Abstract: How bidders’ jump bidding behavior is influenced by auctioneers’ strategy in online auction has not yet been effectively verified by previous studies. This paper analyzes empirically the impact of auctioneers’ strategy on the type of bidders’ jump bidding behavior based on the auction data collected from 7788.com (an online collections auction platform). The research consists of two parts. First, it distinguishes the type of bidders’ jump bidding behavior by cluster analysis, and finds three different types: tentative one, competitive one and sniper one. Second, it builds a Mlogit model to analyze the impact of auctioneers’ ex-ante strategies such as auction rule setting, reputation management and auction item information displaying on the type of bidders’ jump bidding behavior. Results show that, compared with bidders’ tentative type of jump bidding behavior, variables related to auction rule can significantly stimulate bidders’ sniper type of jump bidding behavior; variables about auctioneers’ reputation can significantly stimulate bidders’ competitive type of jump bidding behavior; variables of auction item information, however, have no significant influence on bidders’ either competitive type or sniper type of jump bidding behavior.

Key words: jump bidding, auction rule, auctioneers’ reputation, auction item information