The most conclusive data on the effect of ADR costs comes from studies of mandatory court-administered nonbinding arbitration. In most jurisdictions too few cases are diverted from trial to balance the public costs of implementing an arbitration program. In addition, some cases that previously would have settled without formal proceedings remain on the court calendar, only to settle just before or just after the arbitration hearing. Therefore, these programs operate more as enhancements to the settlement process than as procedures for diverting cases from trials. The results indicate that any new procedure is likely to produce unanticipated outcomes and that noneconomic as well as economic factors influence the behavior of litigants and attorneys. Similarly, research on family law mediation programs reveal little cost savings, although the evidence is stronger than for court-administered arbitration. Overall, findings suggest that cost results will depend less on the process or entity chosen than on how much ADR programs affect attorneys' and disputants' behaviors. 12 reference notes