fact, if recorded correctly and accurately, do not change and thus are permanent, although their meaning is subject to interpretation and those interpretations may change—e.g., what to make of an individual who undergoes a sex change operation. Names and addresses do change with some frequency, although one may be able to make some general sociodemographic inferences with knowledge of such changes over time. An individual’s DNA sequence does not change throughout his or her lifetime, but the longevity and stability of many other biometric indicators have not been definitively established.

Individuals vary considerably in their privacy demands or expectations for different kinds of data and for the same individual data element in different situations. That is, in one situation, an individual may regard a particular data element as highly private (one that might require a large bin size) and in a different situation regard the same data element as not at all private (i.e., he would be perfectly fine with a bin size of one). Relevant situational factors may include:

  • The specific value of the data element and whether or not it stigmatizes or disadvantages. For example, an HIV-positive individual may require a bin size of one million to feel that his HIV status is private; an HIV-negative individual may feel entirely comfortable with a bin size of one (i.e., being identified with certainty as being HIV-negative).

  • The stated purpose for which any given data element is requested. The closer the fit between the goals of the supplier and the requester of information and between the information requested and the goal, the more likely it is to be provided. In most doctor-patient contexts, the patient is only too glad to offer information. If a newspaper’s Web site asks a visitor her income, she may refuse to provide it, whereas she would willingly supply that same information in filling out an online application for a mortgage. Note also that if there is an incentive or reward for supplying personal information, many consumers sell that information more cheaply than their statements about the value of their personal information would lead one to expect.

  • The accessibility of the given data element. Data that are public and hard to access (e.g., paper records, such as property taxes or divorce proceedings, that are kept in the physical facilities of many jurisdictions) are very different from data that are public and very easy to access (e.g., the same public information posted online). The ease or difficulty of finding a particular type of data element may also contribute to accessibility.

  • The transience of the given data element. For example, when information is stored in paper form, it may be discarded eventually because it is expensive to store. There may be different privacy implications if the data

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement