点击查看或下载该文件

570 views

Published on

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
570
On SlideShare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
2
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

点击查看或下载该文件

  1. 1. Characteristics of Information Goods Pricing and Profit Yong Y. Sohn Dept. of Economics, Chonnam National University 300 Yongbong-Dong, Buk-Gu, Gwnagju 500-757, Korea +82 62 539-1549 yysohn@chonnam.ac.kr Abstract A monopolistic firm’s behavior of pricing its information good is modeled, provided that the firm produces upgraded versions continuously. It is shown that demand function of new invented information good has a reversed U-shape because of its network externality but that of its upgraded version is negatively sloped. A special feature of positively sloped demand curve of an information good is not always the case. Firms producing information goods are found not to always take the introductory pricing strategy. When a firm invents a good of a new concept, it is likely to take the strategy. But if the product is an upgraded one of its previous version and is compatible to its previous version, the firm would rather charge profit-maximizing price at the first moment upon the installed base left from the previous version and then lower the price to help secure enough customer base of the next version. It is also analyzed that an information good with network externality is likely to bring forth less profit to the firm than other ordinary goods since the information good has to pay attention to keep its customer base large. Key words: information good, introductory pricing, installed base 1
  2. 2. I. Introduction Generally, a firm producing an information good with network externality is known to take the introductory pricing strategy. To secure the installed base of customers, it sets price lower than its marginal cost at the first moment of introducing a new information good to the market. For example, a web browser, Navigator, allowed people to download it for free in the early 1990s and Microsoft Korea announced to have a plan to donate 1 million copies of MS Word97 Korean to middle and high schools in 1998. But, there are some other cases that indicate the introductory pricing may not be so general in information goods industry. A certain brand of the Intel mainframe computer started to sell high at the first period of its releasing and fell low at the last part of its life. Price of a software program such as a word processor was high at first period of its selling but turned low as the time of introducing its upgraded version gets closer. The Hangul word processor (HWP) was free in the very early 1990s when HWP newly popped up in the Korean word processor market but its upgrade version HWP3.0 was sold at a considerable price and protected from pirate copying. The firm producing HWP put very high price on its first Windows version, HWP95, and lowered it later on. These strategies of pricing information goods are more similar to those of pricing durable goods. Even if there are different types of pricing strategies in information goods sector, most studies on pricing strategies of information goods are focused on the introductory pricing. And hence students studying economics and ordinary people are likely to misunderstand the reality so that all of information goods would take the introductory pricing strategy. In the 1980s and 1990s, there had been many economists who proved the existence of the introductory pricing strategy in the information goods market.1 Katz and Shapiro(1985) showed that consumers were willing to pay more, with expecting higher demand in the future. Bensaid and Lense(1996) investigated the fact that a monopolist producing information good maximized its profit to set a low price at first and raise its price later, with the presumption that network externality brought about by learning by doing and by mouth transmission benefited only the consumers coming later to the 1 In the early 1970s, some papers, which dealt with pricing strategies of communication industry, illuminated the existence of introductory pricing. Rohlfs(1974) showed that a communication firm would charge higher price on a new customer than an old one since a new one would get higher utility with many other customers already connected in the service. 2
