We study a general equilibrium environment in which the only activity of interest is armed robbery. Agents choose to be citizens or robbers, and to purchase handguns or not. By arming, citizens can protect themselves from robbery. The government chooses the intensity of police efforts to arrest would-be robbers and to arrest citizens who arm for self-defense. Properties of an equilibrium are characterized and comparative statics results are obtained. We then show why empirical work that examines variations in "shall issue" laws could lead to wrong conclusions. Our analysis produces counterexamples to the following propositions: raising the arrest rate of robbers reduces crime; increasing the arrest rate of armed citizens reduces the number of armed citizens (crime rate, gun death rate).Here is a link to an earlier ungated version of the paper. Interestingly, the abstract for the earlier version is somewhat different.
We study a theoretical general equilibrium environment in which the only activity of interest is armed robbery. Agents choose whether to be citizens or robbers, and whether to purchase handguns. Armed citizens can protect themselves from robbery but any armed agent runs the risk of accidentally shooting himself or another agent. The government chooses a gun tax, and the intensity of police efforts to arrest would-be robbers and citizens who arm for self-defense. Properties of equilibrium are characterized and the model is calibrated and solved. In all cases unique equilibria are obtained. We find that guns are an inefficient way of redistributing wealth, in the sense that social costs are very large relative to the actual wealth redistribution. In this model society would be vastly better off if handguns could be eliminated. We do find, however, that handguns substantially deter crime when crime is defined as taking another's wealth by force. Yet handguns cause accidental deaths and resultantly in this model policymakers confront a fundamental trade-off between property rights and gun deaths.I have not read both versions of the paper to see how much the content of the paper had changed, but it is interesting to see how the peer review / journal publication process changes the tone of an abstract. There are plenty of changes, but to me, the most substantive change to the abstract is the target audience. The published abstract (top) appeals much more to an economic theorist, while the working paper abstract (bottom) appeals more to someone interested in the determinants of crime. This is no accident. The paper is published in Economic Theory.