The Chilling Effect of Governance-by-Data on Data Markets
24 Pages Posted: 30 Jul 2018
Date Written: July 1, 2018
Big data has become an important resource not only for commerce, but also for governance. Governance-by-data seeks to take advantage of the bulk of data collected by private firms, to make law enforcement more efficient. It can take many forms, including setting enforcement priorities, affecting methods of proof, and even changing the content of legal norms. For instance, car manufacturers can use real-time data on the driving habits of drivers, to learn how their cars respond to different driving patterns. If shared with the government, the same data can be used to enforce speeding limits or even to craft personalized speeding limits for each driver.
The sharing of data for the purpose of law enforcement, raises obvious concerns to civil liberties. Indeed, over the past two decades, scholars have focused on the risks arising from such data sharing for privacy and freedom. So far, however, the literature has generally overlooked the implications of such dual use of data for data markets.
In this Article we argue that governance-by-data may create chilling effects that could distort data collection and data-driven innovation. We challenge the assumption that incentives to collect data are given, and that firms will continue to collect data, notwithstanding governmental access to such data. We show that in some instances an inverse relationship exists between incentives for collecting data, and sharing it for the purpose of governance. Moreover, the incentives of data subjects to allow the collection of data by private entities might also change, thereby potentially affecting the efficiency of data-driven markets and subsequently, data-driven-innovation. As a result, data markets might not provide sufficient and adequate data to support digital-governance. This, in turn, might significantly affect welfare.
Keywords: data, governance, big data, algorithm, enforcement, privacy, personalized law
JEL Classification: K00
Suggested Citation: Suggested Citation