The constant multiplier assumption misestimates long-term sex offender recidivism rates

Many clinical psychologists have claimed that long-term sexual recidivism rates are a fixed multiple of short-term rates and have estimated that the true value of this constant falls somewhere between 1.5 and 3.0. They have also proposed that it is "mathematically sound" for evaluators to estimate the long-term rate for any actuarial score in sexually violent predator civil commitment cases by multiplying its short-term rate by this constant. The present paper questions the "constant multiplier assumption" and summarizes disconfirming data collected by its proponents and others showing that the fixed ratios for groups with low short-term rates are actually greater than the ratios for groups with high short-term rates. These results rule out the use of the constant multiplier assumption by risk evaluators. It is concerning that this assumption has not been previously tested. The authors call on the developers of risk assessment systems to collect and report data that clearly validate the assumptions that underpin their actuarial tables before they are disseminated or administered. The American Psychological Association ethical standards also require forensic evaluators to acknowledge the limitations of their risk assessments when they testify, a practice that is invaluable to the trier of fact.