Is Your Privacy for Sale? An Experiment on the Willingness to Reveal Sensitive Information

Round 1

Reviewer 1 Report

This paper investigates whether individuals' self-stated privacy behavior is correlated with their reservation price for the disclosure of personal and potentially sensitive information. And find that there is a significant positive correlation between the willingness to accept (WTA) the disclosure of private information and respondents’ stated privacy behavior and most subjects are willing to forego considerable potential earnings in order to protect both private details and DOT decisions. Although this paper studies a meaningful problem, from my perspective, it does have the following problems that the author may want to address.

1.      This paper does not explain clearly the reasons for choosing the problems used in the experiment. In my opinion, the setting of these questions greatly determines the value of the experiment. Especially after the experiment concluded that the result is mainly driven by the individuals’ stated decision to cover up the webcam of their laptop, there is no further discussion on the problem setting module.

2.      Participants were already endowed with a sizeable monetary amount at the start of the DOT decision section. This could possibility change participants’ decisions. The paper uses a hypothesis to circumvent the above problem but does not test the correctness of the hypothesis.

3.      The authors fail to cite several past literatures (e.g., [1-4]) that are highly related to this work.  

[1] Arpita Ghosh, Aaron Roth, “Selling Privacy at Auction”, in EC 2011.

[2] Haiming Jin, Lu Su, Houping Xiao, Klara Nahrstedt, "INCEPTION: Incentivizing Privacy-Preserving Data Aggregation for Mobile Crowd Sensing Systems", in MobiHoc 2016.

[3] Chaoyue Niu, Zhenzhe Zheng, Fan Wu, Shaojie Tang, Xiaofeng Gao, Guihai Chen, “Unlocking the Value of Privacy: Trading Aggregate Statistics over Private Correlated Data”, in KDD 2018.

[4] Haiming Jin, Lu Su, Houping Xiao, Klara Nahrstedt, "Incentive Mechanism for Privacy-Aware Data Aggregation in Mobile Crowd Sensing Systems", in TON 2018.

Author Response

Reply to Reviewer 1

Manuscript-ID: games-510441

We would like to thank both reviewers for their valuable comments and the time and effort invested. We edited the manuscript substantially to address all of the raised issues and we think that the manuscript has become much stronger in this process. Below we address the reviewers’ comments point by point and explain the revisions we have made to the manuscript.

1. We agree with the reviewer that we have not yet explained how we selected the problems used. This issue was also mentioned by the other reviewer. In Section 3.1 (Experimental design) we now discuss the rationale for selecting the problems.

2. We thank the reviewer for mentioning this point. We agree that the absolute values of the DOT decisions should be interpreted with care. However, we believe that windfall money effects should not matter for the results of our study, since they will lead to level effects.  In the first manuscript we briefly discussed the results for Hypothesis 3 in section 4.2 (Main regression results). In the revised manuscript we now also discuss the results for Hypotheses 2 and 3 in section 4.3 (Robustness). In addition, we report how many participants have taken or donated money from at least one or two organizations in section 4.1 (Descriptive statistics). We also note that the WTA_DOT do not differ significantly within these groups.

3. We thank the reviewer for making us aware of relevant literature that we overlooked in the first manuscript. In the revised manuscript, we cite the Ghosh and Roth (2015), Jin et al. (2018) and Niu et al. (2018) in a further sentence in section 2 (Related literature). We also cite Ghosh and Roth (2015) at the end of section 3.1 (Experimental design).

In addition, we would like to inform the reviewer that we have made another change to the first manuscript. In the first manuscript in Table B6 the labels of models 2 and 5 were reversed. We have corrected this mistake in the revised manuscript.

Literature

Ghosh, A.; Roth, A. Selling privacy at auction. Games and Economic Behavior 2015, 91, 334-346; DOI: 10.1016/j.geb.2013.06.013.

Jin, H.; Su, L.; Xiao, H.; Nahrstedt, K. Incentive mechanism for privacy-aware data aggregation in mobile crowd sensing systems. IEEE/ACM Transactions on Networking (TON) 2018, 26, 2019-2032; DOI: 10.1109/TNET.2018.2840098.

