A system analysis of the spam problem

Abstract

This thesis considers the problem of the large amount of unwanted email that is being sent and received, which lowers the aggregate value of email as a communication medium from what it would otherwise be. This problem is commonly known as the "spam problem." Solutions to the spam problem involve curbing the adverse affects of existing technology as well as steering technology development in a socially beneficial direction. Unlike some other technology and policy problems, the reasons for the existence of this problem are well known and the desired effects of ideal solutions can be readily articulated. However, attempted solutions to date have not made much progress at solving the problem. I posit that this failure stems from the fact the spam problem is really a complex system, and that solutions to date have not been designed to interact with this system in a useful manner. I show that the spam problem is a complex system, and should be dealt with by developing strategies to holistically interact with it. Such strategies must embrace both technical and legal realities simultaneously in order to be successful. They must also avoid causing negative side effects that negate their purpose. First, I build a model of the system surrounding the spam problem in the form of a Causal Loop Diagram. This diagram shows the causal interactions between the various technical, legal, social, and economic forces that are present in the spam system. Using this diagram, I then identify a number of places that solutions could interact with this system. These places comprise a set of possible levers that could be pulled to alleviate the spam problem. This set of levers is then used to make sense of the attempted and suggested solutions to date.

(cont.) Various solutions are grouped by how they interact with the system. These solution categories are then presented in detail by showing, diagrammatically, how they positively and negatively affect the spam system through their interactions with it. In so doing, I attempt to argue persuasively that much of the current energy expended toward the spam problem is largely unnecessary, and in some cases, counterproductive. I additionally argue that because of the current reality of the spam problem, i.e. particular facts, we are already in a decent position to largely solve this problem by just redirecting current efforts toward more appropriate activity. Such appropriate activity is suggested, which includes steps to increase the identifiability of email in order to enable more successful litigation. Finally, an optimistic conclusion is reached that there are no fundamental reasons why the spam problem can not be dealt with in such a manner to ensure the continued usefulness of email as a communication medium.