网络价格及价格离散

——基于网络图书行业的研究

原文作者 Karen Clay, Ramayya Krishnan 单位 卡内基·梅隆大学

摘要：收集从1999年8月到2000年1月使用覆盖399本书的数据，包括纽约时报畅销书，计算机的畅销书，和随机的书籍，我们研究三十二家网上书店的定价。一个普遍的预测是，在互联网上搜索成本的减少，将导致价格分散。在样本期间，无论是价格或价格分散，我们都没有发现变化。另一个文献的预测检索，购买多次的价格或广告项目，将低于未公开的或不经常购买的物品价格分散。不同类别的图书价格似乎符合这一预测，纽约时报畅销书，具有最低的价格的一小部分的出版商的建议价格，和随机的书籍有着最高的价格。有趣的是，价格离散不符合这一预测，显然是相关的商店决定售卖特定的书的原因。我们为什么不可以观察价格收敛，其中的一个原因是因为商店已经成功地出售他们的商品。我们观察到分化（或试图分化）的一个显著的企业数量。

关键词：价格离散； 网络书店； 预测

引言

一个广泛的报道预测，低成本的信息可用性的价格，特别是比较网购引擎的上升将导致所有的互联网零售商改变批量生产的实物商品的价格，收取的价格将接近于成本。例如，在1999年9月的纽约时报的文章，Bob Tedeschi说“新的购物搜索引擎正在迅速发展，使得保证消费者能得到更低的价格，网络零售商的利润下降。”1.1999年8月的经济学家的文章说：“激烈的竞争迫使网络公司削减价格。”2.这些陈述是基于贝特朗的竞争模型，依赖利用比较购物引擎，以便于消费者充分了解书的价格。

搜索的文献预测二，价格和价格分散的广告商品，或是购买多次将低于项目未公开的或不经常购买的物品。在价格和价格的差异信息结果的差异，该预测假设个体可能会更好的了解对于某些类型的书的价格，比如新的纽约时报畅销书，因为书店经常和离线广告这些书打折。

实证的文献主要关注的是网络渠道和线下渠道的价格和价格离散的比较，结果却显示并没有什么区别。

实证的研究也检验了Stigler的预测，广告由于提高了价格信息所以会使价格或价格离散降低。通过对眼镜等商品的价格比较显示了广告与低价和低价格离散相关。最近Milyo和Waldfogel发现广告的确会降低价格，但是不是所有都是这样。他们也发现价格离散在短期内是稳定的或者增长的。

第三个预测使第一个关于低成本价格信息会降低价格和价格离散的预测陷入了危机。第三个预测公司会寻求产品差异化。零售商为了避免价格竞争，更愿意使产品差异化。因此由于这种产品差异化，价格实际上是离散的而不是集中于均衡点。

一些研究者发现了产品差异化的证据。在网上旅游行业的研究中，Clemons, Hitt, and Hann发现旅行社对于同一请求有不同的时间、价格组合。这个证据表明网上的旅行社采用的显著的产品差异化策略。

为了探究以上三个假设，我们使用从1999年8月到2000年一月的网上图书价格的数据，花了25周时间自动收集了三分之二的网上书店的399本图书。图书包括纽约时报的最畅销书，前纽约时报的最畅销书，计算机最畅销书，前计算机畅销书和任意选择的书。商店包括了建设成熟的电子零售商像亚马逊，BarnesanNoble.com，and Borders.com还有规模较小的Wordsworth和BCY。

背景

一些因素刻画了图书行业的结构。每一本的版本都有一个唯一的识别码叫做International Standard Book Number(ISBN)。ISBN重要的原因有两个。第一，这些唯一的识别码的存在使得批发商可以用电脑进行分类登记。电脑分类登记使得订单的处理更加有效，这使得整个行业都从中收益。第一家网上书店就是建立在这些分类之上的。第二，消费者可以通过搜索ISBN来确定他们是否拿到了正确的书。

四个关于市场的假设是导致预测低成本的价格信息会使得网络零售商定相同的价格并且接近成本价的基础。在勃兰特竞争模型中，1.消费者完全知道所有公司的价格，2.商品是同质的，3.公司选择价格，4.公司可以根据不变的边际成本供应任何数量。在这样的市场中，如果公司收取了比竞争者较高的价格，他将面临零需求。相反的，如果一个公司收取比竞争者较低的价格，他将会赢得整个市场。因此，价格会降至商品的边际成本，所有的公司都将定同样的价格。

关于第二个商品是同质的假设也很有道理。消费者的选择显示了他们愿意多支付百分之五的钱在亚马逊这样的商店购买同样书，即使这些商店的价格比其他的商店要高一些。因为像亚马逊，Barnes和Noble，Borders进行了大量的品牌营销还有一些特性如一键购买，书评和推荐系统。所以消费者愿意支付更多的钱这并不奇怪。这个理论文献同样预测了在完全竞争市场中公司为了是价格竞争降低而事产品差异化，而这将导致价格离散。

