The “Somewhat Optimal” Attributes of Capitalist Growth: Oligopolistic Competition and Routinization of Innovation
innovations are rarely the dramatic breakthroughs that Schumpeter may have had in mind but rather small improvements in a new process or product in which genuine novelty and imitation-with-a-difference shade imperceptibly into one another.
—Blaug, 1999, p. 110
It is something of a puzzle, therefore, why the capitalist innovation system has performed so well. There certainly is nothing like the twin theorems of welfare economics around to support an argument that capitalism “can’t be beat.”
—Nelson, 1996, pp. 54–55
The phrase “somewhat optimal” appears to be an oxymoron, but actually describes the growth processes of capitalism effectively. Indeed, there is no foundation for a claim that free-enterprise economies automatically tend to satisfy the requirements of ideal economic efficiency in the growth process. Yet there are substantial reasons to conclude that the patently extraordinary growth record of the free-enterprise form of economic organization is hardly accidental, and that it is in large part attributable to the pressures of the free market upon the business firm, which force it to spend liberally and continually on the innovation process and to make its innovations available to others if those others are willing to pay an attractive price. Moreover, I will argue that the capitalist growth mechanism has welfare properties far more desirable than the literature of economics seems hitherto to have suggested.
INNOVATION AND THE GROWTH PERFORMANCE OF CAPITALISM
Per capita income in the leading capitalist economies is growing at a rate that apparently permits something like an eightfold multiplication in a century (as John Maynard Keynes predicted in 1932).1 I suggest this number is so large that it defies comprehension. What would our lives be like if we were recipients of an average family income today, and then seven-eighths of that amount were suddenly removed? In contrast, it is estimated (very crudely, of course) that in wealthy eighteenth-century England real per capita income had just about reattained the level it had reached in third-century Rome, some fifteen centuries earlier. Words do, indeed, fail to convey the incredible growth record of the industrialized free-market economies. Undoubtedly, it is these spectacular and unmatched growth rates that distinguish them most from all other economic systems. In no other system, current or in the past, has the average income of the general public risen anywhere nearly as much or as quickly as it has in North America, Western Europe, and Japan. Though the Soviet Union planned its economy and forced its population to invest heavily in factories and hydroelectric dams, it failed to produce enough materially to raise the standard of living of its population, as the free-market economies have been able to do.
What is the secret of their extraordinary success? That is the economic puzzle that undoubtedly is critical to the degree of prosperity achievable in the future. Its answer is what the world’s poorer countries are anxious to learn. The central purpose of Part One of this book is to provide some of the missing explanations—to describe the features of the free-market economy that make it into a fantastically effective machine for the production of innovations and for the consequent growth in its output. This book will emphasize three components of that explanation: the fierce competition among many of the economy’s enterprises, seeking to come up with the better new mousetrap or the better way to produce the old mousetrap; the resulting routinization of the innovation process that reduces the firm’s dependence upon fortunate happenstance in the form of an appropriate invention that just happens to appear; and the competitive pressures to disseminate proprietary technology voluntarily—to make it available, for a suitable return of course, even to direct competitors.
In particular, I will contend here that routine innovation processes—those guided by standard business-decision principles—are a key part of the story. They are, indeed, of great and probably growing importance, with more than two-thirds of U.S. R&D expenditure channeled through business firms. That is one of the key components of the free market’s innovation mechanism, and one that will be emphasized in this book. It thereby focuses on a source of the economy’s stream of new products and processes that is more dependable than the lone inventors who are so widely regarded as the prime creators of technical change.