  3. 3. market. Cabal et al. (1999) also showed by using a dynamic model that duopolistic firms chose the introductory pricing strategy. As mentioned above, since there are many examples of information goods prices of which fall as time goes on, it is sensible to say that information goods are not always taking the introductory pricing. But there are rare academic works explaining different types of pricing in the industry, it is needed to investigate types of pricing strategies of firms involved in information goods business and factors affecting those. This paper will do the job. In the next section, I look into the characteristic features of demand for information goods. The section three shows two types of price discrimination in time of the monopolist, which produces an information good and its upgraded version continuously; the introductory pricing and price cutting in time, depending on the size of the installed base of its customers. The section four deals with actual pricing strategies of the monopolist. Conclusion and some suggestions are provided with in the final section. II. Characteristics of Demand for Information Goods A monopolistic firm produces an information good, provided that there is no threat of entry at least for a while. Developing a new information good is very costly but marginal cost of reproducing the good after its development is trivial, c ≥ 0 . The monopolist knows that a consumer gets more satisfaction from the good as the size of consumers using the same one. In consideration of the network externality, the firm tries to secure the critical size of customer base to survive. The firm develops and sells its original version of information good and its upgraded versions to the market in the future. The population of consumers is η . Consumers are heterogeneous in terms of their preference to the good and are uniformly distributed over x ∈ [0,1] . A consumer located close to zero has strong preference to the good whereas a consumer represented by the index close to one has low preference to the good. He is assumed to buy a unit of q p each version of the good. Let denote the quantity demanded at its price . Then the utility of the x-typed consumer is Ux = (1 − x)q e − p , (1) 3
  4. 4. e where q is the quantity demanded expected by the consumer. In the equation, utility is e proportional to q , which reflects network externality. Since a consumer of high preference is likely to buy the good ahead of that of low preference, the last purchaser must be the one that is indifferent between buying and ˆ ˆ not buying. It we denote this type of a person by x , then x satisfies qe − p x= ˆ qe . (2) All the consumers located over x ≤ x are going to buy the good but the rest of ˆ consumers located in the area, x > x , won’t but it. The number of consumers buying ˆ the good is q = ηx . ˆ Assuming perfect information, it holds that q = q = ηx . Substituting it into the e ˆ equation (2), demand function is derived by p = (1 − x)ηx . ˆ ˆ (3) This demand curve is drawn in <Figure 1> as a reversed hyperbola, whose intercepts with x-axis are x = {0,1} .2 ˆ As seen in <Figure 1>, consumers are willing to pay more as the number of consumers grows until portion of buying consumers reaches the midpoint of population, Over the range, network externality outride price effect in demand for the information good. But willingness to pay is going down after the midpoint as price effect is greater than network externality. This graph makes it clear that demand for an information good will not increase indefinitely owing to ever-exerting network externality. If the monopolist sets price at p 0 = c , the quantities demanded satisfying the equation (3) are3 2 The idea up to this point was shown in Rohlfs(1974), and Shy(2001) simplified it clearly. Economides and Himmelberg(19995) proved, with the fulfilled expectation assumption that this demand function also existed in perfect competitive market. 3 In reality, the exact quantity demanded is ηx but for simplicity x is used instead here, assuming η = 1 . 4
  5. 5. 1 − (1 − 4c) 1 + (1 − 4c ) x0 = ˆL , x0 = ˆH . (4) 2 2 L ˆ The point x0 is an unstable equilibrium. If the size of the consumers purchasing the L ˆ good were less than x0 , then demand for the good would shrink to disappear. However, L H ˆ ˆ if demand for the good is above x0 even slightly, then it increases in itself to x0 . In L ˆ the sense, x0 is a critical base of consumers for the firm’s survival. Over the range L ˆ where the number of consumers happens to less than x0 , the firm had better subsidize L ˆ consumers to buy the good until demand is above x0 . <Figure I> Demand Curve of an Information Good When the monopolistic firm manages to induce demand for its good to be greater than ˆL ˆH x0 , the demand is going to increase in itself to x0 . During the period where demand 5
  6. 6. L H ˆ ˆ expands from x0 to x0 , the firm can charge profit-maximizing price. The final demand H ˆ for the original version of the good is x0 . H ˆ Now, with x0 to be its installed base of customers, demand function of the next upgraded version, which is compatible backward with the original version, would be p = (1 − x) x 0 = x 0 − x ⋅ x 0 , ˆH ˆH ˆH (5) which is a negatively sloped straight line in terms of x . To derive the negatively sloped demand curve of the upgraded version, it needs to be compatible with its old version. And backward compatibility is common in information goods industry since the late 1980s. The special feature of the demand curve deserving to draw attention is that it is a negatively sloped line even with network externality. With the installed base secured enough, a consumer’s willingness to pay is proportional to its preference to the good. In this negatively sloped demand curve, network externality is now affecting the slope and vertical intercept of the demand curve. Even if it would be the case, many economists have presumed that demand functions of information goods with network externality are like ones specified in the equation (3), and shown enthusiastically the critical value and introductory pricing strategy in the information goods market. (Rohlfs(1974); Shy(2001)) One of the reasons is that they have concerned only about the prior demand for brand-new original information goods but not the demand for their upgraded versions, which have their own installed base of customers inherited by their respective previous versions. III. Two Types of Pricing Strategies One of the characteristics of information goods is the existence of economy of scale in the demand side, network externality. A firm producing an information good manages to take the installed base big enough to make demand for its product grow in itself. For the purpose, the firm charges low price with a small customer base and then raises it when the size of customer base passes the critical value. But as it is mentioned above, the firm wouldn’t take the introductory pricing if the good has its installed base to be 6
  7. 7. used. With enough installed base, its demand curve is the same as the ordinary good, negatively sloped in price. A firm producing an information good seems to be irrational on its pricing strategies: raising its price sometimes or pushing it down other times in terms of time. Those self- contradictory pricing behaviors are to be explained to be a logically rational choice of the firm. A monopolist having invented and sold a brand new product and its subsequent upgraded versions maximizes its long-run profit. If the firm producing an information good of strong network externality chooses its profit-maximizing price, then it is more probable for consumers to pay less for the good with the shrunk installed base caused by the monopolistic pricing. Therefore, the monopolist’s effort to maximize profit would be limited by its negative impact on the demand for its upgraded version. This feature needs to get proper explanation. Suppose that the monopolist places a new version on the market every two periods. A consumer buys at most a unit of each version. Even if a version of the good is sold for two periods, the size of its consumers is very important for the demand for the next version with the new version compatible backward with its last previous version. The firm is assumed to set price at each period and then two times for a version. As we have seen in the previous section, the larger demand for the first version is, the larger that for the second version, with high willingness to pay of consumers caused by its increased installed base. And hence the second version can bring forth high returns to the monopolist. For simplicity, assume that the Coase conjecture does not matter and that there is no time discount both to consumers and the firm. The cost of developing a version is sunk at the time of reproducing it as a commodity and need not be taken into consideration in the choice of the monopolist. Marginal cost of production of each version is assumed to be equal at c per unit. In the first period of the first version, the monopolist does its best to secure the critical size of installed base by taking the introductory pricing strategy. But a kind of subsidy to consumers given by the introductory pricing is related to the level of price chosen by the firm in the second period of the first version, since the critical size of L ˆ consumers, x0 , is directly bound to the price level in the second period. The higher the L ˆ price of the second period is, the higher x0 and the amount of the subsidy as a result. 7
