Luohan Academy

The Great Debate: Data Externalities

As the development of digital technology makes data a crucial driver of innovation and growth, the issue of data governance and privacy protection becomes one of the most urgent issues to address. In a recent paper, "Privacy and personal data collection with information externalities" from the Journal of Public Economics addresses this issue.

Don't know what information externalities are?

It's just the cost or benefits of information to a third party.

You see, when information is shared by one person, it's always correlated with other data types, such as friends, classmates, or people of the same demographic. Therefore, when a person decides to share their data, it may hurt someone's privacy, that is, if they care or even know. Economists call this effect an "information externality" of data sharing. It’s important to remember that information externality can be both positive or negative. For example, patient health data can be used to better predict or forecast disease and illness.

Let's use Facebook as an example – when I agree to share my data with Facebook, say a picture, no one asks each subject in the photo for their permission, but I share their individual information. It is, for this reason, Facebook tends to collect more data than is socially desirable. The third-party, in this case (the subject in the picture), is possibly being harmed at the benefit of Facebook's facial recognition algorithm or the poster’s social life.

This paper is interesting because it highlights the consequences of data externalities and goes on to provide examples for policies and market-based solutions. One such solution is data brokerage that, when used together with General Data Protection Regulation GDPR-like policies, can reduce privacy externalities and increase data productivity by allowing more market participates to use the data while at the same time protecting individuals most sensitive data.  However, it should be noted that plans like the GDPR, which imposes strict rules on a brokerage, may hinder both for exploring the value of data and for privacy protection.

The value of data has the property of increasing return to scale. The paper's conclusion might be underscored by the simple fact that we need to better balance the positive and negative effects of data externalities. That said, while discussed in layman terms here, the topic of data governance and privacy protection is no trivial matter, mainly due to the complicated nature of data usage -- harnessing the value of data while meeting the concerns for privacy.


to leave a comment