THE CONTINUING IMPORTANCE OF UNROUTINIZED ENDOGENOUS INNOVATION
However, the critical and growing role of routine innovative activity does not mean that independent innovation no longer plays a significant role.2 F. M. Scherer (1980, p. 438) provides a long list of major technical inventions introduced by newly created firms, whose personnel were presumably not subject to the pressures for routinization that one expects in large, established enterprises. His examples include the incandescent lamp, alternating electric current, radio telegraph and telephony, the dial telephone, the synchronous orbit communications satellite, the turbojet engine, the sound motion picture, self-developing photography, and the electronic calculator, among many others. One can even offer the plausible conjecture that most revolutionary new ideas have been, and are likely to continue to be, provided preponderantly by independent innovators.3
But once their initial undertaking proves successful, the inventors or the associated entrepreneurs often establish new firms that in many cases grow large and themselves routinize their innovation. Henry Ford and his automobiles provide only one of many examples. Such subsequent innovation, like most innovations produced by routine processes, is primarily devoted to product improvement, increased reliability, enhanced user-friendliness of products, and the search for new uses for those products. Though the routine innovations tend to be less dramatic, both the independent and the routinized innovation activities undoubtedly contribute significantly to economic growth, as Nathan Rosenberg has emphasized (see, e.g., 1976, p. 66).
The main point here is that there is no reason to expect the independent inventor or innovator to become obsolete any time in the foreseeable future. Indeed, there is a serendipitous relationship between the routine and the independent economic activities, whose results are arguably superadditive, with the whole greater than the sum of the parts.
PROBLEMS BESETTING INNOVATION AND GROWTH UNDER FREE ENTERPRISE
For all its achievements, it cannot be claimed that the capitalist economy has no shortcomings as an engine of growth. There are good reasons to suspect that innovation under free enterprise will be far from optimal, including the following.
Benefit Spillovers. The first and most widely recognized source of imperfection is the existence of substantial spillovers of the benefits generated by innovation. Generally, a considerable share of the benefits of a particular innovation goes, without compensation of the innovators, to individuals or groups who have made no contribution to the discovery and development of that invention. Often, these non-paying beneficiaries even include competitors of those who hold the rights to the innovation. Thus, it is generally inferred that investors are unlikely to devote the socially optimal quantity of resources to the innovation process.
Other Externalities. Recently, the economics literature has begun to emphasize other sources of differences between the private and social costs of innovation. For example, Philippe Aghion and Peter Howitt (1998) have drawn attention to this side of Schumpeter’s “creative destruction” concept. New products and processes can destroy the marketability of their predecessors, even though these would otherwise have retained considerable value. If the proprietors of the older assets are individuals or groups different from the innovators, the latter will have no incentive to take these losses in value into account in deciding on the magnitude of their innovative activities. For example, the creators of Microsoft Excel were not deterred by the prospect that they would drive the suppliers of other spreadsheet software out of business, even though the now dominant product is arguably only marginally better than the others. An innovation with an expected market value of $10 million will be an attractive proposition to the innovator if its expected private cost is $7 million. But it will be a net loss to society if the process also makes $8 million in older assets obsolete.
Proprietors’ Withholding of Technological Information. A third and less commonly discussed source of inefficiency in the growth process is the incentive for the proprietors of an invention to deny its use to others, notably to competitors, by means of secrecy, patents, and recourse to litigation to enforce those patents. It would obviously be inefficient for a superior process or product feature to be restricted to only one of the firms in an industry, because the others might prove to be the more efficient producers if the innovation became available to them.4 Even if the innovation were not to become available to other firms, they might manage to survive but they would then be supplying products of inferior quality, produced by obsolete processes. If the products of that industry are varied, many suppliers of close-substitute goods may be able to remain in the market, despite the resulting inefficiency in their operations, if each possesses proprietary technology or product features that it can prevent the others from using. In other words, deliberately erected obstacles to dissemination need not prevent simultaneous survival of a number of firms, but the result can be highly inefficient. In addition, mitigation of this problem for growth entails an offsetting danger: weakening of this third obstacle to optimality in the free-enterprise growth process can exacerbate the first. That is, anything that facilitates the easy and rapid dissemination of technology appears to enhance the spillovers of innovation by making it easier for non-innovators to derive a share of the benefits.
Inappropriate License-Fee Levels. A fourth difficulty relevant for the efficiency of the free-enterprise growth process also has not been discussed as widely as the spillover problem. This relates to the pricing of technology licenses. An obvious way simultaneously to reduce the severity of both the first (spillovers of benefits) and third (technology withholding) of the problems that have just been discussed is for firms with proprietary technology to find it profitable to license its use. This clearly ameliorates the dissemination problem and simultaneously reduces spillovers by offering direct compensation to investors in innovation. But profitability depends on the license fee that the owners of the technology are able to obtain. If it is too low they will have no incentive to license. And if they do license, will they not have an incentive to extract a license fee that is far too high from the viewpoint of economic efficiency?
