Economic Value of Data: Quantification Using Online Experiments

How to cite

Sen, Ananya; Moehring, Alex; Collis, Avinash; Acquisti, Alessandro. 2021. "Economic Value of Data: Quantification Using Online Experiments." AEA RCT Registry. AEARCTR-0006727. https://doi.org/10.1257/rct.6727-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Ananya Sen* , Alex Moehring , Avinash Collis , Alessandro Acquisti

* Primary investigator

Registry ID
AEARCTR-0006727
DOI
https://doi.org/10.1257/rct.6727-1.0
First registered
Last updated
Status
In development
Trial dates
2021-06-21 – 2021-08-25
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/6727

Abstract

We conduct incentive-compatible choice experiments to measure the economic value of digital data. We focus on the value that users place on their personal data related to the biggest social media platform: Facebook. We will use BDM methods to elicit their valuation for their Facebook data. For the baseline valuations, we simply ask them their WTA to part with all of their Facebook data. After eliciting their baseline valuation, we will randomize information to individuals about Facebook settlements after data breaches and as well as the amount of revenue per user generated by Facebook. We will then evaluate how individuals update (or not) their valuations based on the information. Finally, we will randomly draw from a distribution and invite selected individuals to upload their data and receive their stated valuation.

Intervention

We will recruit individuals to our survey page and ask them to take the survey on the value of data. We will elicit baseline valuations about their WTA for their Facebook data. After eliciting these baseline valuations, we will randomly assign each individual to one of two information treatment conditions. One condition highlights the fact that Facebook paid each individual $400 in a data breach settlement as part of a lawsuit. Another condition highlights the revenue per user earned by Facebook is about $400 over three years going forward. We allow individuals to revise their valuations if they want to and we measure their valuations again.

These are incentive-compatible since we will draw valuations at the end of the experiment (in real-time) and if their stated valuation is less than the drawn number then the individual will be asked to part with their Facebook data. We use an algorithm to ensure that we never actually access their data but simply record some de-identified summary stats allowing us to ensure that the data is actually legitimate.

Experimental design

Our survey experiment will start with the baseline valuation question:

(1) What is the minimum amount of money (in US Dollars) you would require to upload all your Facebook data? This includes your posts, photos, messages, likes, and comments.

The information treatments about value accruing to individuals from data breaches and as well as how much money Facebook makes per user will be the following:

(2) To provide some additional context, Facebook recently lost a class action lawsuit for harvesting user data and violating privacy laws and agreed to pay around $400 per user for eligible users (source).

(3) To provide some additional context, Facebook is expected to earn over $400 per North American user over the next three years (source).

Design details

Randomization method
Randomization will be carried by the survey software - Qualtrics.
Randomization unit
Randomization will be at the individual level.
Clusters
The number of clusters will be the number of individuals recruited for the survey experiment.
Observations
We expect about 5000 individuals in the study.
Arms
We expect a balanced number of individuals across different treatment arms.
Primary outcomes
The main outcome of interest will be the value individuals assign to their Facebook data. We will ask them for a monetary amount. Additionally, we will also analyze how individuals update their data valuations based on the information treatments.

Cited by

No citation data available for this registration yet.

Keywords