This chapter discusses the particular economic characteristics of high-technology markets which are considered to be relevant for competition policy. Instead of coming up with an exact definition on which markets are included and which excluded I will provide in this chapter an overview of the in the academic literature generally agreed upon set of characteristics of high technology markets.
This set include economies of scale in production, network effects, durable goods, the vital role of intellectual property right protection, compatibility and standardization and the dynamic nature of competition on high-technology markets. [2]
1.2 Economies of scale
Standard economic textbooks assume diminishing returns to scale, which results in firms eventually facing barriers to expansion. This is reflected in the cost structure of the firm exhibiting increasing marginal costs along an increasing production. When production is increased and scale advantages are exhausted this enables multiple firms to co-exist on the market. [3] The standard assumption is relatively accurate for traditional markets but Information goods sold at high-technology markets alter this assumption. This goods exhibit increasing returns along all production levels/whole production range. [4]
This characteristic of increasing returns is also called supply side economies of scale. This is the result of high upfront investments in research and developments or investments in physical or virtual networks (which are necessary) to deliver the product to the consumer. This high fixed costs are often combined with low or even negligible marginal costs. A software package is the typical example of an information good for which the incremental cost of producing an additional copy is negligible, once the software has been developed and successfully tested. As a result of the low marginal production cost, the production cost per unit or average production cost, decreases when the production scale is increased. Firms which are serving a larger part of the market on which significant economies of scale are present are able to produce more cheaply and efficiently than their smaller competitors. Market which exhibit increasing returns to scale tends to be concentrated for this reason. [5]
Network effects
1.3.1 General theory
Economies of scale are however not an unique characteristic of high-technology markets. A couple of (different) traditional industries like the gas, electricity or steel industries for instance are also subject to economies of scale. What however is really new of many high-technology markets is that economies of scale are combined with network effects. [6] The two concepts must be clearly distinguished. Both economies of scale and network effects exhibit increasing returns over a substantial part of the demand [7] , the source of these returns are however different. [8] In the former the source must strictly be found in the supply side of the market, independent of the utility consumers can derive from using the product or other demand side effects. (Direct) network effects, on the other side, can be defined as a situation in which the utility a user derives from the consumption of a good increases with the number of other users consuming the same good. [9] Network effects are thus demand-side phenomena as the nature and shape of the demand is affected by an additional user conferring value to the network.
We can clarify the implications of the combination of network effects and economies with a simplified graphical illustration. [10] It must be born in mind that this graphs are not fully accurate as a result of simplification. But the intuition behind it can make our argument better understood.
In figure 1 the supply and demand schedules of a standard good are drawn. As is explained in standard microeconomic textbooks the demand schedule is downward sloping, which shows that when the production is increasing, individuals with lower willingness to pay must be induced to buy the product on the market. The marginal cost schedule represents the supply curve and is either upward sloping or constant. This shows that the extra cost of producing one more unit are rising or constant respectively. In a market of perfect competition, at the point where the two curves cross each other a market equilibrium exists. Perfect competition will thus lead to an equilibrium where market price equals marginal cost, which is one of the conditions for allocative efficiency.
An industry in which network effects and economies of scale play an important role will exhibit the reverse graphical results. The marginal cost schedule which represents the supply curve will be negatively sloped instead of horizontally or upward sloping. This is the result of the existence of economies of scale on the market. Network effects will lead to the phenomena that the demand curve is upward sloping as the network is growing. When more users are added to the network, the value of the network rises for existing and prospect users and for this reason consumers will have an increasing willingness to pay as the network successfully internalizes network effects. The schedule in figure 2 thus clearly depicts the particularities in high-technology markets in which with an increasing output the willingness to pay of consumers rises, while the unit costs of production fall. [11]
In this kind of markets the intersection of the demand and supply curves no longer represent a stable market outcome. To the left of the intersection point we can see that the marginal costs are higher than the willingness to pay by the consumers. This would imply that the market search for equilibrium would be pushed further to the left and finally reach it solution at the output of zero in which no market will exist at all. There exists however an alternative solution in this market. This outcome will be reached as soon as output moves beyond the point of intersection of the two curves. To the right of this point, known as the point of critical mass, consumers are willing to pay more for the product than the marginal cost of producing it. As the output is increased from this point the combination of network effects and economies of scale result in an ever increasing consumer surplus. The result of this will be that there is a tendency for networks which have reached the point of critical mass and which keep successfully internalizing network effects to grow further. In this simplified example consumers value bigger networks more.
1.3.2 Implications of Network Effects
In the law and economics literature two particular market failures have been recognized in relation to network markets: market tipping and resulting inefficient standards. Several authors argue these market failures demand for active intervention by competition authorities. In the general debate around high-technology markets I would coin this group of authors the neo-structuralists. Especially in Europe the competition authorities often find reasons to intervene in network markets, while their American colleagues might be a little bit more cautionary. Good reasons exist for caution as network effects and its related phenomena such as market tipping, first mover advantage and path dependence remain theoretically and empirically ambiguous topics in the literature. [12] In addition the welfare benefits of network effects must not be neglected. Intervention by competition authorities might thus not necessarily be an example of welfare enhancing policy.
1.3.2.1 Tipping markets
A thorough discussed topic in the literature is the phenomena of market tipping. This is related to the earlier discussed concept of critical mass. Tipping mainly relates to virtual networks instead of physical networks. [13] For this reason it is highly relevant for competition policy in high-technology markets. The example par excellence to explain the concept of tipping is the market for computer operating systems. [14] On a market competition between operating system for users exists. At a certain moment one operating system, by bundling it to PC's or maybe just by coincidence, gets ahead in the number of users. As a result of the higher number of users the operating system becomes more attractive for programmers to write applications for it, this effect is called an indirect network effect. As a result of more available software the operating system becomes even more attractive to users and for this reason again more attractive for programmers. This is a self-reinforcing process which is characterized by the term 'positive feedback effects'. At some point consumers see the operating system as a necessity because it has the greatest number of users and applications. At this point the market for operating systems has 'tipped'. When the market have reached its tipping point, and the mechanism is strong enough, market share will rapidly increase and a 'winner takes all' market outcome will result.
1.3.2.2 Inefficient market standard
A second market failure associated with network effects is that the choice of the final dominating network is not necessarily the most efficient or superior available network. This is the result of consumers getting locked in to an inferior network. [15] Alternative networks which represent more potential value to consumers are not able to displace the dominant network as they lack the size to internalize all the potential network effects. [16] It is thus of crucial importance to obtain the first-mover-advantage and become the dominant network as quickly as possible. In short, 'history matters' in network industries as the adoption of an inferior network is possible as a result of path-dependencies.
Consumers wanting to join the alternative in potential superior network are confronted with a collective action problem. The decision by consumers to join the alternative network entails a high level of switching costs. Examples of costs which have to be beard by consumers are sunk investments in equipment, learning effects, uncertainty regarding the quality of the new product and the exist out of former loyalty programmes. [17] The more consumers are using the same network, the higher switching costs can be, as the result of the loss of compatibility and the high opportunity costs it represents. In addition to the prohibitive costs of switching a high level of coordination is required between different consumer groups which have different expectations. Consumers are only able to switch to the superior network if they are able to coordinate successfully so that they can overcome the collective switching costs. The problem consumers are confronted with can be described by the famous prisoners dilemma.
TABEL BLZ 37
As is shown in the table both players could obtain the highest network benefits if they agree to join the network. However, as a result of coordination problems and the uncertainty regarding the question which network the other player will join, consumers will decide to stick to the inferior standard. This is represented by the lower right cell. [18] In the literature the heavenly debated example of markets ending up with inferior standards is the QWERTY keyboard which is regarded by several authors as inferior to the DVORKIN keyboard. [19]
Technological lock-in have been heavenly critised by several authors on theoretical and empiral grounds. [20] It seems also hard to disregard the decline in market share of once dominant product as WordPerfect, Paradox and Lotus 1-2-3, which have been replaced by superior products despite the existence of network effects at these markets. This examples prove consumers are able to escape an inefficiënt standard at one moment in time irrespective of the high switching costs and coordination problems.