Niu, C.; Zheng, Z.; Wu, F.; Tang, S.; Gao, X.; Chen, G. Unlocking the Value of Privacy: Trading Aggregate Statistics over Private Correlated Data. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining; ACM, 2018, pp. 2031-2040; DOI: 10.1145/3219819.3220013.

Reviewer 2 Report

This study is concerned with the so-called privacy paradox, namely, the finding by privacy researchers that there is a dichotomy between privacy attitudes and behavior.

The paper carries out an experiment with undergraduate students in which the author(s) elicit both privacy attitudes through a survey and the willingness to accept the disclosure of private information in an incentive compatible way.

The main finding is that there is evidence of a correlation between the stated attitudes toward privacy and the WTA, which refutes the privacy paradox. However, this correlation is sensitive to the way the measure of privacy concerns is constructed.

This is a well-written paper on a topic that is very topical.

COMMENTS

- The contribution of the current study to our understanding of the phenomenon in question (privacy paradox) is not clear to me. The author(s) need to do a more convincing case of explaining what is original about the study and how it advances the literature.

- The type of personal details and the identity of the party to which they are released to are bound to be important for privacy behavior. This study makes use of a very specific and limited set of items of personal information that are elicited relative to previous studies and offers to share this personal information with classmates and not a third party, such as an internet company etc. These choices need to be discussed and reflected upon as they could have an important bearing on the findings.

- Please be clear as to how you define the variable "privacy" in Table 2. 

- Results are not robust to alternative measures of privacy concerns. This is disconcerting as it seems that depending on how you construct your measure you can find evidence for or against the privacy paradox. The abstract and introduction should reflect this fact.

Author Response

Reply to Reviewer 2

Manuscript-ID: games-510441

We would like to thank both reviewers for their valuable comments and the time and effort invested. We edited the manuscript substantially to address all of the raised issues and we think that the manuscript has become much stronger in this process. Below we address the reviewers’ comments point by point and explain the revisions we have made to the manuscript.

1. We are very grateful to the reviewer for making us aware that the contribution of our paper to the existing literature has not yet been sufficiently explained. In the introduction we now discuss that although our experiment provides only inconclusive evidence on the privacy paradox, it nevertheless makes an important contribution to the literature, since it provides a design that monetarily quantifies the value that is individually assigned to the information that can potentially be revealed. In this context, we also address the advantages and disadvantages of requesting existing Internet data. Our paper further highlights that future research would benefit from a more comprehensive index for privacy on the internet.

2. We fully agree with the reviewer that we have not yet explained why we have chosen this very specific and limited set of items of personal information.  This point was also noted by the other reviewer and is now explained in section 3.1 (Experimental design). The fact that the information in our experiment could be revealed to classmates and that this disclosure method has no equivalent in real life is now discussed in section 5 (Conclusion). Regarding this point, we also mention the fact that our results are only partially comparable with the results of Benndorf and Normann (2018) and Schudy and Utikal (2017).

3. We thank the reviewer for pointing out the missing explanation of the privacy variable in Table 2. We now specify this variable at the beginning of section 4.2.

4. We are grateful to the reviewer for drawing our attention to the fact that so far we have not made sufficient reference to this point. In the revision of the manuscript, both the abstract and the introduction address the fact that our results on the privacy paradox are inconclusive. Furthermore, in section 4.3 (Robustness) we now point to the fact, that we still find partial support for Hypotheses 2 and 3 when we use alternative privacy indices.

In addition, we would like to inform the reviewer that we have made another change to the first manuscript. In the first manuscript in Table B6 the labels of models 2 and 5 were reversed. We have corrected this mistake in the revised manuscript.

Literature

Benndorf, V.; Normann, H.-T. The willingness to sell personal data. Scand. J. Econ. 2018, 120, 1260–1278; DOI: 10.1111/sjoe.12247.

Schudy, S.; Utikal, V. `You must not know about me’ On the willingness to share personal data. J. Econ. Behav. Organ. 2017, 141, 1–13; DOI: 10.1016/j.jebo.2017.05.023.

Round 2

Reviewer 2 Report

Thank you for responding to my comments

Author Response

Again, we would like to thank the reviewer for his/her valuable comments and suggestions. A native English speaker has read the new version of the manuscript. Based on his comments, we have slightly modified the wording in lines 163 and 205.