银行支付卡双面市场的经济学：定价、应用和使用【外文翻译】.doc

1、 外文翻译 The Economics of Two-Sided Payment Card Markets: Pricing, Adoption and Usage Material Source: JEL Classification: L10, D40, O30 Author: James Mc Andrews and Zhu Wang This paper provides a new theory for two-sided payment card markets by positing better micro foundations. Adopting payment cards

2、 by consumers and merchants requires a fixed cost, but yields lower marginal costs of making payments. Considering this together with the heterogeneity of consumer income and merchant size, our theory derives card adoption and usage pattern consistent with cross-section and time-series evidence. Our

3、 analyses also help explain the observed card pricing pattern, particularly the rising merchant (interchange) fees over time. This is because a private card network, besides internalizing the two-sided market externality, has the incentive to inflate the card transaction value. We show that privatel

4、y determined card pricing, adoption and usage tend to deviate from the social optimum, and imposing a ceiling on interchange fees may improve consumer welfare. Keywords: Payment Cards, Two-sided Market, Interchange Fees. 1. Introduction As credit and debit cards become increasingly prominent forms o

5、f payments, the structure and performance of payment card markets are drawing increasing scrutiny. Many controversial issues are raised about interchange fees the fees paid to card issuers when merchants accept their credit or debit cards for purchase. Interchange fees are typically set by card netw

6、orks and in many instances they are considered by competition authorities to be too high.2 Particularly, interchange fees in the US are among the highest in the world, and they have been increasing in recent years despite falling costs in the card industry. Following the pioneering work of Baxter (1

7、983), an important antitrust literature has discussed the potential anticompetitive effects of the collective determination of interchange fees within payment card systems. However, no systematic theoretical analysis was available until very recently when several formal models of the payment card in

8、dustry were developed (e.g., Schmalensee 2002, Rochet and Tirole 2003, Wright 2003, 2004). These models aim to provide a more rigorous analysis of pricing and volumes in card payment systems. The framework that they use highlights the existence of common patterns between this industry and other netw

9、ork industries including the Internet, media, video games and software, which have been termed as “two-sided” markets. The two-sided market theories emphasize the fundamental externality in card payment systems. Every card transaction necessarily involves two users: a cardholder and a merchant. Card

10、holders benefit from their holding a card only if their cards are accepted by a wide range of merchants, and merchants benefit from the card only if a sufficient number of consumers use it. Therefore, it is reasonable for the card network to price differently to cardholders and merchants in order to

11、 effectively balance the demand on the two sides of the market. Meanwhile, an important question is whether interchange fees or card fees in general could beset at the “wrong level.” These theories show that, although the socially optimal and privately optimal levels of card fees both depend on the

12、same factors (e.g., issuing costs, acquiring costs, cardholders and merchants demand elastic ties, market structure, and bargaining power of the parties), they are not equal in general. However, given various complications of the models, including imperfect competition of merchants, there is general

13、ly no way to tell that the card fees are systematically too high or too low, as compared with socially optimal levels. While these two-sided market theories advanced our understanding of the card payment systems, there are weaknesses in their micro foundations. First, these theories typically assume

14、 payment cards charge fixed-dollar fees. However, in reality, it is those cards charging proportional fees that have pricing controversies. Second, they assume a distribution of “convenience benefits” from the use of a payment card for both sides of the market, but do not really explain where those

15、benefits come from. As a result, those benefits are often referred to in non-pecuniary terms, and consumers have a fixed demand for goods invariant of their payment choices. While that assumption may fit other two-sided markets, it does not describe payment demand well, as consumers derive their dem

16、and for payment services from their underlying demand to purchase a good or service. In other words, consumers purchasing power depends on their payment choices, and the monetary benefit has always been the most important consideration for consumers to choose among different payment options. A simil