  8. 8. L ˆ If price in the second period is set lower, then x0 and the amount of subsidy are consequently low. When the second period price is set at t, the subsidy given to consumers amounts to t ⋅ x0 , which is an increasing function of t. L ˆ There is another thing that the firm should concern about in choosing the second period price. That is its impact on the second period profit by way of its influence on the installed base of the second version since the quantity demanded in the second period of the first version is the installed base of the second version. There may be various ways of pricing, but here two extreme cases of them are investigated among them. First, the firm is assumed to set its price equal to its marginal cost in order to increase demand for the first version and then the installed base of the second version as much as possible. Second, the firm chooses the profit-maximizing price to get the monopolistic profit in the second period of the first version at the cost of installed base of the second version. (1) Marginal Cost Pricing Marginal cost of reproduction of the first version is c . Suppose the firm picks its price at the level of the marginal cost. Then it tries to secure the installed base of 1 − (1 − 4c) x1L* = (4-1) 2 by offering subsidy to consumers of strong preference to the good. If it sells them free, total amount of the subsidy is s1 = x1 ⋅ c . L* L* After getting the critical value of consumers, x1 , demand for the good rises automatically since the willingness to pay of consumers is higher than the price with the L* installed base of x1 . At last the quantity demanded of the first version is settled at 1 + (1 − 4c) x1H * = . (4-2) 2 8
  9. 9. Net revenue of the firm except for the subsidy is NR = c( x1 − x1L* ) . When the firm H* takes the marginal cost pricing, its loss from the first version amounts to the subsidy offered to consumers in its first period. As the firm markets its second version that is the upgraded version of its first one, the economic life of the first version ends. The second version is enabled to be compatible backward so that it can utilize the customers of the first version as its installed base for the second version. Therefore, demand for the second version is p = k 2 (1 − x) x1H * , (7) where k 2 stands for the technological advance of the second version. k 2 is not less than one, but it is assumed to be one from now on, just for simplicity. The monopolist exerts its monopoly power to choose price and output, which are respectively x1H * − c x1H * + c x = 1* p 1* = 2 x1H * , 2 , (8) 2 2 and profit is ( x1H * − c) 2 π 2* = 1 4 x1H * . (9) 1 If marginal cost is zero, profit is 4 . In the second period of the second version, with taking the residual demand into consideration, demand for the good is  (1 − x )( x1H * − x 1M ), x 1M ≤ x ≤ x1H * p=  (1 − x) x, 2 2 (10)  x1H * < x ≤ 1 At this point, the firm needs to decide whether to take profit-maximizing or marginal 9
  10. 10. cost pricing. If the third, further upgraded version of the second one, is put out to the market, it would have the demand function like the equation (6). Firm’s decision-making will be similar to the one of the second version. (2) Monopolistic Pricing Back to the second period of the first version, now suppose the firm acts as a monopoly instead of a price-taker. Since demand function is reversed U-shaped, profit function without fixed costs is now represented by π = ( p − c) = ((1 − x) x − c ) x . (11) Profit maximizing output and price are respectively 1 + 1 − 3c x1HM = , p1M = (1 − x1HM ) x1HM . (12) 3 1 − 1 − 3c Then profit is π 12 M = ( p1M − c)( x1HM − x1LM ) , where x1LM = . The firm earns 3 4 profit of 27 in the second period of the first version when marginal cost is zero. But the subsidy transferred to consumers to lure them up to the critical point of installed base in the first period is s1 = p1 ⋅ x1 M M LM . Hence profit of the first version as a whole is 4 π 1M = π 12 M − s1M and converges to 27 as c gets closer to zero, since the subsidy also gets closer to zero in that case. However, π 1M get smaller as c becomes larger and profit would be less than zero if c were greater than a certain point. As the upgraded version places on the market, the firm sets monopolistic price on the installed base secured in the first version. The monopolist could choose price at the level of marginal cost to expand customer base for the second version. But that strategy is not beneficial to the firm since residual demand for the second version in its second period plummets because of expanded first-period consumption. HM With the installed base x1 handed over by the first version, demand function of the 10
  11. 11. second version in its first period is p = k 2 (1 − x) x1HM . (13) As before it is assumed that k 2 = 1 , output and price maximizing profit are, respectively, 1 c x 1M = 2 − HM 2 2 x1 , (14) 1 + 3c + 1 − 3c p 1M = (1 − x 1M ) x1HM = 2 2 . (15) 6 1 Profit, π 2M = ( p 1M − c) x1HM , is close to 1 2 6 as c goes to zero. Now we turn to the second period of the second version. Demand for the good in the period is composed of the residual from the demand in the first period and a new one of the reversed U-shape, as seen in <Figure 2>.  (1 − x ) x1HM , x 1M ≤ x ≤ x1HM p=  (1 − x) x, 2 (16)  x1HM < x ≤ 1 11
