BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Revisiting the Economics of Privacy: Population Statistics and Con
 fidentiality Protection as Public Goods - Ian Schumutte (University of Geo
 rgia )
DTSTART:20161028T103000Z
DTEND:20161028T113000Z
UID:TALK68656@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:<span>Co-author: John M. Abowd (Cornell University and U.S.  C
 ensus Bureau) <br></span> <span><br>We consider the problem of the public 
 release of statistical information  about a population&ndash\;explicitly a
 ccounting for the public-good properties of both  data accuracy and privac
 y loss. We first consider the implications of adding the  public-good comp
 onent to recently published models of private data publication  under diff
 erential privacy guarantees using a Vickery-Clark-Groves mechanism and  a 
 Lindahl mechanism. We show that data quality will be inefficiently  under-
 supplied. Next\, we develop a standard social planner&rsquo\;s problem usi
 ng the  technology set implied by <span>(&epsilon\;\,&delta\;)<img alt="" 
 width="28" height="30"></span>  -differential privacy with <span>(&alpha\;
 \,&beta\;)<img alt="" width="28" height="30"></span>  -accuracy for the Pr
 ivate Multiplicative Weights query release mechanism to  study the propert
 ies of optimal provision of data accuracy and privacy loss when  both are 
 public goods. Using the production possibilities frontier implied by  this
  technology\, explicitly parameterized interdependent preferences\, and th
 e  social welfare function\, we disp lay properties of the solution to the
  social  planner&rsquo\;s problem. Our results directly quantify the optim
 al choice of data  accuracy and privacy loss as functions of the technolog
 y and preference  parameters. Some of these properties can be quantified u
 sing population  statistics on marginal preferences and correlations betwe
 en income\, data  accuracy preferences\, and privacy loss preferences that
  are available from  survey data. Our results show that government data cu
 stodians should publish  more accurate statistics with weaker privacy guar
 antees than would occur with  purely private data publishing. Our statisti
 cal results using the General Social  Survey and the Cornell National Soci
 al Survey indicate that the welfare losses  from under-providing data accu
 racy while over-providing privacy protection can  be substantial.</span> <
 br><br>Related Links <ul> <li><a target="_blank" rel="nofollow">http://dig
 italcommons.ilr.cornell.edu/ldi/22/</a>  - Working paper&nbsp\;</li></ul>
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR