Alessandro Acquisti and Catherine E. Tucker, Guns, Privacy, and Crime
Comment by: Aaron Burstein
PLSC 2010
Workshop draft abstract:
In December 2008, a Memphis newspaper made publicly available an online, searchable database of all gun permit holders in Tennessee. The database included information such as the permit holder’s name, ZIP code, and his or her permit’s start and expiration dates. It did not receive much attention until, in February, an article about a parking argument that ended in a deadly shooting referred to it. The fierce debate which thereafter arose – with the NRA accusing the newspaper of a “hateful, shameful form of public irresponsibility,” and the newspaper standing by a “right to know” argument – exemplifies the complex interactions, and sometimes collisions, between privacy and other rights and needs. In this case, individual privacy rights collided with the collective right to know, and, arguably, with both individual and communal issues of security.
By preventing the release of personal data, individuals often hope to prevent harm to themselves. However, the publication of the gun permits data highlights one case where privacy and personal security may appear to be in conflict. Whereas gun rights advocates suggested that the publication exposed gun owners to risk (for instance, of criminals targeting houses known to hold guns, in order to steal them), those defending it argued that gun owners may be less likely to be targeted, precisely because the information was made publicly available. In this manuscript we attempt to quantify the actual impact that the publication of TN gun permits data had on 1) crime rates and 2) gun permit requests in the city of Memphis. Combining gun, crime, demographic, and location data from an array of sources and databases, we measured how rates of occurrences of different classes of crime changed, as function of the local density of gun ownership made public by the newspaper, before and after the publication of the database. Our results suggest that the publication of the database reduced more significantly the occurrence of violent crimes (such as robberies and burglaries) in ZIP codes with higher gun ownership density. At the same time, the publication was accompanied by a more significant percentage increase in gun permits requests in areas with pre-existing higher rates of gun ownership. To address concerns about unobserved heterogeneity, we also performed a falsification test by studying crime trends in a similar town (Jackson) in a neighboring state. We found no similar trends in crime during the time period in such town.
This paper contributes not just to the policy debate on the openness or secrecy of gun data (19 states allow the public to access gun permits information; other states either have no laws addressing the issue, or keep the information outside the public domain), but to the broader discourse on the boundaries and connections between privacy and security.