  12. 12. <Figure 2> Demand Curve of the Second Version If the firm maximizes profit on the base of the residual demand, then the type of the consumer buying at the last moment is 3 3c x2 M = 2 − HM 4 4 x1 , (17) and price is 1 + 9c + 1 − 3c p 2 M = (1 − x 2 M ) x1HM = 2 2 (18) 12 1 In this period, profit, π 2 M = ( p 2 M − c )( x 2 M − x 1M ) , converges to 2 2 2 2 24 as c goes close to 0. If the firm sets price to maximize profit on the reversed U-shaped demand curve, M then it is the same as p1 and then profit earned in the second period of the second 12
  13. 13. 1 version, π 2 = ( p1M − c)( x1HM − x 1M ) , gets closer to HM 2 27 as c converges to zero. But when the firm plans to market the third version, the upgraded one of the second 2M version, it is going to face a problem that profit on the installed base of x 2 is much smaller than profit on the base of x HM 1 since x 2M 2 <xHM 1 . Therefore, it is better for the M HM firm to set price at p and to secure the customer base of x1 . 1 But there is the way in which the firm gets the highest profit from the third version. If it prices its good at c, then it would get more profit because of the expanded installed base of customers. For the versions of fourth and on, the firm acts as a monopolist in the first period and as a price-taker in the second period. IV. Pricing Strategies of a Firm Producing Information Goods (1) Firm’s life ends at the first version With the monopolist selling the first version and no more, it maximizes its profit by supporting consumers with low price in the first period and by setting monopolistic price in the second period. It is obvious that the firm sets the monopolistic price in the second period since it cannot recover the subsidy and get net profit by setting price at c . And it is inevitable for the firm to support consumers to get a critical size of consumers in the first period. If, in the first period, the firm fixes the size of subsidy independent of price of the second period, equilibrium price in the second period is x1 HM and then profit π 1HM which 4 is 27 with zero marginal cost. But the subsidy to consumers is so that the firm needs to serious to choose price, since the higher price in the second period, the greater the size of the subsidy in the first period. Suppose that the firm supports a consumer by the amount of price minus his willingness to pay. Then, as seen in <Figure 3>, production under the marginal cost pricing incurs variable cost of A+B+C+D+E, which is the same as its revenue. But total subsidy to consumers is equal to the area A and then the firm loses by the amount 13
  14. 14. of A. On the contrary, when the firm takes monopolistic pricing strategy, it earns A+B+C+D+F+G+H as its revenue and pays A+B+C+D as cost and A+F as subsidy to consumers of high preference. The monopolistic firm can get net profit when G+H is greater than A+F. It is possible for it to get loss, depending on the marginal cost of reproduction of the good. <Figure 3> Two Types of Pricing of the First Version (2) Firm’s life ends at the second version Even with the successful business in the first version, the firm won’t stop doing business at the point but continues to develop and market a new good, an upgraded version of its first one. If the functional savvy of the upgraded good were taken into account, its price would be much higher than its real one. But price of the new version is usually almost same as the first version in reality, probably according to a certain improvement of an operational technology. Following the reality, it is assumed that demand function of the new upgraded version coincides with that of the first version. (That is, k i = 1, i = 2,3,..., t ) 2M In the last period of the firm’s life, the firm charges the monopolistic price of p 2 . It is important to note that the firm would not take the marginal cost pricing in the first 14
  15. 15. period of the second version, since profit in the first period is low and furthermore that in the second period is also low with its very small residual demand. And hence the firm takes profit-maximizing price based on the installed base left from the first version, which relies on the strategy taken by the firm in the second period of the first version. If it makes high profit by monopolistic pricing from the first version, the firm is forced to be satisfied at small profit in the next period because of lowered installed base, and vice versa. When marginal cost of the firm is zero, then sum of profits gained from monopolistic pricing in the second period of the first version and the first period of the 4 1 second version is π 1MH + π 2M = 1 + 27 6 , but that from the marginal cost pricing over 1 the same periods is π 12* + π 2* = 0 + , as seen in the previous sections. But as c is 1 4 greater than zero, net profit excluding supporting expenditure in the first period of the first version shrinks fast enough to bring the sum of profits brought forth from the monopolistic pricing in its second period to be negative. Finally, it can be said that the firm’s decision on which pricing strategy to