The magnitude and variation in the arrest risk per crime for individual offenders, termed Q,1 are fundamental to any characterization of crime and crime control. Obviously, Q has intrinsic value as an indicator of criminal justice performance, as well as a measure of the risk of detection faced by offenders. The variation in Q across offenders is also important as a methodological support to enable use of the widely available arrest data to make inferences about offending behavior. In particular, knowing Q is vital in attempts to estimate individual crime rates, λ, from such arrest data (Blumstein and Cohen 1979; Cohen 1986).