Rent-Seeking and Destructive Entrepreneurial Activity. Entrepreneurs are widely credited with a key role in economic growth. But business persons and innovators can be expected to direct their efforts to where the money is. Many will be tempted to choose activities that promise to be lucrative regardless of the benefits these activities offer or do not offer to society, as in the many cases when firms institute lawsuits against rivals, seeking to curb the vigor of the rivals’ competitive activities and hoping, in addition, for some monetary payoff in the form of damage payments imposed by the courts.5 Entrepreneurs will sometimes devote innovative ideas as well as their energies to the creation of monopolies, and even to criminal activities. That is hardly the most effective avenue to growth.
”Winner Take All” Patent Races. A sixth problem that can handicap the efficiency of the innovation process is the “winner take all” property of some patent races. Partha Dasgupta and Joseph Stiglitz (1980) have demonstrated that R&D expenditure can be excessive in terms of social welfare when competing firms, in an attempt to win an innovation race, duplicate efforts and rush to get ahead of one another’s innovation programs. In addition, if the winner takes all, the efforts of those who have sought the desired innovation, but failed to obtain the patent, are wasted. Although such duplicative effort does reduce the risk that a sought-after invention will not be discovered, there seems to be no assurance that this decrease in risk will be worth the expenditure of resources that goes into the race. Moreover, the absence of any reward to the runners-up may exacerbate the risks of the innovation process and it may mean that somewhat inferior, but nevertheless valuable, substitutes for the winning invention are denied to potential purchasers whose special needs they may meet satisfactorily.
HOW THE MARKET DEALS WITH THE IMPEDIMENTS TO EFFICIENCY IN INNOVATION
The market economy does have attributes that mitigate these problems, though it can hardly be claimed that it deals with them optimally. This is a second central theme of Part I of this book. With the exception of the problems associated with “winner take all” patent races, on which only a few comments will be offered (in the appendix to this chapter), all of the efficiency impediments just listed will be dealt with in the chapters that follow. And it will be shown that for each of these difficulties the market mechanism has features that make them less serious than they might otherwise have been.
For example, contrary to what may at first be thought, it will be shown that firms emphatically do have strong reasons to disseminate their technology, reasons not commonly noted in the literature. Thus, firms other than the owner of the proprietary technology are not driven to reliance on obsolete products and processes.
One reason for voluntary dissemination is that proprietary technology is like any other asset in this respect—it is profitable to rent it out if the price is right. After all, if the price is sufficiently high, employment of its technology by others will be the most profitable option to the owner. And it will pay the prospective renter to offer such an attractive price whenever it is in a position to put the technology to better use than the proprietor. As a result, it is not surprising that substantial markets in technology licenses have emerged.
A second reason for voluntary dissemination arises, paradoxically, from the pressures of competition. Exchange of technology need not undermine the competitive position of the enterprise. Indeed, it may well strengthen that position. If a firm supplies its technology to a rival and the favor is reciprocated, both firms, having the use of their combined inventions, will end up strengthened relative to a third competitor that has only its own resources to rely upon for new products and processes. Because of this, market forces provide a strong incentive for the formation of what amount to informal technology consortia, with extensive or even complete sharing of the information. I will also show that such technology consortia provide an incentive for enhancement of investment in innovation, rather than discouraging such outlays. I will provide evidence that these consortia are not theoretical fictions, but that they exist and are widespread in reality. In sum, it will be shown that market forces provide strong incentives for the rapid dissemination of innovation, and do not generally encourage the hoarding of such proprietary property that could significantly impede economic growth.
Each of the other main impediments to optimality in the growth process will also be analyzed, and in each case it will be shown that there is reason to conclude that the handicap is at least far less serious than seems often to be believed. On the contrary, reasons will be given to expect a fairly impressive performance by the free market in at least some of these areas.