take in the second period of the first version depends on whether (π 2 − s1 ) − (π 2 + π 1HM − s M ) is greater than zero or not. 1* 1M If it is greater than zero, the firm will price its good at the marginal cost level. (3) Firm’s life never ends If the firm continues to produce the third version, fourth version, and so on, we need to keep an eye on the firm’s decision-making over the second periods of the second version and the first period of the third version. There is a tradeoff once again between profit of the second period of the previous version and that of the following period of the upgraded version. If the firm decides to take high profit in the second period of the second version, then the equation (17) and the following show that sum of profits in the two neighboring 1 1 periods is π 2HM + π 3 M , which is 1 + 27 6 with zero marginal cost. On the contrary, marginal cost pricing in the second period of the second version and monopolistic 1 pricing in the following period makes sum of profit of π 2 * + π 3* , which is 0 + 2 1 4 when c is equal to zero. Therefore, the firm wants to price at the marginal cost level in the second period of the second version and then to maximize profit in the next period on the large installed base left from the previous period. This type of pricing strategies 15
  16. 16. over the second period of any version and the first period of the following upgraded version is likely to be taken by the monopoly on and on over more higher versions. Summarizing the investigation up to this point, we have the following proposition without any formal proof. Proposition 1. Suppose that a monopolistic firm develops and markets backward compatible upgraded versions continuously. The firm takes the introductory pricing only for the first version, subsidizing consumers in its first period and pricing monopolistically in the second period. From the second version and on, the firm charges monopolistic price in the first period and price of marginal cost in the second period, except just for a case where the firm decides to exit. In the exiting period, it acts as a monopolist. Now it is clear that firms producing information goods do not always take the introductory pricing strategy. When a firm invents a good of a new concept, it is likely to take the strategy. But if a product is an upgraded one of its previous version and is compatible to its previous version, then it is not necessary for the firm to charge low price at the first period of marketing it. The firm would rather charge profit-maximizing price at first upon the installed base left down from the previous version and then lower the price to help secure enough customer base of the next version. It is probable that the result we have got would change, depending on the degree of discounting of participants in the trading and on the distribution of the consumers. High degree of discounting causes the firm to be more interested in monopolistic pricing now rather than waiting for high profit tomorrow. And if consumers are distributed to have a mass around the middle point of the full range, then the firm has much less interest in discriminating price in time except for introductory pricing of the first version. Now we turn our attention to profit. It is easy to guess that, without network externality, demand function of an information good be p = 1 − x , where x stands for the type of consumers uniformly distributed over a unit range. The monopolistic firm just charges price maximizing its profit from the first moment of trading, instead of supporting customers who want to buy its good. As for later versions, there is no need to sacrifice its earnings to maintain its customer base meaningfully large. From those reasoning, we can suggest the nest proposition. 16
  17. 17. Proposition 2. An information good with network externality can get less profit than other ordinary goods since the information good has to sacrifice some of its profit to keep its customer base as large as possible. There is only an exception of the last version in the firm’s lifetime that enables the firm to get monopolistic profit. One thing to be mentioned here is that switching cost should be taken into consideration to investigate profit of a firm producing an information good. Customers of an information good are likely to be locked in the good and need to burden high switching cost when they try to move to another alternative good. The incumbent firm can take advantage of the switching cost to charge higher price and get more profit. (Shapiro and Varian(2000)) However, the monopolistic market structure has been assumed in this paper, we don’t need to consider in this direction anymore. V. Conclusion A firm producing an information good characterized by network externality is generally known to take the introductory pricing strategy. To secure