Profiling, Screening and Criminal Recruitment

We model major criminal activity as a game in which a law enforcement officer chooses the rate at which to screen different population groups and a criminal organization (e.g., drug cartel, terrorist cell) chooses the observable characteristics of its recruits. Our model best describes smuggling or terrorism activities at borders, airports, and other security checkpoints. When the social costs of crime are high, law enforcement is most effective when it is unconstrained in its ability to profile, that is its ability to screen different population groups with different probabilities. For more moderate costs, the most effective law enforcement policy imposes only moderate restrictions on the officer's ability to profile. In contrast to models of decentralized crime, eliminating profiling by law enforcement is never optimal