数据收集

自动的数据收集爬虫系统来对比价网站和一个个体的网站的数据进行收集。每本书的网页数据中提取的信息为价格，邮费，派送时间和到达天数。这些数据将在下午2点进行收集这个可以避免拥挤的网络。

为了实现文章的目的，我们集中于从1999年8月也到2000年1月的每周对这399本书的价格进行收集。由于计算的限制，我们集中于手机第一，第九和地十七个星级的最畅销书。

关于统计

总的数据统计结果呈现在表1中。我们的样本涵盖了181本任意书，136本纽约时报的最畅销书和82本计算机畅销书。

对于这五类书的横向对比，价格在平均单价和出版商推荐的价格之间显示了很大的不同。为了促进价格比价，我们使用了标准化的价格，即零售价格和推荐价格是分离的。纽约时报畅销书和前纽约时报畅销书的平均价格最低（分别为推荐价格的0.69和0.76），计算机畅销书和前计算机畅销书较高一些（分别为0.78和0.79），任意书的价格最高（0.86）。

批发商的价格基本是推荐价格的百分之五十。任意书则是例外，他们为推荐价格的备份至六十。纽约时报畅销书的平均利润最低（0.19），前纽约时报的畅销书较高（0.24），任意书（0.26），最高的是计算机畅销书和前计算机畅销书（0.29和0.29）。所以尽管任意书比计算机书有较高的价格，但是利润却更低。

价格离散的测量与价格的关系的相反的。纽约时报畅销书的标准差离散最高（28%），前纽约畅销书和当前的计算机畅销书（18%和16%），前计算机畅销书和任意书最低（10%和13%）。最低价和最高价格之前的百分之比差异也遵循这个排序。

结论

像之前人们所预测的那样的竞争并没有实现。通过对批发商和订单实现成本的计算，表明平均的净利润为9%—17%,正在销售纽约时报畅销书的亚马逊，Barnes和Noble和Borders和其他一些书店的价格等于或甚至低于成本。净利润为7-18个百分点。没有加权重的价格和标准差在1999年8月到2000年1月之间稳定的或者略微上升，表明了竞争并没有加剧。亚马逊，Barnes和Noble，和Border的销售量占据了总销售量的大多数。这些商店的价格和标准差就和稳固，标准差很小但是不为0。

我们观察到选项与价格的差异。一些有品牌的和没有品牌的商店提供了大量的选择和具有竞争力的价格。一些高价的商店的确有网站存在，但是似乎网站主要是用来做广告的，而不是作为销售的渠道。一些商店专门列出了一些特别的产品如计算机书，儿童图书，或廉价出售的书。我们的一个疑惑是商店提供了很多选择，但是没有品牌名字也没有例外的具有竞争的价格。

参考文献

[1]“The real Internet revolution.” The Economist, Aug 21, 1999, pp. 53-54.

[2]Arnold, M. A. “Costly Search, Capacity Constraints, and Bertrand Equilibrium Price Dispersion.” International Economic Review, Feb. 2000, 41 (1), pp. 117-131.

[3]Bailey, J “Intermediation and Electronic Markets: Aggregation and Pricing in Internet Commerce.” Ph.D Thesis, Department of Electrical Engineering and Computer Science, MIT, May 20, 1998.

[4]Benabou, R. “Search Market Equilibrium, Bilateral Heterogeneity, and Repeat

Purchases.” Journal of Economic Theory. 1993, 60, pp. 140-158.

[5]Benham, L. “The Effect of Advertising on the Price of Eyeglasses.” Journal of Law and Economics, October 1972, 15(2), pp. 337-52.

[6]Bertrand, J. Theorie mathematique de la richess sociale. Journal des Savants, 499-508 (1883)

[7]Book Industry Study Group. Press Release, June 2, 2000. Available at:

http://www.bisg.org/pressrelease_june2_2000.html

[8]Bookweb.org. “Industry Fair Trade Practices, 1999.”

http://www.bookweb.org/news/444.html

[9]Brynjolfsson, E and M. Smith. “Frictionless Commerce? A Comparison of Internet and Conventional Retailers” Management Science 46(4) (April 2000): 563-585. [10]Available at http://ecommerce.mit.edu/papers/friction/friction.pdf

[11]Brynjolfsson, E and M. Smith. “The Great Equalizer? Consumer Choice Behavior at Internet Shopbots.” Available at http://ecommerce.mit.edu/papers/tge/tge.pdf [12]Burdett, K. and M. G. Coles. “Steady State Price Distribution in a Noisy Search

Equilibrium.” Journal of Economic Theory. 1977, 72, pp. 1-32.

[13]Burdett, K. and K. L. Judd. “Equilibrium Price Dispersion.” Econometrica. July 1983, 51 (4), pp. 955-970.

[14]Butters, G. R. “Equilibrium Distribution of Sales and Advertising Prices.” Review of Economic Studies. Oct. 1977, 44(3), pp. 465-491.

