities; “public good” aspects strongest and spillovers likely to be bene…cial. hence, increasing appropriability of the fruits of university R&D potentially harmful Applied Research, Development: carried out mostly in the private sector, where results will be applied; absorptive capacity more important. c 2010 () Innovation I 03/10 37 / 40 R&D and market structure Schumpeter Mark I: turnover of leading …rms — today’s dominant …rm displaced by an innovative entrant, which is in its turn displaced by a later innovator Schumpeter Mark II: the dominant …rm’s comparative advantage in innovation allows it to maintain its dominant position, and this goes hand in hand with good dynamic market performance. c 2010 () Innovation I 03/10 38 / 40 R&D and market structure Strategic entry deterrence The prediction of many models is that the duopoly pro…t gained by an entrant will be less than the pro…t lost by an incumbent should entry occur — it follows that an incumbent could (à contre cœur ) pay a potential entrant enough to make the entrant indi¤erent to the prospect of entry, leaving the incumbent better o¤ Example: reverse payment patent settlement agreements in which a pharmaceutical company holding a patent pays the producer of a generic substitute to keep the generic o¤ the market. c 2010 () Innovation I 03/10 39 / 40 R&D and market structure Strategic entry deterrence By the same token, if the incumbent can be certain of innovating …rst by spending more than a potential entrant, it will have a greater incentive to do so — the pro…t it would lose should entry occur is more than the duopoly pro…t the entrant would gain. But if the outcome of innovation is uncertain, the replacement e¤ect identi…ed in Arrow’s static model comes into play — the incumbent is certain to get economic pro…t until entry occurs, and this reduces its incentive to invest in innovation, relative to an entrant for which all post-innovation economic pro…t would be a gain. c 2010 () Innovation I 03/10 40 / 40