17、ar argument holds for merchant benefits. Third, these theories typically assume merchants engage in a special form of imperfect competition (e.g., Hotelling). While this is handy for considering merchants business stealing motives for accepting cards, it complicates the overall picture and makes the

18、 social welfare analysis difficult. In fact, when one combines the unspecified merchant/consumer benefits from the use of a payment device with the strategic effects of imperfect competition and differences in bargaining power, the welfare analyses in the existing theories are quite formidable. 2. A

19、 New Approach In this paper, we provide a new analysis of two-sided payment card markets by positing better Micro foundations. First, to be consistent with the reality, we assume payment cards charge proportional fees. Second, the payment methods that consumers use do not yield utility directly, but

20、 instead imposes a frictional cost on their purchases. Third, in contrast to the existing model, we assume a contestable market for merchants, which greatly simplify the welfare analysis. Using the model, we are able to systematically compare the differences between privately and socially optimal ca

21、rd fees: Because the card network makes profit from serving card users, it does not care about cash users in its pricing decision. Meanwhile, lowering card fees to consumers help inflate the value of card transactions so the network prefers high merchant fees. In contrast, the social planner cares a

22、bout both card users and cash users, and prefers lower merchant fees because this helps to lower retail prices and raise consumers real purchases. Therefore, these are the fundamental differences between the private network and the social planner in dealing with the “two-sided market effects” and th

23、e “inflation effect.” Meanwhile, we found that under the private network, cards are not over adopted and used as suggested by many existing models. Rather, because the card network charges a high markup, it leads to a lower card adoption and usage compared with the social optimum. Also, we found tha

24、t cash users are disadvantaged under the private network. However, this is not because they have to subsidize card users as suggested by many existing theories (In fact, cash users are subsidized by card uses because they use less efficient payments), but rather because under the private network, fe

25、wer stores serve both card and cash users, so fewer cash users get subsidized by card users who use more efficient payments. 3. The Model Our model studies pricing, adoption and usage of monetary payment devices. We first lay out the environment in which only one payment device, referred to as cash,

26、 is in use. Then we will consider the introduction of an alternative device, which we refer to as a payment card. We model the consumers as having generalized Cobb-Douglas preferences across a range of goods. They take prices as given. Each consumer is endowed with income, which is distributed acros

27、s the population of consumers according to known cumulative distribution function. The merchant side of our model is quite stylized. Each merchant competes in a contestable market for a single good that the merchant sells, and prices are set at the zero profit level. Consumers and merchants are both

28、 presented with the option to adopt a new payment device that offers a lower variable cost of use, but a higher fixed cost relative to the pre-existing alternative. They each make their optimal adoption decision taking the others choice as given. The model yields a two-sided market under our structu

29、re of the heterogeneity of consumer income and merchant size, the fixed adoption costs, and, finally, under the assumption of price coherence by merchants that accept both payment devices. 3. Conclusions This paper provides a new two-sided market theory to study the pricing, adoption and usage of pa

30、yment devices. Using this framework, we are able to study the evolution of payment card markets both in the short run and in the long-run, and also shed new light on related competition policy issues. Our analyses show that as payment card markets evolve the interchange fees increase over time. This

31、 is because a monopoly card network, besides internalizing the two-sided market externality, has the incentive to inflate the value of card transactions. Lowering card fees to consumers but raising them to merchants help inflate the card transaction value, so the card network prefers high interchang

32、e fees. As card service costs decline over time, the card network is able to further raise interchange fees to extract more profits out of the system. In contrast, our model shows that the social planner would prefer lower interchange fees than the monopoly network, and imposing a ceiling on interch

33、ange fees may improve consumer welfare. These findings provide some support for the ongoing investigations that competition authorities around the world are taking on the payment card interchange fees. There are many avenues for further research. Wang (2007) examines four-party systems in mature car

34、d markets without adoption externalities. Further extensions to this environment would be useful. Models of the vertical restraints imposed by card networks would also be of interest in this model and its extensions. Finally, modeling consumer credit constraints and the provision of credit via the p