This may all sound as if I am hinting that, from the point of view of innovation and growth, the market mechanism almost is all for the best in this best of all possible worlds. But this is far from my purpose. As in the case of the efficiency analysis that deals with the performance of the economy in matters other than growth, application to the real world is characterized by a multitude of warts and blemishes. Monopoly power, ill-considered government intervention, private miscalculation, and ignorance all can and do degrade the performance of the economy. Moreover, nothing that will be said in Part One of this book is meant to argue that even a theoretical market equilibrium is necessarily characterized by optimality properties. Rather, my much more limited claim is that market forces tend to yield an economic performance far closer to the requirements of economic efficiency than at least a superficial reading of the standard analyses might lead one to believe.
SUMMARY
We are left with the following account of the production and distribution of technology in the capitalist growth process. First, continued investment in innovation is ensured by the arms-race character of competition in the high-tech oligopoly industries that will be explored in the following chapter. Such competition also forces firms to routinize the innovation process as a means to reduce their risks. These incentives are enhanced, not undermined, by technology trading and licensing—primarily because those processes serve, via the access (license) fees, partially to internalize the externalities of innovative activity. Second, innovative activity by the firm is stimulated by the requirement for success in technology-exchange negotiations that a negotiating party have something of value to offer to the firm whose technology it hopes to acquire. Because of licensing and technology trading, innovations are now disseminated with historically unprecedented rapidity. Rather than benefiting just a severely limited subsector of the industry and the economy, leaving other producers to fend with obsolete techniques and products, the advantages of technical advances are quickly made available to all. This, too, can be expected to make a significant contribution to economic growth.
These conclusions suggest that the innovation process may be characterized by some efficiency properties that do not seem to be widely recognized in the literature. Although it cannot be denied that the activity is beset by a number of significant imperfections, that still leaves the free-market economies with a flow of innovations of unprecedented magnitude. And even the efficiency-handicapping spillovers of innovation offer a valuable tradeoff between naked productive efficiency and acceptable division of its benefits. By distributing the benefits of technical progress widely among the population, spillovers enhance the economic health of society and surely add social value to the growth accomplishments of the free-market economy.
The role of innovation, investment, and education in the growth of the industrialized economies is, of course, widely recognized. But usually these are, at least by implication, treated as exogenous products of happenstance, not as a predictable product of the free-market growth machine. Thus, consider the following passage by Angus Maddison (2001, p. 21), plainly an astute and well-informed student of growth on whose data many growth analysts quite appropriately rely:
Between 1820 and 1913, British per capita income grew faster than at any time in the past—three times as fast as in 1700–1820. The basic reason for improved performance was the acceleration of technical progress, accompanied by rapid growth of the physical capital stock and improvement in the education and skills of the labour force, but changes in commercial policy also made a substantial contribution. In 1846 protective duties on agricultural imports were removed and in 1849 the Navigation Acts were terminated. By 1860, all trade and tariff restrictions had been removed unilaterally. In 1860 there were reciprocal treaties for freer trade with France and other European countries. These had most-favoured nation clauses which meant that bilateral liberalisation applied equally to all countries.
It is clear that this passage does not connect the innovation and other immediate stimuli of economic growth to what I maintain is their fundamental source—the capitalist economy. Yet without this connection, without recognition of the market economy as an innovation- and growth-producing machine of unparalleled effectiveness, one cannot hope to offer a systematic explanation of the outpouring in the industrialized economies in the past two centuries. Without that, the story appears as one great set of coincidences with little internal coherence, one that could just as easily have happened elsewhere, in radically different circumstances, and that could end as abruptly, as it may appear to have begun. But it will be shown in this book that there is every reason not to accept such a view.
APPENDIX
REMARKS ON HETEROGENEOUS-PRODUCT RACES
Since it is not a central concern of this book, only a few comments will be offered on the patent race issue. Much of the literature on competition among innovators interprets the process as one in which there is a single prize—the winner takes all. The inventor who reaches the patent office first, as when Alexander Graham Bell beat his rival in the invention of the telephone by several hours, becomes the sole possessor of the legal right to profit from the invention. There are, of course, inventions of this sort, particularly when the new products or processes that the rival inventors have tried to create are homogeneous—essentially identical. It is apt to be true of technical advances whose sole purpose is to reduce the cost of a homogeneous final product, because the innovation that provides the greatest cost saving can clearly outcompete any rival, leaving no place in the market for an inferior cost-cutting substitute.
