Skip to content

Cloos et al. 2019 – Is Your Privacy for Sale?

In brief

An economics experiment with 105 first-semester students at the University of Kassel, testing whether people's self-stated privacy behaviour correlates with the price they demand to have personal information revealed to fellow students (pp. 1, 6) [U]. The authors find a positive correlation with one version of their privacy index, but it disappears with other specifications, so they find no conclusive evidence for or against the ‘privacy paradox’ (pp. 1–2, 9–10) [U]. Regardless of index, the price depends on how delicate the information is, for example a stated wish to lose weight (pp. 1, 10) [U].

Research question and main claim

  • Question: does the privacy paradox hold when individuals decide whether to reveal potentially sensitive information about actual behaviour (p. 2) [U]?
  • Three hypotheses: (H1) self-stated privacy behaviour does not correlate with willingness to accept (WTA) disclosure; (H2) a wish to lose weight, smoking and a poor school grade increase WTA; (H3) taking from (giving to) organisations increases (decreases) WTA (pp. 5–6) [U].
  • Main finding: no conclusive and robust evidence against the privacy paradox; the willingness to protect information depends on how delicate it is (pp. 1, 10) [U].

Theoretical perspective

  • Experimental and behavioural economics of privacy: the literature on the privacy paradox and on the monetary valuation of privacy (Acquisti and others) (pp. 1–4) [U].
  • Mixed evidence in the literature is explained by differences in incentives, salience and transparency (Benndorf and Normann) (p. 1) [U].

Empirical material and context

  • Experiment in October 2018 at the University of Kassel, Germany, during introduction week; two sessions, 105 participants, 99 observations analysed (p. 6) [U].
  • Participants answered a privacy questionnaire, three personal questions (weight, smoking, school grade) and made decisions to donate to or take from three organisations: a local football club, ILGA and UNHCR (pp. 4, 6) [U].
  • The Becker–DeGroot–Marschak (BDM) mechanism elicited the minimum price for having the information read aloud to fellow students (p. 5) [U].

Method

  • Incentivised laboratory-type experiment with questionnaire, OLS and tobit regressions, and factor analysis of privacy indices (pp. 4–10) [U].
  • The privacy index is provisional; the authors note that no adequate index for online privacy behaviour exists (p. 4) [U].

Key concepts

Concept How the source defines or uses it Page U/T
privacy paradox The intention–action gap between stated privacy preferences and actual behaviour 1 U
WTA (willingness to accept) The minimum price a participant demands to have information revealed 2, 5 U
privacy Subjective, idiosyncratic and intertemporal; its value varies between people and information and over time 1 U
DOT decisions Decisions to donate to or take from organisations, with moral loading 2, 4 U
BDM mechanism A standard incentive-compatible method for eliciting WTA 2, 5 U

How privacy and private data are understood

  • Privacy is called a fundamental human right in the first sentence (p. 1) [U].
  • Mainly treated as something with a monetary value that individuals can sell (pp. 1–3) [U].
  • Personal data are described as increasingly used as ‘a kind of currency’ to pay for ‘free’ online services (p. 3) [U].
  • The value of privacy depends on the type of information and on avoiding shameful exposure in front of peers (pp. 3, 10) [U].
  • Protection and sharing of personal information involve costs and benefits for individuals and society (p. 3) [U].

Arguments and findings

  • With the additive index, a one-unit increase in privacy behaviour increases WTA for personal details by about €11–12 (p. 8) [U].
  • The effect disappears with indices based on factor analysis (p. 9) [U].
  • The result from the additive index is driven mainly by the question about covering the webcam (pp. 9–10) [U].
  • Stating a wish to lose weight increases WTA by about €14–16; smoking and school grade have no significant effect (p. 8) [U].
  • Reduced donations to ILGA increase WTA for DOT decisions; not for the other two organisations (p. 8) [U].
  • Most participants are willing to forgo considerable earnings to protect their information (p. 10) [U].

Limitations the source itself mentions

  • The privacy index is provisional, and the factor-analysis indices do not meet the usual threshold for acceptable measures (pp. 4, 9) [U].
  • The data disclosure in the experiment has no real-life equivalent, so results are only comparable to other studies to a limited extent (p. 10) [U].
  • Stating a reservation price is itself a disclosure of private information (p. 5) [U].
  • An upper bound on WTA may create anchoring effects (p. 5) [U].
  • Future research needs better indices for privacy preferences (p. 10) [U].

What the source does not cover

  • The state or government use of data [T].
  • Norway; the participants are German students [T].
  • Political or legal analysis [T].

Possible relevance for the thesis

  • An example of privacy being understood as an individual's economic choice with a price, which can be contrasted with collective or rights-based understandings in other sources [T].
  • If Norwegian documents refer to people's ‘choice’ or ‘consent’, the privacy paradox literature is relevant background [T].
  • Limited direct relevance for analysis of government documents [T].

Connections to other sources

  • Fairfield & Engel 2015 – thematic: both address why people say they value privacy but still disclose information. Fairfield & Engel explain it as a social dilemma (p. 425 there), Cloos et al. test it experimentally [T].
  • Solove 2011 – thematic contrast: Solove criticises reducing privacy to hiding bad things; this experiment measures willingness to reveal ‘bad deeds’ [T].

Key quotes

“Privacy is a fundamental human right; but is yours for sale?” (p. 1)

“Privacy is subjective, idiosyncratic and intertemporal; its value varies between different people and different information, and it may change over time” (p. 1)

“indicating that the willingness to protect private information depends on the delicacy of the information at stake.” (p. 1)

“However, we do not find conclusive and robust evidence for behavior which is not in line with the privacy paradox.” (p. 10)

Open questions for Rebekka

  • What role does this source play for you – as an example of the economic understanding of privacy, or as background on the privacy paradox?
  • Do Norwegian documents assume that people make informed choices about their data? Does the privacy paradox literature challenge that?
  • Is it relevant that the experiment concerns disclosure to peers, not to the state or companies?