35、ayment device is another important direction for future research. 银行支付卡双面市场的经济学：定价、应用和使用 资料来源： JEL Classification: L10, D40, O30 作者： James Mc Andrews and Zhu Wang 本文提供了关于微观经济学基础和银行支付卡双面市场不相抵触的一种更好的新理论。在消费者和商家使用信用卡时，银行在制作信用卡时需要支付一定的固定成本， 得到较低的边际效益。考虑到消费者的收入差距和商人的规模大小，我们对信用卡的理论要采纳和使用截面和时序模式的相关证据来推导。我们

36、的分析也有助于解释还没有投入使用的处于观测阶 段的信用卡的定价模式（特别是新兴的商业交流费用）。这是因为对于个人业务的银行卡系统来说，除去互利的表现外，还需要增加银行卡的交易价值。 我们的分析表明：私下确定价格来使用信用卡会偏离社会最优选择，对交易费实行上限制度也有可能增加消费者的利益。 关键词：信用卡、双面市场、交易费 信用卡和借记卡在如今的社会上作为越来越重要的付款形式，在付款方式市场引起了越来越多的关注和审视。 许多有争议的问题在于银行卡的交易费（交易费的费用是指当商人接受他们的信用卡或借记卡进行采购时支付给银行卡发行者的费用）。银行卡的 交换费通常是 由银行卡的网络环境或者在许多情况下

37、 ,他们被认为是由更多的竞争部门决定的。交易费，尤其在美国，是银行卡收费业务中最高比重之一。近年来，尽管银行卡产业的成本下降，但是他们还是一直在增加银行卡交易的交易费。 交易费的提出首先是从 Baxter (1983)展开的 ,他在文章中提出 反对垄断主义的观念，并探讨了在付款卡系统中潜在的反竞争性在交易费的确定中的影响。然而，之后就一直没有系统的理论分析 Baxter的观点， 直到最近关于支付卡中的正规模型 (e.g., Schmalensee 2002, Rochet and Tirole 2003, Wright 2003, 2004)的产生。这些模型提供了一种比较最严谨的方法来分析支付

38、卡的定价和价值。他们所使用的框架强调了在这个工业和其它包括互联网、媒体、视频游戏和软件等网络产业已经被称为“双面”市场的一般规律。 双面市场的基本理论强调银行卡付款体系的外部性。每张银行卡的交易必然包括两个用户：持卡人和商人 。持卡者受益于其持有的卡片应当拥有广泛的商人群体，而商人受益于这些银行卡在于只要有足够数量的消费者去使用它。因此，为了有效地平衡双方市场的需求，对银行卡系统的持卡人和商人分别定价是合 理的。 与此同时，一个重要的问题是：是否交易费或银行卡手续费在一般情况下会存在“错误的状态”。这些理论表明：虽然社会最优选择和个人信用卡费用的最佳水平都依赖于相同的因素（例如：发行成本，获得

39、成本，持卡者和客商的需求、市场结构弹性，团体的议价能力），但它们一般并不相等。然而，考虑到各种复杂的模型，包括商人的不完全竞争，这些情况通常没有办法被告知这张银行卡片所需要的费用到底是太高还是太低，相比其他的支付卡而言是否是社会最优水平。 虽 然这些双边市场理论让我们对银行卡支付系统有更先进的认识，但在微观经济基础理论上也存在薄弱的环节。首先，这些理论通常认为信用卡的卡费是固定的费用。然而，在现实中，对这些支付卡定价的收费比例还是存在争议的。其次，在双方的市场上，他们假定使用一种“便利好处”的银行付款卡，但是却不解释这些好处在到底从哪里来的，到底有哪些好处。这些好处通常被称为非经济条件，并且是

