Wednesday, 16 December 2009

"I've got nothing to hide", rebutted - summary of Solove's paper

I only came across Prof Daniel Solove and his work relatively recently.

For anyone who's not yet read his brilliant analyses, here's a summary of his superb 2007 paper 'I've Got Nothing to Hide' and Other Misunderstandings of Privacy which pulls apart the "I've got nothing to hide" argument commonly put forward as a reason to justify and legitimise invasion of privacy. The paper was mentioned during the discussion after Bruce Schneier's talk on the future of privacy.

This is my own take on it so it's not necessarily in the same order as in his paper, but I hope and believe it correctly reflects his masterly dissection of the issues. I wholeheartedly recommend that everyone interested in this area should read his original paper.

  1. The right to "privacy" is usually seen as a kind of right to "secrecy", the right to hide bad things, to hide things about you which are negative (or embarrassing, or that you just plain think is no one else's business to know).
  2. The most seemingly compelling form of the "I've got nothing to hide from the government" argument is put as one involving a necessary trade-off between invading individual privacy and protecting national security - in which case of course most people will agree that national security must win out.
  3. But privacy is not an issue of individual interests vs. society's interests. Privacy, freedom from friction and others' intrusiveness, is itself a social value, a form of social control based on society's norms; and its protection is not a question of individual right vs. "greater social good", but a question of one social interest vs. other social interests. The true issue is, how can we strike the right balance between different social goods?
  4. Privacy is very hard to define as a single all-encompassing concept. Attempts to do so have resulted in views of privacy which are either too vague and broad, or too narrow - and this harms rather than helps analyses of the position.
  5. Privacy should be seen as a set of related issues, and privacy problems as a web or cluster of related problems - not necessarily connected by one common element, but which resemble each other. Rather than asking "Is this a "privacy" breach or isn't it?", we should ask: "Is this a problem that should be protected against because it causes harm to something valuable to society?"
  6. Solove defines a privacy problem or privacy violation as occurring when an activity (whether by individuals, government or business) causes harm by disrupting the socially valuable activities of others, e.g. by chilling free speech/association or by resulting in adverse power imbalances in society such as excessive executive power.
  7. He proposed a taxonomy of privacy to model the problems of privacy harms:
    • Information Collection
      • Surveillance
      • Interrogation
    • Information Processing
      • Aggregation
      • Identification
      • Insecurity
      • Secondary Use
      • Exclusion
    • Information Dissemination
      • Breach of Confidentiality
      • Disclosure
      • Exposure
      • Increased Accessibility
      • Blackmail
      • Appropriation
      • Distortion
    • Invasion
      • Intrusion
      • Decisional Interference
  8. Many people approach the privacy problems arising from the collection and use of personal data from the angle of Orwell's 1984 Big Brother metaphor, which focuses on the harms of surveillance (social control, inhibiting free speech etc) - but in fact most collected personal data isn't sensitive in itself, its collection wouldn't inhibit free speech, and most people wouldn't care very much about its being collected. Hence, many people will say, "Go ahead, I've nothing to hide".
  9. In fact, the better metaphor in relation to collection/use of personal data is Kafka's The Trial, about a faceless bureaucracy that uses personal data to make important decisions about people who are excluded from having any control or even say about the use of their data for purposes unknown.
  10. The problem isn't so much data collection but information processing, which alters the power relationship between citizen and state and also creates, not necessarily inhibition or chilling, but a feeling of helplessness and powerlessness.
  11. Data aggregation, by combining seemingly non-sensitive separate bits of information, may well reveal additional and possibly even sensitive information; so, without knowing exactly what information is deduced by the data mining software, we can't say definitively that it won't reveal any data that we'd want to hide.
  12. Initiatives such as the US National Security Agency's data collection and data mining, even if they don't uncover any information that people might want to hide, still cause privacy problems because they result in the Kafkaesque problems of bureaucracy (rather than surveillance) - "suffocating powerlessness and vulnerability… indifference, errors, abuses, frustrations, lack of transparency and accountability".
  13. Data mining and profiling also tries to predict future behaviour. If you are matched to a particular profile, that means that they (or the software) think that you are likely to follow a particular pattern of behaviour in future. But how can you deny something you've not done yet? "Having nothing to hide will not always dispel predictions of future activity". Are you happy to be judged as being a dodgy person in some way (whether to national security, or for insurance purposes) because of your "profile", which you may not even know about let alone have any control over?
  14. Data mining causes "exclusion" problems. You don't know what data is being held about you (or sometimes even that data is being held about you), let alone have the power to correct any errors in the data (e.g. that you are a bad credit, when in fact it's someone else who has the same name as you or used to live at your address).
  15. Thus the problem is not about whether the data collected is or is not something people want to hide; it's really about the structure and power of government: how government treats citizens, and the power imbalance between citizen and the executive branch of government [i.e. a separation of powers and checks and balances issue, fundamental to democracy].
  16. Similarly with the secondary use of data, where data collected for one purpose is then used for a different unrelated purpose without the person's consent. The potential uses may be endless, but people just can't properly evaluate the risks of the government (or whoever) having their data, because there are no limits, transparency or accountability. Again it's a power imbalance issue.
  17. Secondary use may also involve breaches of confidentiality or contract - e.g. some US airlines gave passenger records to government agencies without passenger consent, in breach of their privacy policies. There is a social interest generally in ensuring that promises are kept and that trust in business/customer relationships is maintained, and specifically in businesses/government keeping within any stated limits on the way they use personal data. If government/businesses can use personal data in any way they choose, then stated limits are meaningless and consumers are powerless. Again, the power imbalance is a structural harm.
  18. Privacy problems are mostly thought boring as generally they "lack dead bodies". Usually the threat to privacy is not from single obviously egregious actions, but "by a slow series of relatively minor acts which gradually begin to add up". Solove draws an analogy with some types of environmental harms which involve gradual pollution from lots of sources, rather than one big spill.
  19. So, in a nutshell, it's not a question of balancing privacy against security, it's not a question of whether government should or should not be allowed to engage in surveillance or data collection/analyses activities - the true question is: should the unelected executive branch of government be allowed to do these things without adequate (or any) judicial oversight (e.g. a requirement to obtain warrants) or data minimisation?
  20. Most democratic societies would hopefully answer "No" to that question. And while Solove's analysis is primarily in relation to information gathering by government, it applies equally, if not more so, to data collection and data mining by businesses in relation to their customers and others.
It's very worthwhile finding the time to read the paper if you're interested in these issues. The link again - 'I've Got Nothing to Hide' and Other Misunderstandings of Privacy.

©WH. This work is licensed under a Creative Commons Attribution Non-Commercial Share-Alike England 2.0 Licence. Please attribute to WH, Tech and Law, and link to the original blog post page. Moral rights asserted.