Resale Price Maintenance Without Services∗ Wendy Mao The University of Hong Kong Email: [email protected]

Travis Ng The Chinese University of Hong Kong Email: [email protected]

April 12, 2016

Abstract A valid defense of the use of resale price maintenance (RPM) in court now must explain how the RPM induces more services (or other non-pricing dimensions customers value). Often, however, services either do not play a significant role, or their significance is subject to debate. This happened in the 2013 China’s RPM case of Rainbow v. Johnson & Johnson (J&J); J&J sold medical equipment to China’s hospitals through various distributors but how significant was the distributors’ services is debatable. We argue that J&J’s RPM, even if it never elicited more services from its distributors, can be pro-competitive. We explain why the RPM did help J&J compete with other brands. We formalize our explanation in a model based on Winter (1993). The model has one upstream and two downstream firms. Unlike Winter (1993), the downstream firms only choose their prices but nothing else. Without services, there exist equilibria under which one downstream firm would under-price when the upstream firm offers a uniform wholesale price together with a lump-sum transfer. The upstream firm can curb such an underpricing incentive by imposing a minimum RPM. In a numerical example, we show that the use of RPM without services can increase the welfare of both the whole supply chain and the consumers. Keywords: Resale price maintenance, free-riding, services hypotheses, antitrust. JEL Classifications: L42, L51

∗ Preliminary and incomplete.

We thank the invaluable comments from Kim-sau Chung, Jiahua Che, and Zhigang Tao. All remaining errors are our own.

1.

Introduction “The reduction in dealers’ rivalry in the price dimension is just the tool the manufacturer uses to induce greater competition in the service dimension.” Frank H. Easterbook. (1984, p. 156).

This paper offers an explanation for the use of resale price maintenance (RPM) by an upstream firm when the demand of its downstream firms only depends on prices. We rationalize the use of RPM as shown in the right oval of figure 1, for which the set of business scenarios do not overlap with the right circle. This right circle represents the set of business scenarios under which the demand of its downstream firms depends on both prices and services (for which, like Judge Easterbrook did, refers to any non-price dimensions customers value). Figure 1: The explanations for RPM the court may accept Universe: Business scenarios

Explanations when downstream firms’ demand depends ONLY on prices [Our paper]

Free-riding explanations [Telser(60)]

Non-free-riding explanations [e.g., Winter (93), Mathewson-Winter (84)]

Note: The shaded areas represent the sets of business scenarios the court may accept in defending the use of RPM. The darker is the shaded area, the more likely it is for the court to accept it as a defense for the use of RPM.

Telser (1960) has greatly influenced how the court adjudicates RPM cases. His argument is that if downstream firms’ demand depends on prices and services, the downstream firms do not always offer enough services. For instance, if one downstream firm offers high level of services, customers can benefit from its services but buy from another downstream firm who offers no services. Customers would do so because the latter would have lower costs and therefore can afford to charge a lower price. The upstream firm can impose a minimum RPM to curb this free-riding

−1−

problem; it prevents any downstream firm that does not offer enough services from lowering the price to steal customers away from those that do offer enough services. This free-riding explanation dominated the literature, as commented by Mathewson and Winter (1998), “Too much of the literature on RPM tries to force the RPM “square cases” into the free-riding “round holes”.” Naturally, the explanation also dominated the court, as shown in Judge Breyer’s dissent in Leegin1 : “Petitioner and some amici have also presented us with newer studies that show that resale price maintenance sometimes brings consumer benefits..... But the proponents of a per se rule have always conceded as much. What is remarkable about the majority’s arguments is that nothing in this respect is new.... The one arguable exception consists of the majority’s claim that, even absent free riding, resale price maintenance may be the most efficient way to expand the manufacturer’s market share by inducing the retailer’s performance and allowing it to use its own initiative and experience in providing valuable services. I cannot count this as an exception, however, because I do not understand how, in the absence of free riding (and assuming competitiveness), an established producer would need resale price maintenance. Why, on these assumptions, would a dealer not expand its market share as best that dealer sees fit, obtaining appropriate payment from consumers in the process? There may be an answer to this question. But I have not seen it. And I do not think that we should place significant weight upon justifications that the parties do not explain with sufficient clarity for a generalist judge to understand.” Leegin Creative Leather Products, Inc. v. PSKS, Inc., 551 U.S. 877 (2007) (5-4 decision) (Breyer, J., dissenting). Non-free-riding explanations do exist. Winter (1993) shows one. Using a model with one monopoly upstream firm and two symmetric downstream firms that are horizontally differentiated. In Winter’s (1993) model, free-riding is just one special case of a more general problem in which downstream firms have a bias in favor of price competition while putting insufficient emphasis on non-price (services) competition, relative to the whole supply chain. As Klein (2009) nicely summarizes, there are nonfree-riding reasons that lead to such a bias, for which a minimum RPM can solve. Perhaps they are less intuitive explanations, the court has shown some resistance in accepting them (as indicated by Judge Breyer’s dissent in Leegin, thus the grey rather than dark shaded area in figure 1). 1 Leegin

Creative Leather Products, Inc. v. PSKS, Inc., 551 U.S. 877 (2007)), is a US antitrust case in which the United States Supreme Court reversed the 96-year-old doctrine that vertical price restraints were illegal per se under Section 1 of the Sherman Act, replacing the older doctrine with the rule of reason.

−2−

However, any explanation of the use of RPM based on services, free-riding or not, must answer one question: how significant are the additional services that a RPM is supposed to induce? Putting it in the language of antitrust, whether the procompetitive services-inducing effects of the accused RPM conduct outweighs the anticompetitive effects. In the context of supermarket, it means how much more sales the supermarket would get if the cashiers are, say, a minute faster, or the staffs are a little friendlier to the customers. In the context of selling jeans, evidence should be presented on how important it is to the sales of jeans if customers can try them out in a fitting room. It is not always easy to answer these questions, and different sides do often disagree on the assumptions/models used to come up with an estimate. Can RPM be justified in court even if the significance of services is debatable? To address this problem, we look into the business scenarios under which the downstream firms’ demand depends only on prices. We then ask two positive questions: Under such a condition, why would a contract between an upstream and its downstream firms without vertical restraints fail to align their incentives? How would RPM resolve such incentive conflicts? Our motivation comes from the 2013 litigation against the use of RPM by Johnson and Johnson (J&J, the upstream firm) in China for its staplers and suturing products sold to Chinese hospitals through its distributors (downstream firms). One of its distributors, Rainbow, sued J&J for the use of RPM. The Shanghai Higher Court ruled against J&J, who paid mere USD87,000 damages. Since the case was the first private litigation in China involving vertical agreements, it served as an important source for businesses to learn how China’s court enforces the Anti-Monopoly Law.2 Staplers and suturing products, together with examination gloves and sutures, and more specialized products such as orthopedic implants and pace makers, are categorized as clinical supplies. Staplers and suturing products are consumables and do not require state-of-the-art medical technologies to make. It is reasonable to say that a doctor is unprofessional if she is incapable of closing a skin wound simply because her favorite brand is out-of-stock. Some British hospitals, for instance, had changed their brands recently too.3 These products have also been sold for many decades. It is not obvious how distributors’ services are significant in determining the demand for staplers and suturing products from hospitals. As such, both sides did debate in the court on the significance of services. Section 2 discusses the case; we highlight 2 Two

hard liquor firms, Moutai and Wuliangye, had lost their RPM cases before J&J lost its case. Their cases were not private litigations and very limited court materials were given to the public. 3 The August 2013 Publication by the Department of Health in England entitled “Better Procurement Better Value Better Care: A Procurement Development Programme for the NHS” has indicated so. On page 10 of their document, it indicates that some of their hospitals had “ recently saved over £100,000 by switching to a lower cost alternative to the brand leader.”

−3−

the unimportance of relying on any services (or any non-price) dimensions in making a case for the use of RPM. In section 3, we present the simplest possible model in which a monopolist upstream firm sells to its downstream firms. We deliberately take away any non-price dimensions, i.e., the demand curves faced by the downstream firms only depend on prices but nothing else. We show the condition under which the use of RPM can better align the interests of the whole supply chain relative to a simple contract with only a wholesale price and a lump-sum transfer. Therefore, it explains why it is unnecessary for the defense of the use of RPM to state the types of services the RPM would induce as well as estimating their significance. Theoretical results A few results stand out. First, it is necessary for the downstream firms to be asymmetric for a simple contract to fail to align the incentives between the upstream firm and the downstream firms. Second, when the downstream firms are asymmetric, either offering different wholesale prices or a vertical price floor can induce the downstream firms to set prices that maximize the whole supply-chain’s profit. Third, other than practical legal concerns or bootlegging among downstream firms, charging different wholesale prices can be inferior to setting a vertical price floor in a plausible theoretical setting within which the upstream firm cannot be certain of the types of each of the downstream firms. Forth, while customer-stealing is necessary in Winter (1993) for RPM to work, our model does not depend on the possibility that customers can walk away from one downstream firm and buy from the other. The intuition Why would not asymmetric downstream firms charge the “right” prices when they face the same wholesale price? When setting their prices, a downstream firm does not internalize the effects of its pricing on the upstream firm’s profit. It also does not internalize the horizontal externalities - the effects of its pricing on the other downstream firm’s profit. For instance, if downstream firm A lowers its own price to attract more sales from customers in region A, downstream firm B may have a slower sale in region B. Perhaps it is because some customers in region B goes to buy from firm A (customer-stealing), or customers in region B simply buy less from firm B without buying from firm A (no customer-stealing). An RPM has a potential to curb these pricing mis-incentives. The gist of the paper, however, is to understand why there exists business scenarios under which firm A’s lowering of its prices slows down firm B’s sales. While the J&J case in Section 2 does so in business-to-business (B-to-B) transactions, we argue in section 6 that it is possible for business-to-consumers (B-to-C) transactions too using a

−4−

European case. We discuss the J&J case in Section 2. Section 3 develops a simple model to justify the use of RPM without services, which also includes a simple numerical example. Section 4 illustrates the advantages of the strategy adopting RPM with a single wholesale price over different wholesale prices strategy. Section 5 analyzes the welfare change from one equilibrium contract (without RPM) to another (with RPM). Section 6 discusses our theory in the context of a European case. We review the relevant literature on RPM in Appendix A.

2.

The RPM case against Johnson & Johnson in China

In any business scenarios under which the significance of services in determining the demand of the products cannot be ascertained, any defense for RPM with services in court is weak. The background In the 2013 case of Beijing Ruibang Yonghe Science and Technology Trade Company (“Rainbow”) v. Johnson & Johnson Medical (China) Ltd. (“J&J”), this problem appears. The products of J&J in this case are medical products: staplers and suturing products. These medical consumables, used mainly in surgeries to close skin wounds, have been used for many decades and there are a variety of brands for hospitals to choose from.4 It would be unprofessional for a doctor not to be able to close a skin wound simply because she does not have any experience with a particular brand of staplers and suturing products. Of course, above and beyond purely technological view, medical doctors can and do have their own preferences for their favorite brands. J&J is a major player in this product market. Since around 1998, J&J has been selling these products to China’s hospitals through its distributors, each of whom enjoys an exclusive territory. Rainbow, who is in charge of Beijing, is one of J&J’s distributors. J&J charges a wholesale price to its distributors and authorizes these distributors to set their own resale prices, but J&J also imposes a vertical price floor.5 In 2008, Rainbow did offer to sell to hospitals in Beijing at a price lower than the vertical price floor. As such, J&J first restricted Rainbow to sell only a subset of hospitals and eventually stopped offering any products to Rainbow. Rainbow claimed to have suffered great loss and sued J&J for the anti-competitive use of RPM. J&J lost. 4 As

indicated in footnote 3, some British hospitals did switch brands. case materials in this case do not show the price. We, however, have found that the publiclyavailable documents of the 2014 case involving the same plaintiff and defendant do provide the price information of J&J’s suturing products. The prices are the following: Benchmark Price: RMB 4,650; Wholesale price: RMB 3348 (72% of the benchmark price); Before the reform of the hospital procurement (we refer to it as “bidding”): The resale price of Rainbow: RMB 4,604 (99% of the benchmark price); After “bidding”: the resale price of Rainbow: RMB 4200 (90% of the benchmark price). 5 The

−5−

The Shanghai Higher Court opined that the staplers and suturing products market was not that competitive: it had a sufficiently high entry barrier, including the technology of the product, the preferences of doctors on the brands they like, and the relationship between the hospitals (buyers) and the distributors (sellers). Based on J&J’s 20.4% market share and the fact that its price has been maintained for almost 15 years, the court opined that J&J possessed sufficiently strong market power. Such market power enabled J&J to set a high price; this high price could not be defended by services explanations and free-riding explanations because the court was not convinced that the product requires much promotion, after-sale services, etc, from the distributors. As such, the RPM could not be seen as pro-competitive. The anticompetitive effects of restricting price competition must outweigh any procompetitive effect. The minimum RPM was ruled as illegal. Were the distributors’ services significant? This is a case in which both sides held polar opposite views on the significance of the services given by the distributors.6 It was fair to say that there is no consensus in court as to how significant are any services in determining the sales of the products. The plaintiff (Rainbow) argued in court that no services can be claimed to be significant except those from J&J, while the defendant (J&J) argued that the distributors’ sales services, delivery services, after-sales services are extremely important and that a vertical price floor would induce the distributors to exert more of these services. Different from the opinion of the Shanghai Higher Court, we would not argue that hospitals (end-buyers and end-users) rarely receive any services or promotion from J&J’s distributors when purchasing these medical products. But given the substitutability among the different brands, and the fact that these medical consumables have been used for so many years, it is hard to argue that any services given would be decisive in determining the demand. Even if they are decisive factors, it is hard to determine how significant they are. Our explanation (not used in the court) We argue that each distributor’s price not only influences their own sales, but also the sales of the other Chinese markets of J&J. This is so even if the distributors have their own exclusive territories (i.e., no customer stealing). Rainbow sold J&J’s products at prices lower than the price floor, hoping to grab a bigger market share among Beijing’s hospitals. We argue that Rainbow’s lowering its prices could and should lower the sales of J&J’s staplers and suturing products in other 6 For instance, the defending lawyers and economist argued that after-sales services by the downstream distributors were important, but Rainbow’s legal team disagreed. Instead, they argued that training and promotions done by J&J were much more decisive a factor in determining their sales.

−6−

Chinese hospital markets. Let us illustrate why so with an example. Think about a hospital in Chongqing, pretty far away from Beijing. Why a hospital in Chongqing would check and care about the price paid by Beijing’s hospitals? Put it in another way, if J&J’s distributor in Chongqing has already offered a reasonable price relative to other brands available in Chongqing, why would the Chongqing hospital care at all about the price paid by a Beijing hospital? We may put ourselves in the shoes of the procurement officer of a hospital rather than the hospital as one decision-making unit. Such an officer is responsible for buying medical inputs, including the staplers and suturing products. The officer does not necessarily have to buy from J&J on behalf of the hospital; there are other brands available. And his job is to buy reasonable products to maintain the hospital services at reasonable prices. Most likely for checks and balances, he would authorize the hospital accounting department to pay for the suppliers he picks. The accounting department’s job is to make sure the transactions are accurate and records are kept. The top management’s role is to make sure the prices authorized by the procurement officers are reasonable and the products satisfactory. The only way for the top management to know, however, is to know the prices other hospitals paying for similar products. In modeling procurement corruption, Burguet and Che (2004), they begin by citing the seriousness of procurement corruption: “Bribe taking in competitive procurement, whether public or private, is widespread. During the first half of the 1970s, more than 450 U.S. companies, 117 of which were listed in the Fortune 500, had made over $400 million in questionable payments to foreign concerns, and the U.S. firms allegedly lost nearly 100 foreign contracts worth $45 billion to foreign competitors through graft in 1994-1995.” The procurement officer would therefore face the risk of being scrutinized if he authorizes the hospital accounting department to pay J&J a price higher than what other hospitals are paying for similar products. This is true even if he cannot switch from buying from J&J’s Chongqing distributor to J&J’s other distributors. Authorizing the accounting department to pay a price higher than do other hospitals is understandable if the products come with complementary services. The more important is the service dimensions, the less likely it is for the procurement officer to get in trouble even if the price he authorizes the hospital to pay is relatively high. If this part is understandable, we then have a rather weird situation: The less significant are services to the products, the more concerned the procurement officers would be if there is a higher chance that other hospitals pay a lower price.

−7−

Paying a higher price raises the eyebrows of the top management, more so if volume or services cannot be an explanation of the procurement officer. The top management would wonder if there are any secret under-the-table deals between the procurement officer and the supplier, or whether illegal kick-backs are involved. Their concerns are not without reasons because the U.S. Foreign Corrupt Practices Act of 1977 (FCPA) was enacted as a result of foreign corruption practices firms did outside of the U.S.7 Why might RPM be a competitive strategy of J&J in this case? To compete, J&J has an incentive to disallow any of its distributor to under-price. Ironically, J&J should have less incentives to do so and allow a large dispersion of prices among the different regions of China if J&J’s products’ sales depend more on services of its distributors. In the case of staplers and suturing products, if it cannot be easily argued that services are significant, then the incentives of the procurement officers to avoid getting into trouble is a concern. If any given procurement officer thinks that it is more likely for another hospital to buy similar products from any distributors of J&J at a lower price than he can negotiate, he would switch to other “safer” brands that would less likely to get him into trouble. J&J’s vertical price floor is a solution to this problem. A minimum RPM, even known to both the procurement officers and all its distributors, seems a reasonable strategy for J&J to compete with other brands. The price floor potentially clears the doubt of the procurement officers, making it easier for the procurement officers to select J&J. This is so not in spite of the insignificance of services, but because of the insignificance of services. Although Orbach (2008, 2010) has argued that prices signal image of a brand, an explanation of RPM that does not need to have services, our explanation does not rely on it. Prices do not signal brand image here for J&J because the medical doctors, the hospitals, and the procurement officers do know that J&J has long been an established brand. 7 Reading

the publicly-available documents of this private litigation, it is surprising to us that there was not much mentioning of the actual procurement practices of China’s hospitals. We therefore investigated on the actual procurement practices of the China’s hospitals (summarized in B) and we are convinced that the procurement processes were complicated enough. Checks and balances are emphasized in official documents to combat bribery and corruption but the actual implementation is not something we can easily tell. Neither the plaintiff nor the defendant gave any details in the private litigation that seems to us to be crucial to the understanding of the use of RPM leaves us wonder why this is so. We wonder if openly talking about the procurement procedures would expose them unnecessary troubles.

−8−

3.

The model

3.1.

A simple numerical example

Consider an upstream firm (denoted M) selling a product through two downstream firms (denoted R1 and R2). Their demand functions are: D1 = 30 − 4p1 + p2 ,

(1)

D2 = 30 − 3p2 + 2p1 .

(2)

Denote w as the wholesale price M charges to them and Ai as the transfer fees, for i = 1, 2. Their profit functions are π1 = ( p1 − w) D1 − A1 ,

(3)

π2 = ( p2 − w) D2 − A2 .

(4)

Profit-maximization results in the following pricing rules:8 1 w+ 2 1 p2 ( p1 , w ) = w+ 2 p1 ( p2 , w ) =

1 15 p2 + , 8 4 1 p1 + 5. 3

(7) (8)

The whole supply chain, however, maximizes: Π = ( p1 − c) D1 + ( p2 − c) D2 ,

(9)

where c is M’s marginal cost. Without loss of generality, we set c = 1. Profitmaximization results in the following the optimal prices:9 p1∗ = 8 The

(12)

pricing rules for each R comes from the following first order conditions:

∂π1 ∂p1 ∂π2 ∂p2 9 The

96 ≈ 7.385, 13

= 30 − 8p1 + p2 + 4w = 0,

(5)

= 30 − 6p2 + 2p1 + 3w = 0.

(6)

optimal prices comes from the following first order conditions:

∂Π ∂p1 ∂Π ∂p2

= 30 − 8p1 + 3p2 + 2c = 0,

(10)

= 30 − 6p2 + 3p1 + 2c = 0.

(11)

−9−

p2∗ =

352 ≈ 9.026. 39

(13)

The whole supply chain would realize a profit of Π( p1∗ , p2∗ ) ≈ 202.56. Can a single uniform wholesale price elicit p1∗ and p2∗ ? Fix a w, the pair of prices, denoted ( pb1 (w) , pb2 (w)), that satisfies (7) and (8) are: 27 w+ 46 16 pb2 (w) = w+ 23

pb1 (w) =

105 , 23 150 . 23

(14) (15)

There exists no w that would make both pb1 (w) = p1∗ and pb2 (w) = p2∗ .10 Can a price floor help achieve Π( p1∗ , p2∗ )? If M wants a w to induce R2 to pick its optimal price, given that p1 is already at its optimal. M can solve for w by setting p2∗ = p2 ( p1∗ , w) from (8), for which w = 3.1282. Facing w = 3.1282 and p2∗ , would R1 pick a price lower than its optimal price? From (7), we can check that p1 ( p2∗ , w = 3.1282) = 6.4423 < p1∗ = 7.385. If R1 is free to choose its own price, R1 would under-price. To curb this under-price incentive, M can impose a price floor of 7.385, which would bind R1 but not R2. Can a price ceiling help achieve Π( p1∗ , p2∗ )? If M wants a w to induce R1 to pick its optimal price, given that p2 is already at its optimal. M can solve for w by setting p1∗ = p1 ( p2∗ , w) from (7), for which w = 5.0128. Facing w = 5.0128 and p1∗ , would R2 pick a price higher than its optimal price? From (8), we can check that p2 ( p1∗ , w = 5.0128) = 9.9679 > p2∗ = 9.026. If R2 is free to choose its own price, R2 would over-price. To curb this over-price incentive, M can impose a price ceiling of 9.026, which would bind R2 but not R1. Can two wholesale prices help achieve Π( p1∗ , p2∗ )? We can check by sub p1 = 96 13 and p2 = 352 into (7) and (8), we get w = 5.0128 and w = 3.1282. 2 1 39 To sum up, with only one single wholesale price w, the whole supply chain would not realize a profit of Π( p1∗ , p2∗ ). What contract then would suffice? Charging two wholesale prices (i.e., w1 = 5.0128 and w2 = 3.1282) would do. Charging one uniform wholesale price at w = 3.1282 while setting a vertical price floor of 7.385 would do too. In this case, R2 would charge the optimal price, while R1 will be bound by this price floor, which is optimal for the whole supply chain too. Charging one uniform wholesale price at w = 5.0128 while setting a vertical price ceiling of 9.026 would do too. In this case, R1 would charge the optimal price, while R2 will be bound by this price ceiling, which is optimal for the whole supply chain too. 10 If

only one single wholesale price w can be used, which w would M pick? M solves max( pb1 (w) − c) (30 − 4b p1 (w) + pb2 (w)) + ( pb2 (w) − c) (30 − 3b p2 (w) + 2b p1 (w)) w ≥0

s.t. c = 1. The solution is w ≈ 4.1737 and a profit of 201.10 only.

− 10 −

(16)

In this simple example, we do not see how the RPM contracts can be better than a contract with two wholesale prices. In section 4, we argue that the RPM contract is more robust when the setting slightly changes. 3.2.

The general model: assumptions and a proposition

A1. An upstream firm, denoted M, produces a single product at the constant cost c selling to consumers through two downstream firms, namely R1 and R2. R1 and R2 serve a certain group of consumers within their own areas. A2. The downstream firms purchase the product from the upstream firm at a constant wholesale price w and their sale prices p1 (for R1) and p2 (for R2) are higher than the cost c. A3. Consumers’ demands are based on the prices of R1 and R2, so the demands for R1 and R2 can be represented as D1 ( p1 , p2 ) and D2 ( p2 , p1 ), respectively, and they ∂D ( p ,p ) ∂D ( p ,p ) have the properties that 1∂p1 2 < 0 and 2∂p22 1 < 0. Moreover, if consumers 1 in R1’s area find R2’s price decreasing (increasing), their demand for this product ∂D ( p ,p ) ∂D ( p ,p ) may also decrease (increase), which means 1∂p12 2 ≥ 0 and 2∂p2 1 ≥ 0. 1

Proposition 1 Under assumptions A1, A2, and A3, (a) A single uniform wholesale price alone is insufficient to achieve the first best equilibrium unless R1 and R2 are symmetric. (b) Different wholesale prices charged to R1 and R2 are also sufficient to reach the first best. (c) A wholesale price and resale price maintenance (a vertical price floor or price ceiling) are sufficient to implement the first best equilibrium when R1 and R2 are asymmetric in a certain way. The potential misalignment of incentives. Comparing a downstream firm’s profit function with the whole supply chain’s profit allows us to see the potential misalignment of incentives in this setting. The R1’s profit function is π1 ( p1 , p2 , w) = ( p1 − w) D1 ( p1 , p2 ) + A1 ,

(17)

where A1 > 0 means subsidy and A1 < 0 means franchise fee. R2’s profit function is similarly defined. The whole supply chain’s profit function is Π( p1 , p2 ) = ( p1 − c) D1 ( p1 , p2 ) + ( p2 − c) D2 ( p2 , p1 ).

− 11 −

(18)

Their respective first derivatives with respect to p1 are: ∂D ( p , p ) ∂π1 = D1 ( p1 , p2 ) + ( p1 − w) 1 1 2 , ∂p1 ∂p1 ∂D ( p , p ) ∂D ( p , p ) ∂Π = D1 ( p1 , p2 ) + ( p1 − c) 1 1 2 + ( p2 − c) 2 2 1 . ∂p1 ∂p1 ∂p1

(19) (20)

Re-do the same exercise for R2. Combining them, we have the following equations: ∂π1 ∂Π h ∂D ( p , p ) ∂D ( p , p ) i = − ( w − c ) 1 1 2 + ( p2 − c ) 2 2 1 , ∂p1 ∂p1 ∂p1 ∂p1 h ∂π2 ∂Π ∂D ( p , p ) ∂D ( p , p ) i = − ( w − c ) 2 2 1 + ( p1 − c ) 1 1 2 . ∂p2 ∂p2 ∂p2 ∂p2

(21) (22)

The mis-alignment of interest of the downstream firms (relative to the whole supply chain) comes from the 2 terms in the square bracket of (21) and (22). The first term in the square bracket is the externalities the downstream firm’s pricing has on the upstream firm’s profit through its own demand changes. The second term is the externalities the downstream firm’s pricing has on the profit of the whole supply chain through the other downstream firm’s demand changes. If the two square bracket ∂Π i terms happens to be zero, we would have ∂π ∂pi = ∂pi for i = 1, 2, i.e., the profitmaximizing price for the downstream firm is the profit-maximizing price for the whole supply chain. Of course, nothing guarantees that the two square bracket terms are ∂Π i zero. If they are not, we have ∂π ∂pi 6 = ∂pi and there would be mis-alignment of interest of the downstream firms. ∂Π The first best (relative to the whole supply chain) ( p1∗ , p2∗ ) is achieved when both ∂p 1

and

∂Π ∂p2

are equal to zero. But downstream firm i = 1, 2 set its own price by equating

∂πi ∂pi

to zero instead. We now turn to both the symmetric and asymmetric cases. Being symmetric means ∂D1 ∂D1 ∂D1 ∂D1 ∂D2 ∂D2 ∂D2 ∂D2 ∂p1 = ∂p2 and ∂p2 = ∂p1 . Being asymmetric means either ∂p1 6 = ∂p2 or ∂p2 6 = ∂p1 or both. 3.3.

Symmetric downstream firms

When the two downstream firms are symmetric, their first best prices are the same ∂π2 ∂Π ∂Π 1 = ∂p = 0. To ensure ∂π (i.e., p1∗ = p2∗ ). At such a price, ∂p ∂p1 = ∂p2 = 0, we need to 2 1 have zero externalities. Therefore, there is a single wholesale price, denoted w∗ , that satisfy the following two equations: ∂D2 ( p2∗ , p1∗ ) i ∂D1 ( p1∗ , p2∗ ) + ( p2∗ − c) , ∂p1 ∂p1 h ∂D2 ( p2∗ , p1∗ ) ∂D1 ( p1∗ , p2∗ ) i 0 = (w∗ − c) + ( p1∗ − c) . ∂p2 ∂p2 0 =

h

(w∗ − c)

− 12 −

(23) (24)

The square brackets in both (21) and (22) become zero. There can never be an underpricing problem if the right wholesale price, w∗ , is picked. Consequently, RPM cannot improve the whole supply chain’s profit, rendering it an unnecessary vertical restraint in this symmetric case. On the other hand, if firms are asymmetric, then there is no single wholesale price that would make both the square brackets become zero. This proves part (a) of the proposition. 3.4.

Asymmetric downstream firms

When the two downstream firms are asymmetric, which is almost always the case in the real world, their first-best prices differ (i.e., p1∗ 6= p2∗ ). 3.4.1.

Different Wholesale Prices

One way to align their interests is for M to set two different wholesale prices that satisfying the following equations: ∂D1 ( p1∗ , p2∗ ) ∂D2 ( p2∗ , p1∗ ) i ∗ 0 = − c) + ( p2 − c ) , ∂p1 ∂p1 h ∂D2 ( p2∗ , p1∗ ) ∂D1 ( p1∗ , p2∗ ) i + ( p1∗ − c) 0 = (w2∗ − c) . ∂p2 ∂p2 h

(w1∗

(25) (26)

Take R1 for example. If M sets a wholesale price w1∗ , from (21) we will see that ∂π1 ∂Π ∂Π ∂Π ∗ ∗ ∂p∗ = ∂p∗ . And if R1 picks p1 and R2 picks p2 , by definition, ∂p = ∂p2 = 0. To ensure 1

1

∂π1 ∂p1

∂π2 ∂p2

1

= 0, we need only to have zero externalities. This applies to R2 in the same = fashion. The two wholesale prices are derived to do so. Through varying the A1 and A2 , the upstream firm can appropriate the rents through franchise fee according to their relative bargaining powers. This proves part (b) of the proposition. 3.4.2.

One Wholesale Price with a Resale Price Maintenance

Suppose (25) and (26) are such that w2∗ < w1∗ . Maximizing the whole supply chain’s profit requires the two Rs to charge p1∗ and p2∗ , respectively. Suppose M can only charge one wholesale price to both Rs. For the sake of argument, suppose R1 picks p1∗ and does not respond to R2’s price and the wholesale price, then the one wholesale price that maximizes the whole supply chain’s profit is w2∗ . R2 would then choose p2∗ . But given that w2∗ < w1∗ , if R1 faces w2∗ and p2∗ , it is not incentive-compatible for R1 to choose p1∗ . Instead, R1 would have an incentive to price lower than p1∗ . M can eliminate such an under-price incentive by disallowing any R to charge a price lower than p1∗ . In order for this minimum RPM to work, however, we need to have p1∗ < p2∗ . If this inequality holds, then the vertical price floor at p1∗ would bind R1 but not R2.

− 13 −

We can rearrange (25) and (26) to get ∂D2 ( p2∗ , p1∗ ) ∂D1 ( p2∗ , p1∗ ) / , ∂p1 ∂p1 ∂D1 ( p2∗ , p1∗ ) ∂D2 ( p2∗ , p1∗ ) = c − ( p1∗ − c) / . ∂p2 ∂p2

w1∗ = c − ( p2∗ − c)

(27)

w2∗

(28)

For w2∗ < w1∗ , we need c − ( p2∗ − c) Rearranging terms, we have For p1∗ < p2∗ , that means w2∗ < w1∗ is

p1∗ −c p2∗ −c

p1∗ −c p2∗ −c

∂D2 ( p2∗ ,p1∗ ) ∂D1 ( p2∗ ,p1∗ ) / ∂p1 ∂p1 ∂D1 ( p2∗ ,p∗ ) ∂D2 ( p2∗ ,p∗ ) 1 / 1 ∂p2 ∂p2

∂D2 ( p2∗ ,p1∗ ) ∂D1 ( p2∗ ,p1∗ ) / ∂p ∂p1 1


c − ( p1∗ − c)

∂D1 ( p2∗ ,p1∗ ) ∂D2 ( p2∗ ,p1∗ ) / ∂p2 . ∂p2

.

< 1. Therefore, a sufficient condition for p1∗ < p2∗ and

≥ 1, or

∂D2 ( p2∗ , p1∗ ) ∂D1 ( p2∗ , p1∗ ) ∂D1 ( p2∗ , p1∗ ) ∂D2 ( p2∗ , p1∗ ) / ≤ / . ∂p1 ∂p1 ∂p2 ∂p2

(29)

Inequality (29) is analogous to condition (6) in Winter (1993).11 Note that downstream firm R1 has an incentive to under-price when M charge w2∗ as the wholesale price. For a vertical price floor to curb this under-pricing incentive, this inequality says that the effects of its lowering price on R2’s demand, as a fraction of the effects on its own demand, has to be larger than the effects of R2’s lowering price on R1’s demand, as a fraction of the effects on its own demand.12 This proves part (c) of the proposition. ∂D2 ( p2∗ ,p1∗ ) ∂D ( p∗ ,p∗ ) , then (29) simply means the downstream firm with Suppose 2∂p22 1 = ∂p2 an incentive to under-price (R1) would exert a larger effect on the other downstream firm’s demand if it lowers its price than would the other firm.13

4.

Different wholesale prices or just one with a RPM?

In an asymmetric situation, while a contract with different wholesale prices (denoted (w1 , w2 )) can always elicit the right prices, a contract with one wholesale price and a RPM (denoted (w, RPM )) works only in a subset of situations. The model cannot say that under this subset of situations, one contract dominates another. We argue in this section that if both contracts can elicit the right prices, in at least 3 types of realistic business scenarios, (w, RPM) appears more attractive to use than (w1 , w2 ). 11 Condition

(6) in Winter (1993) compares the retailer’s and market’s elasticities of demands with respect to price and service. He finds that when these cross elasticities are higher in price than in service, RPM is sufficient to induce the efficient service level. 12 It is larger because the fractions in both sides are negative value. 13 The example in Section 3.1 satisfies (29): The RHS is +1 = − 1 , while the LHS is +2 = − 1 . −3 3 −4 2

− 14 −

Bootlegging concern While (w1 , w2 ) invites bootlegging, (w, RPM ) does not give any incentives for downstream firms to engage in bootlegging, as they are facing the same wholesale prices. Bootlegging is not modeled. If it is difficult for M to discourage bootlegging among Rs, (w, RPM ) seems to better ensure the maximum profit for the whole supply chain. Price discrimination legal concern Discriminating downstream firms in price may invite legal trouble, and charging a uniform wholesale price eliminates this risk.14 This uniformity helps to rid manufacturers of suspicion of conducting price discrimination. On the other hand, RPM is not without its own potential legal trouble. If price discrimination invites more potential legal trouble than does RPM, RPM appears a better option to choose. The choice among different strategies can thus be regarded as the result of balancing their respective benefits and potential litigation costs. Asymmetric information concern Suppose M knows that R are asymmetric and they are of two different types. What M does not know is which R is of which type. Under this information asymmetry setting, M cannot be certain which R should charge which wholesale price if M is contemplating to adopt (w1 , w2 ). Using contract (w, RPM) eliminates this concern because both downstream firms would faced the same contract. The one who has an incentive to under-price would automatically be binded by the RPM. There is no need for M to figure out which R is of which type. The RPM contract is also robust if the types of the R can change over-time but M cannot keep track of their types.15

5.