40、消费者对商品需求的一个固定不变的选择。这个假设也许适合其他双边市场的情况，因为消费者的支付购买需求是他们潜在的购买货物或服务的需求。换句话说，消费者的购买力 要取决于他们的付款方式，金钱上的利益一直是消费者选择不同的付款方式最重要的考虑因素。类似的争论会给商人来带一定的好处。第三，这些理论通常认为商人从事一种特殊形式的不完全竞争 (e.g， Hotelling)。虽然这便于考虑商人接受银行支付卡的商业动机，但它以复杂化了的全局的角度出发，使社会福利分析起来更加困难。事实上，从未指明的商人或者消费者使用一种付款装置所得到的收益与不完全竞争和议价能力的影响差异的角度上，以现有的理论来分析支付卡的收

41、益福利还是相当艰巨的。 在这篇论文中，我们将使用一个与微观经济基础不相抵触 的更好的模型来分析银行支付卡的双边市场。 首先，根据现实的情况，我们认为信用卡的费用是有比例的费用。 第二，消费者使用的付款方式在购买商品时会产生摩擦成本。 第三，与现有的模式进行对照，我们为商人假设一个有争议性市场来大大简化分析福利制度。 利用该模型系统，我们可以比较个人的和社会最优支付卡之间费用的差异。因为支付卡的网络性使它从它的使用者那里获得利润，它不在乎现金用户在定价上面的决定。与此同时，消费者支付更少的银行卡费用，会助长其交易费。其网络性会使得商人需要支付更高的交易费。相比之下，社会规划对用卡用户和现金用户两

42、者都很关心，更喜欢比较低的商人费用的支出是因为这有助于降低零售价格来提高消费者对于采购的欲望。因此 ,这些是私人的网络性与社会规划在 “双面市场效应”和“通货膨胀的效果”根本区别。 与此同时，我们发现，在一个专用网络下，支付卡并没有超支采用，而且在现有的模型中反而被更多的使用。相反，因为网络上银行支付卡收费的涨价，导致在社会最优状态下银行支付卡的低采纳。同时，我们发现那些用钱的用户在专用网络下处于不利地位。然而，这并不是因为他们在支付卡现有的理论上提出更多的建议，而是因为在专用网络服务下，更少的商 店提供刷卡和现金两种支付体系，以至于现金支付的用户得到更少的刷卡补助。 我们的模型将要研究定价、

43、应用和货币支付设备的使用。我们首先定下的环境中，只有一种付款装置，被称为现金，并且是在使用着。那我们将会引进一种设备，我们所说的付款卡。 我们假设消费者通过一些商品具有广义柯布 -道格拉斯偏好。每个客户都被赋予了收入，消费者的人口分布是根据已知的累积的分布函数。商人模型的另一个方面是相当程式化。在市场竞争比赛中 , 商人出售是单一的商品 , 价格设置在零利润水平。 消费者和商家两者都提出了选择采取一种新的可变成本较低但 是使用效率较高的相对固定成本已经存在的货币支付设备。他们各自采用对方给予的最优选择决策。该模型使得在消费者的非均质性收入和厂商规模的大小、固定成本的采用的双面市场结构下，最后，

44、在价格连贯的假设中，接受商家的支付设备。 . 本文提供了一种新的理论模式研究双面市场的定价、应用和使用付款方式。使用这个框架，我们无论在长期或者短期都能够学习到演变的支付卡市场的变化，以及新的政策执行时的提供了相关依据。 我们的分析表明：随着付款卡市场的发展，会产生越来越多的交易费。这是因为一个垄断具有网络性的支付卡，除了各种外部性的双 面市场的存在，如果要取得利益，就要抬高银行卡的交易的价值。到消费者身上的信用卡费用降低，而到商人身上的费用增长，然则会增加信用卡交易价值。而支付卡的服务成本的降低，随着时间的推移，支付卡能进一步提高交换费得到额外的利润。相反，我们的模型表明：社会计划希望交易费比垄断的网络低，并对交易费实行上限制度来有可能提高消费者的福利。