We discuss the rationale for benefit carve-out contracts in general an
d for mental health and substance abuse in particular. We focus on the
control of adverse selection as a principal explanation and find that
this is consistent with the widespread use of sole-source contracting
with periodic rebidding. We also find that some degree of risk sharin
g is common; we interpret this as a method of balancing cost-containme
nt incentives with incentives to maintain access and quality on unmeas
ured dimensions.