Inventions, however, are generally not so homogeneous. The research and development laboratories of two competing firms often turn out innovations that are imperfect substitutes for one another. One may be superior to the other in some features but inferior in others. Or one may be only marginally inferior and therefore salable at a price that is a bit lower than the other’s. In such cases, as already suggested, the race offers many prizes; the winner receives the highest payoff, but those who come in close behind obtain compensation commensurate with the value of their performances.6
In a highly competitive market the relationship among the payoffs to the different innovators in the race is readily analyzed with the help of the standard rent model associated with David Ricardo. In such markets, competition forces the imperfect substitute innovations to be priced so as to offer purchasers of their services exactly the same net benefits. Thus, if purchasers are homogeneous, the difference between the rent payments to the suppliers of any two (imperfect) substitute innovations must be equal to the difference in the benefits they provide to buyers.
This result follows not only from the behavior of product buyers but also from that of profit-seeking investors in innovation. Investment in the innovation process will be increased only if it is expected to raise the probability of winning a more valuable prize, if, for example, it increases the likelihood of winning the second profit prize rather than the third. If freedom of entry into the innovation process reduces expected profits to zero, however, higher expected payoff prizes will be offset precisely by the higher cost of obtaining a superior invention. So the expected gross earnings from the second prize should exceed those of the third by precisely the incremental cost necessary to advance the firm from being the expected winner of the third prize to winner of the second.
Ricardian rent theory teaches us that such a differential rent arrangement should be consistent with efficiency in allocation of resources among activities that are expected to produce heterogeneous but substitute innovations. Indeed, that is what the writings of Chung Yi Tse (1996) indicate more formally for a rather more restricted set of circumstances. Clearly, these arguments are fully valid if entry into the innovation race faces no barriers. We will see later that the need to sink costs in an innovation process does constitute an entry barrier that may sometimes be substantial. In that case, however, the problem for economic efficiency is attributable to the entry barriers, not to the innovation races that occur in free-market economies.
___________
1. Keynes (1932, pp. 364–65). I must thank Senator Daniel Patrick Moynihan for calling this passage to my attention.
2. Though I obviously do not agree fully, Richard Nelson (1996) even concludes that “Schumpeter’s prognostication that as science grew stronger technical innovation would become more predictable and routine has turned out to be a bad call” (p. 81). However, Nelson does not seem to be disputing that a good deal of R&D has become a routine corporate activity.
3. There is also a good deal of R&D activity in universities and government laboratories. Clearly, this is not research conducted by business, but much of it is different from the work of the independent innovator under discussion. For example, much of the activity of the independent innovator is conducted in pursuit of wealth, and it consists primarily of applied rather than basic research.
4. Thus, Litton sued Honeywell for patent infringement in the latter’s use of a superior mirror coating in the ring laser gyroscope now used to guide and steer most commercial aircraft. As this book was being written, Litton’s initial court victory was still under appeal. There is much evidence indicating that Honeywell was indeed the superior producer in the period to which the dispute pertains. Litton, of course, denies Honeywell’s superiority, and recently improved technology may be changing the competitive balance.
5. A number of concrete illustrations are provided in chapter 4 of my 1993 book on the theory of entrepreneurship. For example, the Sewell Plastics company sued a group of Coca-Cola bottlers in the U.S. Southeast for starting a rival bottle-producing cooperative that supplied bottles at less than half Sewell’s earlier price. Sewell sued, openly demanding that the bottlers agree to purchase only from Sewell. The court threw the case out.
6. In his Ph.D. dissertation, Chung Yi Tse (1996) introduced innovation races very close to what has just been described and analyzed their welfare properties systematically and far more deeply than is done in the brief discussion here. His results also indicate that the races need not conflict with economic efficiency.