This Article considers a widely overlooked consequence of having a criminal record in the digital age: the spread of inaccurate or outdated criminal record information. Remarkably common, errors in criminal record data quickly multiply across digital platforms and are nearly impossible for people to manage. Error can begin in governmental sources and spread into the private sector or can be introduced by data aggregators as information across jurisdictions and agencies is compiled into databases and web content. For the subject of the record, error can pose enormous obstacles to securing employment and housing, particularly as automated decision-making and algorithmic governance transform traditional institutional processes. Yet, those who are harmed have very few rights regarding the ability to identify and remedy data error.
Part I of the Article introduces the issue of data error in criminal background checks and describes the scope of the problem. Parts II and III describe how and why criminal record data occurs and detail the specific harms through several theoretical lenses: data error as a due process and equal protection harm, as an informational privacy harm, and as a reputational harm. Part IV analyzes legal obstacles that limit remedies, with a particular focus on the practical obscurity doctrine, the Fair Credit Reporting Act, standing, and various legal immunities available to governments and the private sector. The analysis shows how regulating criminal record data has failed in a digital environment and how existing law fails to protect people from unfounded and illegal discrimination on the basis of inaccurate criminal record information. Part V argues that bad data should be conceptualized under broader critiques of racialized, algorithmic injustice and offers solutions for better regulating and using criminal records.
* Associate Professor, Rutgers University-Newark School of Criminal Justice. The author thanks David Noll, Wayne Logan, Eric Goldman, Marty Berger, Vincent Chiao, Hadar Dancig-Rosenberg, Aya Gruber, Itay Ravid, David Alan Sklansky, Jim Francis, and David Lopez for their helpful comments and guidance.
The full text of this Symposium is available to download as a PDF.