Policy Brief Habeas Data as a Policy Toward Corporate Data Aggregation Brett Thomas Submitted 5 14 2007 Harvard University CS 199r Introduction Background of Data Aggregation Consumer data is an important aspect of business in America 1 Businesses strive to maximize the information they possess about consumers and minimize the costs of collecting and maintaining data The process of analyzing consumer data within a business is known as data aggregation and its successful practice is essential to the survival of most corporations Data aggregation has many uses A Businesses often collect and analyze data to learn more about their business These conclusions can be useful both to improve efficiency and to explore new markets For example a retailer such as Macy s that collects information about where their customers live often by asking for a zip code with a purchase may decide to open a store in a new area if they realize that a disproportionate number of their customers reside in a certain area B Businesses also use data aggregation for strategic marketing purposes Efficient businesses market their products to consumers who are more likely to become customers Discovering those consumers however requires the collection and analysis of consumer data For example a drug store such as drugstore com would only want to email advertisements about a drug for diabetes to consumers with diabetes but it must aggregate a list of people with diabetes to do so C Businesses use data aggregation as a means to price discriminate Price discrimination is a goal for businesses that is optimally efficient but often unattainable without sufficient data aggregation For example a bookseller such as Amazon com would be smart to charge its wealthier customers more but it cannot without a knowledge of its customers financial state Data aggregation is increasing in both scope and depth It is expanding in many ways A Data storage is becoming cheaper It has been shown to adhere to Moore s law2 because the capacity for corporations to store data increases by a factor of two every 12 months3 Interestingly purchasing the disk space to store data is now often less expensive than consuming the power necessary to delete it so data storage is now often an opt out policy Increased data storage means that businesses are developing the capability to possess increased amounts of data 1 This paper studies data aggregation in corporate America For simplicity it ignores data aggregation that is used by Governments 2 Cramming more components onto integrated circuits Electronics Magazine 1965 Retrieved on 5 13 2007 3 Lecture by Jim Waldo 3 20 2007 CS199r Harvard University B Data transmission is also becoming cheaper and more powerful Data transmission capacity through network enhancement and transmission innovation has been shown to increase by a factor of two every 24 months 4 This means that businesses are able to use consumer data more easily so data aggregation that was previously inefficient is now economic C New forms of media such as blogs and internet television are creating new ways for firms to use strategic marketing As a result firms are facing pressure from marketing specialists to produce more specific consumer data which requires enhanced data aggregation It should be noted that reverse pressure is also present innovations in information technology are allowing marketing specialists to produce more specified forms of advertising Marketing consultant Sally Hertz agrees right now what we are doing is beginning to catch up with what we can do 5 D Data commerce is becoming more common as specialty data aggregation companies become increasingly more powerful These companies have become embedded into the infrastructure of other businesses that rely on their services For example ChoicePoint provides background employee background checks for many companies which have been able to decrease legal obligations because they have fewer legal issues E Because of reasons A B and C data aggregation is shifting from the matching of higher level processed information to the mining of lower level atomic data This is evident in business firms are abandoning traditional methods of data cubes and data aggregates in favor of running specific data queries The latter are becoming more costeffective because of the decreased costs of data aggregation and more useful because of new forms of marketing CVS data consultant Adrian R Sosa who manages the CVS ExtraCare card program agrees We as a department are moving away from data cubes and aggregates and toward more atomic data because of strains on our IT department 6 Case Study CVS Competitive pressure recently forced CVS to develop an on line digital photograph division Rather than build new infrastructure though they partnered with Kodak com for efficiency purposes Customers now submit pictures via their Kodak com account and pick them up printed in CVS stores CVS however is unable to link customers Kodak com accounts to their CVS instore account so they cannot provide them with the same price discrimination and strategic marketing services However as is typical in many businesses IT specialists are feeling pressure to integrate the two accounts Sosa predicts that in five to ten years customers will possess one account for both web sites7 whether they are aware or not Threat Model Dangers of Data Aggregation Although data aggregation is useful and often essential it poses significant dangers to consumers A Data aggregation as an activity can harm consumers by exposing personal information to outsiders that individuals wish to keep private This occurs in two ways Data collection increases the amount of personal information that could be revealed to 4 Lecture by Jim Waldo 3 20 2007 CS199r Harvard University Conversation with Sally Hertz 5 7 2007 6 Conversation with Adrian R Sosa 5 8 2007 7 Conversation with Adrian R Sosa 5 8 2007 5 unwonted parties and data analysis increases the number of parties that a given piece of information is exposed to B Targeted advertising that is enabled by data aggregation can harm consumers as an irritant and also because it has significant persuasive power Recent findings in behavioral psychology have shown that consumers often do not purchase goods rationally and it is likely that advertising has the power to capitalize on such shortcomings C Price discrimination that is enabled by data aggregation can also be harmful to consumers If data is inaccurate