Welfare Analysis

While the RPM improves the whole supply chain’s profit, does it benefit the consumers? In court, it is necessary to measure the welfare effect of the RPM. We show in our example in section 3.1 that it is possible for consumers to be better off with a RPM. 14 The

U.S. Robinson-Patman Act of 1936 prohibits some forms of price discrimination, which are more applicable to B-to-B transactions. The J&J case does involve B-to-B transactions (from a manufacturer to a distributor, then the distributor sells to a hospital. 15 Specifically, referring to the general model again, what the upstream firm knows are that there are two optimal sale prices ( p1∗ , p2∗ ) and their corresponding wholesale prices (w1∗ , w2∗ ). But it does not know the actual type of each downstream firm. Without information asymmetry, M knows their types and adopting different wholesale prices can well resolve the underpricing problem. In contrast, under asymmetric information, M would have some chances to make a mistake on the types of R, resulting in potential profit loss. The application of RPM with a uniform wholesale price, however, would avoid such a problem. With information of w2∗ and p1∗ , which is already known to M, the optimal prices of both R can be reached for sure. With the effect of w2∗ , the retailer of type R2, will always set the optimal retail price p2∗ . The other retailer who has an incentive to under-price will be constrained at the vertical price floor p1∗ . M does not need to figure out the specific types of its downstream firms.

− 15 −

We can compute the welfare of a uniform per-unit wholesale price with and without 352 ∗ RPM. With a minimum RPM, p1∗ = 96 13 and p2 = 39 . Without a minimum RPM, footnote 10 calculates that the uniform per-unit wholesale price that maximizes the 8002 whole supply chain’s profit is around 4.1737. This results in p1 = 7941 1132 and p2 = 849 . Feeding these prices into the demand functions D1 = 30 − 4p1 + p2 ,

(30)

D2 = 30 − 3p2 + 2p1 ,

(31)

it is then mechanical to calculate the profits and the consumer surplus. Imposing a minimum RPM reduces the consumer surplus in market 1 from around 16.146 to around 11.251, but it increases the consumer surplus in market 2 from 41.367 to 52.170. Therefore, the RPM actually increases the total consumer surplus of the two markets from 57.513 to 63.42.16

6.

Concluding remarks

The European Community (EC) competition law does not treat vertical restrictions such as price fixing as certainly illegal, although they hold the view that they “restrict competition per se and there is always (though very insignificant) an opportunity to find them legal under the Article 81(3) of the EC Treaty” (Svirinas, 2009)17 . For this purpose, European Commission published Regulation 330/2010 to guide vertical restriction and established a presumption of legality for vertical agreements that satisfy certain criteria. In this regulation guidance, the only way to justify RPM is the use of non-price competition and free riding issues (i.e, services), which is the same with U.S. law.18 The EU, therefore, has the same issue faced by the U.S. courts, namely, if the court cannot ascertain the significance of services, there seems to be no way to explain RPM. 16 The

tedious calculations are available upon request. 85(3) of European Economic Community (EEC) Treaty provides a defense for agreements that create efficiencies which contribute to improving the production or distribution of goods or to promoting technical or economic progress, while allowing consumers a fair share of the resulting benefit. 18 Commission Regulation (EU) No. 330/2010 of 20 April 2010 on the Application of Article 101(3) of the Treaty on the Functioning of the European Union to Categories of Vertical Agreements and Concerted Practices [2010] OJ L102/1-7. Regulation 330/2010 paras. 106. “It is important to recognize that vertical restraints may have positive effects by, in particular, promoting non- price competition and improved quality of services... ” Regulation 330/2010 paras. 225. ”...RPM may be helpful during the introductory period of expanding demand to induce distributors to better take into account the manufacturer’s interest to promote the product...In some situations, the extra margin provided by RPM may allow retailers to provide (additional) pre-sales services...The parties will have to convincingly demonstrate that the RPM agreement can be expected to not only provide the means but also the incentive to overcome possible free riding between retailers on these services and that the pre-sales services overall benefit consumers...” 17 Article

− 16 −

Take Pronuptia de Paris v. Schillgalis case as an example.19 Pronuptia de Paris GmbH was a French leader in wedding dresses and accessories and operated through a franchising system. Pronuptia de Paris Irmgard Schillgallis was a franchisee, who sold in Hamburg, Oldenburg, and Hannover under the name Pronuptia de Paris. The French franchiser Pronuptia sued this German franchisee Schillgallis for not paying the royalty fees, DM 158,502 (approximately $93,815) in royalties and interest on her turnover from 1978 to 1980. However, German intermediate appellate court “held that the franchise agreement violated Article 85(1) and was null and void.” (Goebel, 1986)20 Among the pleadings of Mrs. Schillgllis, the RPM imposed by Pronuptia was a strong evidence of violating the Article 85(1)21 . The EU’s approach regarding RPM is that RPM is regarded as restricting price competition. Based on the guidelines of Regulation 330/2010, the products in this case are not able to satisfy the instructions mentioned in the guidelines, since the franchise system always gives consumers an impression that they provide uniform services. Indeed, the franchiser usually sets standards and provides assistance for its franchisees on training staff, promotion and so on to build up its reputation and image. Therefore, no matter how prices vary, the services were supposed to be similar among different franchisees and no free-riding problem existed in a franchising system. No services argument can be used to justify RPM. This B-to-C case belongs to the cases involving RPM but without services. Can our model apply to this case? Suppose Pronupita, the franchiser, has imposed a minimum RPM for all franchisees. This price floor may represent the value and image of the product. Its franchisee in Hamburg, Germany, reduces the price in its district to attract consumers who are less likely to pay a high price for their products. The consumers in Hamburg certainly enjoy more welfare, but this reduction has horizontal externalities on the other franchisees in other areas - consumers of those areas would question the true value of the product and would be reluctant to purchase products. The inconsistent recognition of the brand will result in the loss of customer adhesiveness, which obviously undermines the image as well as profitability of Pronupita. As Hamburg is just one of the tiny markets for Pronupita, the increased sale of the product may not offset the consequences of harming image leading to the loss of consumers in other areas. Our paper offers a feasible explanation for the use of RPM without the services. 19 Case 161/84, Pronuptia de Paris GmbH v.

Pronuptia de Paris Irmgard Schillgallis, 28.1.1986, (1986) ECR

353. 20 Article

85(1) of European Economic Community (EEC) Treaty prohibits agreements that restrict competition, such as directly or indirectly fix purchase or selling prices or any other trading conditions and limit or control production, markets, technical development, or investment. 21 Actually, the Supreme Court ruled that the franchiser did not impose actual RPM since the franchiser only made price recommendations to the franchisee, which did not constitute a restriction of competition. But it confirms the fact again that the use of RPM is illegal.

− 17 −

Although Leegin has made the current U.S. law in adjudicating RPM to move a step forward: from the rule of per se illegal to rule of reason standard, only relying on services argument confronts the difficulties of estimating and agreeing on the significance of services.

