Factors that affect the HMO enrollment decision: a tale of two cities.

Although there is agreement that utilization patterns in HMOs differ from those in non-HMO systems, there is less agreement on how and why these differences are achieved. Selection bias is one frequently offered explanation. In an attempt to identify those individual and site-specific factors that help determine employees' choice of coverage, we developed and tested a logistic model of the enrollment decision. We used both survey data and claims data to study the decision to enroll in either an HMO or a traditional indemnity plan among state employees in Tallahassee, Florida, and Salt Lake City, Utah. We found that many factors affected the decision--factors that could be altered by an HMO or insurance company to correct selection bias if and when it is encountered.