Effects of Seller Reputation on Success at Online Auctions: An Empirical Study

Article excerpt

In online auctions, success for sellers typically means making sales and getting the price they want at the auction. To test the effectiveness of such auctions, as defined by the degree to which sellers achieve these twin results for themselves in auctions, we examined 8250 on-line auctions conducted by the giant mobile phone company Nokia from September 21 to November 18 in 2003 in Yahoo!. Two step estimations were used to simultaneously examine the probability of auction success and transaction prices. Empirical findings support that the view that the seller's reputation positively affects the probability of auction success, whereas it does not affect the transaction price.

(ProQuest: ... denotes formulae omitted.)

Introduction

Online auctions recently have become the most popular activities of electronic commerce. There are more than billion dollars traded by online auctions each month since 1999, and the trading quantum is growing more than 10% every month (Lucking-Reiley, 2000). That induces many studies to investigate in online auctions, which have focused on the type of online auction, the behavior interaction between sellers and buyers, and the design for auction mechanisms. (Bajari and Hortacsu, 2004) This study focuses on the effect of the reputation mechanism, which is the most important designation in online auctions, and we believe reputation plays an important role in online auction.

Information asymmetry may be the most important problem existed between the sellers and the buyers in online auctions. The problem may be derived from inability to directly inspect the auction item, and market failure might happen due to the imperfectly signaling. Auction websites such as Yahoo and eBay try to alleviate asymmetric information by offering reputation mechanism. How does the reputation mechanism work? Take Yahoo for example: when the auction ends, Yahoo auction system automatically sends email to notify the seller and the winning bidder. Consequently, the seller and the buyer can give a score to each other for the transaction. The score can be rated by a positive (+1), neutral (0), or negative (-1) score, and each score will be along with text comment. After every successful transaction, the scores will be cumulated, which is represented as their own reputation.

Previous researches have studied the effect of reputation for online auction price. They almost conclude that reputation could indeed affect the auction price positively (LuckingReiley et al., 2000; Houser and Wooders, 2006; Resnick and Zeckhauser, 2001 ; Melnik and Alm, 2002, 2003; McDonal and Slawson, 2002). Among these, Lucking-Reiley et al. (2000) find that if a seller gets every additional 1% positive score, the auction price would be raised by 0.03%. But if a seller gets additional 1% negative score, the auction price would be decreased by 0. 1 1%. Otherwise, Houser and Wooders (2006) suggest that if the seller gains every additional 1 0% positive score, it will lead to the rise of auction price by 0. 17%. Consequently, the additional 10% negative or neutral score would make the auction price descended by 0.24%.

Unlike the aforementioned literatures, Eaton (2002) concludes that negative score positively affects the auction price. He discussed that most sellers or buyers on the online auctions may have more or less negative scores. Once the seller has a negative score, it tends to have more communication records between the sellers and bidders before bidders bidding in the auction. Consequently, the negative scores can have positive effect on auction price.

Based on the above considerations, the role of reputation in determining the auction price seems to be still debatable. The goal of this paper is to re-evaluate the relationship between the auction price and the reputation of the sellers by two-step estimation method. The seller's reputation seems to indeed affect the probability of auction success, but insignificantly affect the transaction prices in online auctions. …