Auctions are often not independent from each other, and the movement of bidders across different auctions is one of the key linkages. We propose different measures of bidder movements (which we call bidder migration in this paper) and how such migration affects the price outcome of later auctions. Moreover, we identify two potentially confounding effects: the learning effect where bidders learn to become more sophisticated bidders, hence driving down the price of later auctions; and the desperation effect where bidders, in a hope to obtain the product that they previous couldn't win, tend to increase the prices. We empirically investigated these effects using bidding history data from eBay and Generalized Linear Model specifications. We further discussed potential applications of bidder migration for online auction platforms, such as bidder segmentation, dynamic promotions, and shill bidder detection. These bidder migration measures can be provided to internet auction sellers as a value-added service.