This paper develops a model of information acquisition prior to an open auc
tion. The common value ("quality") of the item for sale has two distinct co
mponents. Each of two bidders must first acquire expertise and specialize i
n one component, then decide how much research to conduct, and finally deci
de how much to bid. The model provides a rationale for bidders to different
iate themselves by conducting different lines of research. Further, it show
s that research differentiation can lead to both inefficiently higher resea
rch efforts and a less efficient allocation of the item, so that the seller
's expected revenue and social welfare are lower than if bidders do not dif
ferentiate their research. (C) 2000 Elsevier Science B.V. All rights reserv
ed.