Tolls on the Information Superhighway: Entitlement Defaults for Clickstream Data

This paper addresses the collection of "clickstream data," and sets forth a theory about the legal rules that should govern it. At the outset, I propose a typology for categorizing privacy invasions. A given state of informational privacy may be represented by: the observed behavior, the collecting agent, and the searching agent. Using this typology, I identify the specific sources of concern about collection of clickstream data. Then, based on expected levels of utility and expected transaction costs of "flipping" to a different rule, I argue for a particular set of privacy defaults for data mining.