the installed base of customers, it sets price lower than its marginal cost at the first moment of marketing a new information good. But the introductory pricing is not so common in reality; there are some cases where firm prices high at the first time of marketing a specific version of an information good and then cuts price just ahead of the birth of its upgraded version. Therefore, we have investigated the pricing strategies of an information good producing firm and analyzed factors that determine those pricing strategies. It is shown by a simple model that demand function of a new invented information good has a reversed U-shape because of its network externality but that of its upgraded version is negatively sloped. That is, it means that the special feature of demand for an information good that the demand curve is positively sloped in price holds true only partially. We model a monopolistic firm’s behavior of pricing its information good, provided that the firm produces upgraded versions continuously. It is made clear that firms producing information goods do not always take the introductory pricing strategy. When a firm invents a good of a new concept, it is likely to take the strategy. But if the product is an upgraded one of its previous version and is compatible to its previous version, then it is not necessary for the firm to charge low price at the first period of marketing it. The firm would rather charge profit-maximizing price at the first moment upon the installed 17
  18. 18. base left from the previous version and then lower the price to help secure enough customer base of the next version. It is also analyzed that an information good with network externality is likely to bring forth less profit than other ordinary goods since the information good has to pay attention to keep its customer base large. This paper is successful to show that the introductory pricing in information goods market is not always common and that profit of the firm producing information goods is likely to lower than other goods, on the same condition. But there are other issues untouched or not touched appropriately, such as distribution of consumers, potential entrants, and switching cost. The model here will be extended to encompass those in the near future. References Bagnoli, Mark, Stephen W. Salant, and Joseph E. Swierzbonski (1989), "Durable-Goods Monopoly with Discrete Demand," Journal of Political Economy 97(61), 1459-1478. Bensaid, Bernard and Jean-Philippe Lense (1996), "Dynamic Monopoly Pricing with Network Externalities," International Journal of Industrial Organization 14: 837-855. Bulow, Jeremy I. (1982), "Durable Goods Monopolists," Journal of Political Economy 90 (2), 314-332. David, Paul A. (1985), "Clio and the Economics of QWERTY," American Economic Review 75, 332-337. Economides, Nicholas (1995), "Network Externalities, Complementarities, and Invitation to Enter," European Journal of Political Economy 12, 211-232. Economides, Nicholas and Charles Himmelberg (1995), "Critical Mass and Network Evolution in Telecommunications," in Gelrald W. Broch ed., Toward a Competitive Telecommunications Industry: Selected Papers from the 1994 Telecommunications Policy Research Conference. Telecommun. Coase, R. H. (!972), "Durability and Monopoly," Journal of Law and Economics 15, 143-149. Cabral, Luis M. B., David J. Salant, and Glenn A. Woroch (1999), 갡onopoly Pricing with Network Externalities," International Journal of Industrial Organization 17: 199-214. Gul, Faruk, Hugo Sonnenschein, and Robert Wilson (1986), "Foundations of Dynamic Monopoly and the Coase Conjecture," Journal of Economic Theory 39, 155-190. Katz, Michael and Carl Shapiro (1985), "Network Externalities, Competition, and 18
  19. 19. Compatibility," American Economic Review 75 (3): 424-440. Klein, Benjamin (2001) "The Microsoft Case: What Can a Dominant Firm Do to Defend Its Market Position?" Journal of Economic Perspective 15(2) 45-62. Lee, Sangho, and Yong Y. Sohn (2001), Principles of Competition in the Cyber Market, (in Korean) Sigmainsightdotcom; Seoul, Korea. Liebowitz, Stan J. and Stephen E. Margolis (1995), "Path Dependence, Lock-in and History," Journal of Law, Economics, and Organization 11, 205-226. ___________________________________________ (1999), Winners, Losers & Microsoft, The Independent Institute: Oakland, CA. Rohlfs, Jeffrey (1974), "A Theory of Interdependent Demand for Communications Service," Bell Journal of Economics 5 (1): 16-37. Shapiro, Carl and Hal R. Varian (1999), Information Rules: A Strategic Guide to Network Economy, Harvard Business School Press: Boston, MA. Sohn, Yong Y. (2001), “Pricing Strategy of Information Goods,”(in Korean) Journal of the Sungkok Academic & Cultural Foundation 32 (2), 47-90. Shy, Oz ((1995), Industrial Organization: Theory and Application, The MIT Press: Cambridge, MA. 19

×