This brief discussion on network effects makes clear they give rise to big policy debates on where to draw the standard of intervention. Detrimental effects of market failures should be prevented but on the other side the positive welfare effects of network effects should not be neglected. [21]
1.4 Durable goods
Goods sold at high-technology markets may have the characteristic that their usable life cycle tend to be long. Since information can be used and reused almost without limit, information goods can be used repeatedly. A typical example of a durable good is software of which the physical components do not depreciate over time. As a result of this the market power of a firm selling durable goods is constrained by its own installed-base. [22] A durable monopolist (also called a durapolist), even if it held a market share of 100%, might as a result of this not be able to exercise its market power to the full extent. This theoretical issue was raised by Coase and is known in the literature as the 'Coase Conjecture'. [23] The Coase conjecture may necessate the durapolist to price its good as low as the marginal cost of production. [24]
1.5 The role of intellectual property rights
Economic actors have to be rewarded for costly investment with a high level of risk. [25] Firms are allowed to reap the rewards of their innovation by giving them the right to exclude competitors from the use of their inventions. Especially in high-technology markets on which innovation plays a central role firms are often allowed to enjoy market power created by intellectual property rights. Intellectual property rights are of crucial importance to incentivize firms to innovate. Without the legal right to exclude competitors no firm would be willing to incur the high costs and risks associated with innovation. Instead, firms would wait and free ride on the inventions of other firms. In the extreme case this may lead to a market on which no innovation takes place at all. [26] Allowing firms to enjoy market power for a certain period of time does however also lead to social costs known as the 'monopoly-loss triangle'. Firms get compensated for their successful innovative efforts by allowing them to price their products above the competitive level and to produce below the competitive output level. The monopoly-loss triangle is the graphical representation of the unrealized welfare consumers could get from the output the monopolist does not produce. This creates the difficult tasks for policymakers to balance the gains of competition on the short term versus the long term welfare loss of reduced investments in innovation. [27] Policymakers should try not to favor the short term allocative efficiencies for the reason that they are much easier to quantify with a reasonable level of certainty than the long term welfare costs of stifled innovation. [28]
1.6 Compatibility and standardization
Many high-technology products require simultaneous functioning of several complementary parts. [29] For instance a personal computer, requires the simultaneous functioning of the physical computer itself, the operating system and the software used on it. [30] The extent to which the different complementary parts in a system are compatible with each other determines the kind of rivalry on the market. Rivalry can exist within a platform between different components (intra-system rivalry) or rivalry can exist between different platforms (inter-system competition). In the former competition will take place on price and quality, while also other considerations are relevant when competition takes place between platforms. [31] Market behavior thus is affected by compatibility and interoperability conditions on a particular market. In addition, there exists a reasonable risks dominant firms on the market for one component are able to leverage their market power to a complementary markets.
Compatibility requirements ask for technology standards set on the market to prevent degradation of the system. When no technology standard is set competition for the market and tipping markets may result in one firm dominating the market. Consumers who have to make investments to join a particular market are vulnerable to end up in a network which lacks widespread support among other consumers. By setting technology standards consumer coordination is facilitated to prevent this.
1.7 Dynamic competition
The defining feature of high-technology markets is innovation and technological change. [32] Firms in innovative industries often engage in a dynamic competitive process with each other in a series of winner-takes-all races. To obtain market control firms invest in the first race heavily in developing a superior product. The resulting dominant position of the industry leader is however temporarily as competitors in subsequent races are trying to leapfrog over the technologic capabilities developed by the dominant firm. [33] In other industries competition takes primarily place through price competition. [34] The dynamic nature of competition on high-technology markets poses the question if the standard framework of competition policy, which is based on a static notion of competition, is still appropriate. The dynamic nature of competition in high-technology markets and its implications for competition policy will be the central issue in this thesis and will be deeper analyzed in the subsequent chapters.
