Published November 5, 2019 | Version v20191105

Suboptimal Provision of Privacy and Statistical Accuracy When They are Public Goods

  • 1. U.S. Census Bureau and Cornell University
  • 2. University of Georgia
  • 3. Cornell University

Description

We explain why population statistics are provided by public statistical agencies rather than private firms. To do so, we focus on inefficiencies in how private providers trade off data privacy and accuracy.  Increasing the accuracy of published statistical summaries necessarily results in a loss of privacy for the data owners. Data publication is based on differential privacy. Privacy protection and accuracy are public goods We find that private provision results in suboptimally low data accuracy. The external benefit of data accuracy to all consumers is not captured by the willingness-to-pay of the consumer with the greatest private value.The provider buys just enough data-use rights (privacy loss) to sell the data accuracy to the consumer with the highest valuation.

Notes

The views expressed in this paper are those of the authors and not those of the U.S. Census Bureau or other sponsors. This research was partially funded through NSF Grant #1131848 (NCRN) and Alfred P. Sloan Foundation Grant G-2015-13903.

Files

AbowdSchmutteSextonVilhuber2019_TPDP_Poster.pdf

Files (231.2 kB)

Name Size Download all
md5:3d11917a4e8035caf6b89649bbc58d4d
231.2 kB Preview Download

Additional details

Funding

U.S. National Science Foundation
NCRN-MN: Cornell Census-NSF Research Node: Integrated Research Support, Training and Data Documentation 1131848