[15]Cady, J. F. “An Estimate of the Price Effects of Restrictions on Drug Price Advertising.” Economic Inquiry 14(4) (Dec. 1976): 493-510

[16]Clay, K. , R. Krishnan, E. Wolff, and D. Fernandes. “Retail Strategies on the Web: Price and Nonprice Competition in the Online Book Industry.” Working paper, 2000.

[17]Available at: http://www.heinz.cmu.edu/~kclay/retailstrategies.pdf

[18]Devine, D. G. and Marion, B. W. “The Influence of Consumer Price Information on Retail Pricing and Consumer Behavior

剩余内容已隐藏，支付完成后下载完整资料

Prices and Price Dispersion on the Web

——Evidence from the Online Book Industry

Karen Clay

Ramayya Krishnan

Eric Wolff

NATIONAL BUREAU OF ECONOMIC RESEARCH

1050 Massachusetts Avenue

Cambridge MA 02138

May 2001

ABSTRACT

Using data collected between August 1999 and January 2000 covering 399 books, including New York Times bestsellers, computer bestsellers, and random books, we examine pricing by thirty-two online bookstores. One common prediction is that the reduction in search costs on the Internet relative to the physical channel would cause both price and price dispersion to fall. Over the sample period, we find no change in either price or price dispersion. Another prediction of the search literature is that the prices and price dispersion of advertised items or items that are purchased repeatedly will be lower than for unadvertised or infrequently purchased items. Prices across categories of books appear to conform to this prediction, with New York Times bestsellers having the lowest prices as a fraction of the publishers suggested price and random books having the highest prices. Interestingly, price dispersion does not conform with this prediction, apparently for reasons related to stores decisions to carry particular books. One reason why we may not observe convergence in prices is because stores have succeeded in differentiating themselves even though they are selling a commodity product. We observe differentiation (or attempted differentiation) by a significant number of firms.

Keyword: price dispersion; online bookstores; prediction

Introduction

One widely reported prediction is that the availability of low-cost information on price – specifically the rise of comparison-shopping engines – will lead all Internet retailers to charge the same price for mass-produced physical goods and that price will be approximately cost. For instance, in a September 1999 New York Times article, Bob Tedeschi noted “New shopping search engines are being developed quickly, bringing the promise of low prices for consumers and thinner margins for Web retailers.”1 An August 1999 Economist article stated “Fierce competition has forced web companies to slash prices.”2 These statements are based on a Bertrand model of competition and rely on consumers using comparison-shopping engines to become fully informed about book prices.

A second prediction of the search literature is that the prices and price dispersion of advertised items or items that are purchased repeatedly will be lower than for unadvertised or infrequently purchased items. The differences in prices and price dispersion are the result of differences in information. This prediction assumes that individuals might be better informed about prices for some types of books, such as New York Times bestsellers, since on and off-line bookstores frequently advertise their discounts on these books. The empirical literature has primarily focused on comparisons of price and price dispersion in the Internet and physical channels, saying relatively little about changes in prices or price dispersion over time. Lee (1997) compared pricing of used autos inInternet and physical channels and found that prices were higher in the Internet channel.

This may be due to unobserved quality variation. Bailey (1998) using matched sets of

books, compact disks, and software also finds higher prices in the Internet channel for

1996-1997. Brynjolfsson and Smith (1999) on books and compact disks for 1998-99

found that Internet retailers had lower prices, made smaller price adjustments, and had

greater or smaller price dispersion than conventional channels, depending on whether

prices were weighted by proxies for market share.

Empirical studies testing Stiglerrsquo;s (1961) prediction that advertising that improves price information would cause both price and price dispersion to fall confirm this intuition. Comparisons of the prices of goods such as eyeglasses, optometry services, and prescription drugs in states that permitted or did not permit advertising showed that advertising was associated with lower prices and price dispersion.3 More recently Milyo and Waldfogel (1999) found that increased advertising did lower prices on advertised products, but not overall. They also found that price dispersion was stable or increasing in the short run. Sorenson (2000) examined the posted price of drugs across drugstores in two communities. In line with the predictions of search theory, he finds that average price-cost margin and price dispersion are lower if the prescription is purchased repeatedly than if it is purchased only occasionally. It is notable that none of these papers examines data from the Internet.

The first prediction that low cost price information on the Internet will force price and price dispersion to fall is in tension with a third prediction. The third prediction states that firms will seek to differentiate their products. Firms prefer to be differentiated,because it mutes price competition. Thus with product differentiation, prices may not converge in equilibrium and could in fact diverge. Several authors have found evidence of product differentiation. In a study of the online travel industry, Clemons, Hitt, and Hann (1998) found agents responded to identical requests with different time/price pairs. This evidence suggests that online ticket agents engaged in significant product differentiation. Clay et al. (1999), using data from April 1999, find indirect evidence that online booksellers were engaging in product differentiation through price, selection, and other nonprice attributes. Although it is not a focus of their paper, Brynjolfsson and Smith (1999) on books and compact disks for 1998-99 also find evidence that is consistent with product differentiation.

To investigate th

剩余内容已隐藏，支付完成后下载完整资料

资料编号：[287183]，资料为PDF文档或Word文档，PDF文档可免费转换为Word