I’ll be talking at the National Institute of Justice (810 7th. St., N.W., Washington) this Thursday, October 12, at 1 pm. If you’re interested in hearing the talk, please email me at markarkleiman (at) gmail (dot) com and I’ll ask my hosts to add you to the guest list. The seminar breaks promptly at 2 p.m. I think they’re serving lunch, though I’m not sure.
I’ll be giving more or less the same talk later in the term at George Mason and at the Harvard Center for Risk Analysis. Beau Kilmer, Andrew Morral, and I should have a draft of the paper sometime this month.
Here’s the abstract:
The Dynamics of Deterrence
When enforcement capacity is constrained, potential offenders face a interdependent choices: the greater the frequency of offending, the less the risk of punishment for any given offender. This can generate a “tipping” situation in which both high violation rates and low violation rates are self-sustaining in the face of the same underlying characteristics of potential offenders and the same level of enforcement capacity.
If instead we fix the ratio of sanctions to violations, then as long as rate of offenses slopes downward as a function of the sanctions risk, and above some level is more than unit-elastic, the amount of sanctions capacity used will form an inverted U.
Under these assumptions, a temporary increment to enforcement capacity can produce lasting benefits by “tipping” a high-violation-rate equilibrium into a low-violation-rate equilibrium. Even where no additional capacity is available, a strategy of sequential concentration, in which a subset of offenders is singled out for attention and that subset is expanded as falling violation rates free up capacity, can have the same effect.
A simple agent-based simulation demonstrates these phenomena, instantiated in a variety of actual programs including the grafitti and squeegee crackdowns in New York, Operation Cease-Fire in Boston, the low-arrest drug crackdowns in High Point, NC, and elsewhere, and the HOPE probation project in Hawai’i. Full exploitation of the underlying principle could lead to reduced crime and reduced incarceration.