References [1] Asker, John, and Heski Bar-Isaac. 2014. “Raising Retailers’ Profits: On Vertical Practices and the Exclusion of Rivals.” American Economic Review, 104(2): 672-86. [2] Burguet, Roberto, and Yeon-Koo Che. 2004. “Competitive Procurement with Corruption.” RAND Journal of Economics, 35(1): 50-68. [3] Comanor, William S. 2013. “Leegin and its Progeny: Implications for Internet Commerce.” Antitrust Bulletin, 58(1): 107-127. [4] Deneckere, Raymond, Howard P. Marvel and James Peck. 1996. “Demand Uncertainty, Inventories, and Resale Price Maintenance.” Quarterly Journal of Economics, 111(3): 885-913. [5] Easterbook, Frank H. 1984. “Vertical Arrangements and the Rule of Reason,” Antitrust Law Journal, 53(1): 135-173. [6] Jullien, Bruno, and Patrick Rey. 2007. “Resale Price Maintenance and Collusion.” RAND Journal of Economics, 38(4): 983-1001. [7] Klein, Benjamin. 2009. “Competitive Resale Price Maintenance in the Absence of Free Riding.” Antitrust Law Journal, 76(2): 431-481. [8] Klein, Benjamin, and Kevin M. Murphy. 1988. “Vertical Restraints as Contract Enforcement Mechanisms.” Journal of Law and Economics, 31(2): 265-297. COMPETITIVE RESALE PRICE MAINTENANCE IN THE ABSENCE OF FREE RIDING Benjamin Klein Antitrust Law Journal Vol. 76, No. 2 (2009), pp. 431-481 [9] Krishnan, Harish, and Ralph A. Winter. 2007. “Vertical Control of Price and Inventory.” American Economic Review, 97(5): 1840-1857. [10] Massell, Benton F. 1969. “Price Stabilization and Welfare.” Quarterly Journal of Economics, 83(2): 284-298. [11] Mathewson, Frank, and Ralph A. Winter. 1984. “An Economic Theory of Vertical Restraints.” RAND Journal of Economics,15(1): 27-38.

− 18 −

[12] Mathewson, Frank, and Ralph A. Winter. 1998. “The Law and Economics of Resale Price Maintenance.” Review of Industrial Organization, 13(1): 57-84. [13] Marvel, Howard P., and Stephen McCafferty. 1984. “Resale Price Maintenance and Quality Certification.” RAND Journal of Economics, 15(3): 346-359. [14] Orbach, Barak Y. 2008. “Antitrust Vertical Myopia: The Allure of High Prices.” Arizona Law Review, 50(1): 261-287. [15] Orbach, Barak Y. 2010. “The Image Theory: RPM and the Allure of High Prices.” The Antitrust Bulletin, 55(2): 277-307. [16] Telser, Lester G. 1960. “Why Should Manufacturers Want Fair Trade?” Journal of law and economics, 3: 86-105. [17] Winter, Ralph A. 1993. “Vertical Control and Price Versus Nonprice Competition.” Quarterly Journal of Economics, 108(1): 61-76. [18] Xia, Lan, Kent B. Monroe, and Jennifer L. Cox. 2004. “The Price is Unfair! A Conceptual Framework of Price Fairness Perceptions.” Journal of marketing, 68(4): 1-15.

Appendix A.

Literature Review

The major explanations for the use of RPM in the literature can be either anticompetitive and pro-competitive. A.1.

Anti-competitive Explanation

First, RPM is viewed as a method to facilitate a retailers’ or manufacturers’ cartel, which is known as conspiracy theory. In fact, this view is the main point of Telser (1960), but not the points that others quoting the paper is referring to. RPM could be used to reduce downstream price variation and detect manufacturers’ cartel deviation (Jullien and Rey, 1996). Especially in a competitive retail market, agreed uniform retail prices could help manufacturers fix their wholesale prices appropriately with stationary retail cost conditions. By removing the variation of retail prices, RPM can increase cartel stability of manufacturers since the wholesale prices are hard to be observed by each cartel member (Mathewson and Winter, 1998). RPM could also act as a bargaining chip through which retailers can impose pressure on the manufacturer if they maintain a cartel. This explanation of forming retailer cartel is most often

− 19 −

referred to retail drug stores in the US and grocery outlets in Europe (Mathewson and Winter, 1998). To block the entry of large discount outlets, those traditional, lowvolume outlets with plentiful assets force manufacturer to use RPM to restructure the cartel retail price. Asker and Bar-Isaac (2014) provide evidence that in an equilibrium situation, to protect or raising retailer’s own profit, incumbent retailers are more likely to use RPM to exclude a lower cost rival from a market. A.2.

Pro-competitive Explanations

Second, RPM as a way to increase margin can encourage retailers to carry larger inventories so that it could reduce the stock-out possibility when facing uncertain demands and thus benefit consumers as a whole (Jullien and Rey, 1996). Krishnan and Winter (2014) also show the distortion in the decision of price and inventory and they prove that only a two-part price and chosen inventory by each outlet, could never coordinate incentives and correct the distortion, but resale price floors or buyback policies (retailer-held options to sell inventory back to the manufacturers) could achieve the first best decision. The third explanation attempts to interpret it in terms of enhancing retailer services by solving free-riding problem. This is a footnote in Telser (1960), perhaps one of the most influential footnote in economics. Free-riding problem refers to that without the use of RPM, consumers would probably simply pass by a retailer with special service, such as knowledgeable personnel, post-sale service. This phenomenon is particularly common for those new products, since customers tend to collect as much information as possible before purchasing the product in a discounting retailer which they consider as the cheapest store. This would apparently cause a loss to the retailer providing add-value services whereas charging a higher price (Telser, 1960). The deployment of RPM tends to reduce the price competition among retailers and induce them to provide an identical, optimal level of services. This proves to be especially crucial since the extensive use of the Internet has been making the purchase more convenient and cheaper. Winter (1993) also explores the optimal combination of price and service instruments which maximizes the overall profit, by focusing on the tastes of consumers on the “product margin.” The profit functions of retailers and the whole system reveal the vertical and horizontal externalities, which are the focus of the distortion in their pricing decisions. Retailers are therefore prone to engage in price competition and provide insufficient services in considering consumers on the wrong margin - on the “shoppers” but not the new customers that could bring the true increased sales to the manufacturer. In this situation, vertical control would align the incentives and interests of retailers and the whole system.

− 20 −

Our paper is based on Winter (1993)’s framework, which makes the comparison of the profit function of retailers and the whole system, but without triggering the need of service. From the first order conditions of these profit functions, we find the distortion. As an independent retailer that can set its own price, it usually only considers its own group of customers and profit, seldom taking the other retailers selling the same product into consideration. This may lead to the profit loss for the whole system. It is easy to observe this outcome in a situation that customer stealing is existed and prevalent among nearby retailers. One retailer lowering its price will attract the customers supposed to belong to other retailers or inter-margin area (Winter, 1993). But not only that, it is also the case even without customer stealing. For instance, a Spanish retailer and a Chinese retailer are selling the same product made by a manufacturer. Suppose it is optimal for the Chinese one to set a higher retail price than the Spanish one. The customers know that the product can be bought both in China and Spain. But the nature of the product does not justify them bootlegging across the two countries. Some Chinese customers would feel alright to buy the product in China at a high retail price, provided that they would not feel “cheated,” or just feel that they enjoy a high quality or a high-end brand of the product. But it may not work in the case that they find the same product sold in Span is at a very low level. A Google search is all it takes to find that out, and that would probably piss the customers off. Feeling unfair, uncertainty of the quality, brand image destroying, waiting price to decline and so on all could be one of the possible reasons lead to this consequence. Various marketing researches have shown that negative emotion/ or lower perceived value would occur if customers think they are facing a less fair price (Xia, Monroe and Cox, 2004). Marvel and McCafferty (1984) provide that a uniform price with the use of RPM is the quality certification that the manufacturer wishes to give to the consumers. Orbach (2008, 2010) puts forward an interesting image theory that the disparity of prices would lead consumers to suspect of brand image and the sales of product would be influenced permanently once the image is damaged. Massell (1969) studies the different consequences of welfare of producers and consumers resulting from a stable price and fluctuating price. The result shows that the price stabilization will enhance the welfare under certain conditions. All the results of these studies help to support that one retailer’s price strategy would influence others’ no matter existing customer stealing or not. If the Spanish retailer does not take the Chinese retailer into consideration, which is always the case, the Spanish retailer is probably setting too low a price that would discourage Chinese customers from buying at the Chinese retailer, something the manufacturer does not want to see. This example illustrates two points. First, the quantity demanded faced by one retailer can depend on the price set by the other retailer, even though there is no

− 21 −

customer stealing among the retailers. Second, underpricing incentives can indeed happen even if there is no customer stealing. Our solution to this problem is that we could simply use a single wholesale price with RPM to achieve optimal sale prices for the collective profit.

B.

Hospital Procurement in China

The procurement system of public hospitals in China contains two parts.22 The first one is the procurement of medical equipment and the other is the procurement of drugs and clinical supplies. The medical equipment covers from the common equipment, such as X-ray and B ultrasound equipment for normal check, to more specialized and technical apparatus, like the laboratory and occupational therapy equipment. The equipment needs professional knowledge and services to operate and usually, the hospital cares more about the function, quality and brand of the equipment than the price. In this category of procurement, the hospital and the related department of the local government will form a group and together make the purchase decision through the open tendering.23 Then this group will first do the investigation about properties, price and after-sale services of different brands of equipment through other hospitals who are in use of this equipment.24 Next, the hospital will hold an open tendering for all medical companies and normally, there are five times inquiry about the negotiation and bargain on the price. After collecting all the advices and comments from different parties of the group, the hospital will make the final decision on which brand to choose with the supervisory of the local Public Health Bureau. Drugs and clinical supplies include generic and branded drugs, examination gloves, and sutures and so on. Before 2015, the procurement of drug and clinical supplies should first follow the catalogue of the National (or Provincial) Public Health Bureau, then the hospital can purchase the drug by their own procurement system.25 Figure ?? shows the procurement system of the public hospital. When the drug or clinical supplies is in the catalogue of the Health Bureau, the hospital usually purchases it 22 We here discuss the system of public hospitals rather than private hospitals, because private hospitals use their own funding to purchase supplies and drugs but public hospitals get the fund from the government, which may induce more problems, such as bribery and monitoring of the government, and require a more complex system. 23 Associated parties: (1) related departments of the local government include Finance Bureau (Department of Purchasing Management) and the Public Health Bureau (Disciplinarum Committee and deputy director of the bureau who is in charge of the procurement system); (2) management of the hospital; (3) related technicians and directors who will use this equipment. 24 The choice of other hospitals for this home hospital can be the hospital which locates in the city near the home hospital’s city. It can be within or outside the province of the home hospital. 25 The directory of the National (or Provincial) Public Health Bureau determines which drugs and clinical supplies that hospitals can purchase.

− 22 −

through an open bidding. After 2015, Chinese government decides to take over the procurement system of the public hospital. It delegates the right to every provincial Public Health Bureau to implement the uniform bidding and procurement.26 Every provincial Public Health Bureau will also carry out an independent opening tendering within its own province. The procedures are similar with those in figure x except the group of implementing the bidding contain the officers (usually the Chair) of related departments of the government and doctors and professions of some provincial hospitals. After the bidding, the provincial Public Health Bureau will publish the price, brand, and properties of every kind of drugs and supplies in the catalogue. Each public hospital in the province should purchase drugs and supplies strictly following the provincial uniform procurement system. The J&J case took place before 2015, thus the procurement of stapler and suture was decided by the hospital alone and the procedures follow the figure 2. It seems that the whole system is transparency and fair for the hospital to get the cheapest and best products from the medical company. However, after we have interviewed several related people, including ones in the hospital and ones in the medical company, we found that there existed a gray zone in this system, where might be full of bribery, corruption, fraud and unlawful activity. The bidding can be insidious. One possibility is that the chosen medical company could first collude with the hospital and forge two other virtual medical companies to attend the bidding. Moreover, it is the person in charge of the procurement system that make the final decision on which company to purchase from or whether they need to switch to a particular brand from the previous one. This person in charge will collect all the information from different parties, but the purchase decision may be inclined to the medical company that offers the heftiest rebate. In this sense, Chinese government in 2015 decides to improve the procurement system of the public hospital by carrying on a provincial uniform procurement and administration system. We can speculate that without this uniform system, the brand that promises a uniform or minimum sale price could also be a good way to suppress this kind of corruption activities. Even for the provincial procurement system, it can also help the government to figure out whether the related procurement officers work in a “right” way and the procurement officers can also clean up the suspicion of potential corruption. We also study the publicly funded health-care system of the England—the National Health Service (NHS).27 It has the greatest purchasing power comparing to other UK organization. However, according to their official document of a procurement 26 Different

China’s provinces do exhibit slight variations in their hospital procurement practice. is the largest and the oldest single-payer health-care system in the world. Primarily funded through the general taxation system and overseen by the Department of Health, the system provides health-care to every legal resident in the United Kingdom, with most services free at the point of use.” From www.nhs.uk. 27 “It

− 23 −

Figure 2: China’s hospital procurement procedures in a flow-chart The Management

Pharmacy Department

The Administration and

Committee of Drugs

Clinical Supplies Department

Leading Group of

and Clinical Supplies

Drugs and Clinical Supplies

Procurement

Procurement

The Committee decides the type, dosage form and

Bidding (drugs and

specification of

supplies in the Others

drugs.

catalogue)

1. Study the type of the

The specific

drug to be procured

approved drugs

Confirm the pharmacy companies attending the bidding;

2.

Twice or more times negotiation and

Go through the procedures

bargaining;

of specific drugs

3.

Lowest price wins the bidding

The approval of procurement

The result of the bidding is Pharmacy keeper fills the form of procurement of drugs and Direct

clinical supplies

delivered to pharmacy and the Committee of Drug Delivery of the hospital

procurement

The Chair of the hospital approves

Procurement officer purchases drugs and clinical supplies according to the plan

Delivery and check

development program, they admit the variation across the NHS, which needs more transparency to improve the productivity and save the cost.28 They are also handling with bribery, corruption, fraud and unlawful activities by founding a more transparent data system and carrying on anti-fraud plans. We can see that transparency and appropriate monitoring are required by every country to help the government and the hospital to save the cost and increase the efficiency. It is thus quite understandable that the implement of RPM is an ideal choice for the medical company to serve both its goal of the acquisition of different markets and the concern of the hospital on the 28 The

variation includes different prices for the same product and different solutions to the same

problem.

− 24 −

procurement of clinical resources.

− 25 −