网上拍卖交易机制有效性研究——来自淘宝网面板数据的证据

Are Online Auction Transaction Mechanisms Effective? Evidence from Panel Data in Taobao

本文考察了淘宝网上信任保障和购物搜索两类交易机制的有效性。网上拍卖的交易机制越来越完善,除了信用度和好评率等基本信用指标外,网站先后建立了卖家服务质量、店铺动态评分、搜索排序等重要交易机制,本文利用淘宝网上的面板数据来验证上述机制的有效性。回归结果表明,信用度作为网上拍卖最早引入的一项信用指标仍发挥着一定的作用;店铺动态评分和卖家服务质量没有发挥应有的作用。进一步的解释是:网上市场存在较高的搜寻成本,买家利用搜索排序来降低搜寻成本,信用度的重要性因而被信用度排序放大;买家无法根据店铺动态评分和卖家服务质量进行筛选排序,导致这两大机制没有发挥应有作用。购物搜索在网上交易中发挥着重要作用,信任机制依赖搜索工具发挥作用。除混合数据OLS回归外,本文还利用面板数据和联立方程回归分析控制了遗漏变量偏误和内生性问题。

We investigate the roles and efficiency of trade mecha- nisms in the dominant C2C marketplaces in China, Taobao.com. Established in 2003, Taobao has now become the largest online electronic marketplace in the world. Given that the laws related to online trade in China are not mature, the rapid development of Tao- bao （including Tmall.com） is a business miracle. The key of Taobao＇s success is that Taobao fully use private orders to support online trade. Reputation, third intermediaries and other governance mechanisms substitute law and function well in online trade. Except traditional feedback measures, such as reputation score and positive feedback ratio, Taobao developed detailed shop ratings （DSR）, seller service quality （SSR）, recent feedback, payment tools, consumer guarantee services and searching ＆ ranking tools to support transaction. We collect large panel data sample from Taobao to test the role of these transaction mechanisms. We find that traditional feedback measures are still useful, while detailed shop ratings and seller service quality don＇t have significant influence on sellers＇ trade volume. Search ＆ ranking tools do affect sellers＇ sales and substitute each other. We give further explanation that high search cost exists in online market, consumers use search and ranking tools to search products, and rank- ing by reputation score magnify the effect of reputation score. How- ever, Taobao doesn＇t provide tools of ranking by detailed shop ratings or seller service quality, so these two important mechanisms cannot work. Therefore, the role of reputation mechanism depends on search ＆ ranking tools. In robust test, we use panel data regression and simultaneous equation model to address potential endogeneity. Our empirical results have important implication for Taobao--Taobao must improve its search and ranking system, add more search options for consumers, such as search ＆ ranking by DSR, SSQ, recent feed- back and add more weight to DSR and SSQ in its search algorithm.