OBJECTIVE: We assess whether physician turnover stems from incorrect physician expectations about the practice environment or from actual constraints or rewards in that environment. DATA SOURCES: Our primary data source contains information about individual HMOs' primary care physicians incentive mechanisms and general HMO characteristics. Our secondary data source is the area resource file (ARF), which contains countywide information about the HMOs' market areas, including physician characteristics, population characteristics, and other market characteristics. DATA COLLECTION: Our primary data source is from a nationwide survey of all HMOs in operation in 1987-1988, as reported to Interstudy. PRINCIPAL FINDINGS: We find that turnover is higher on the part of physicians whose HMO enrollment comprises a greater percentage of their total practice. Our results further indicate that physicians whose compensation is dependent on the behavior of a group of other physicians are more likely to leave the plan than those who bear the risk (and control it) more directly. On the other hand, turnover is increased by basing bonuses on individual productivity and by not sharing surpluses among a group. Market characteristics also are significant in explaining physician turnover in HMOs. CONCLUSIONS: It appears that physicians accurately forecast how they will react to individual financial risk, although they dislike restrictions imposed by HMOs.
OBJECTIVE: We assess whether physician turnover stems from incorrect physician expectations about the practice environment or from actual constraints or rewards in that environment. DATA SOURCES: Our primary data source contains information about individual HMOs' primary care physicians incentive mechanisms and general HMO characteristics. Our secondary data source is the area resource file (ARF), which contains countywide information about the HMOs' market areas, including physician characteristics, population characteristics, and other market characteristics. DATA COLLECTION: Our primary data source is from a nationwide survey of all HMOs in operation in 1987-1988, as reported to Interstudy. PRINCIPAL FINDINGS: We find that turnover is higher on the part of physicians whose HMO enrollment comprises a greater percentage of their total practice. Our results further indicate that physicians whose compensation is dependent on the behavior of a group of other physicians are more likely to leave the plan than those who bear the risk (and control it) more directly. On the other hand, turnover is increased by basing bonuses on individual productivity and by not sharing surpluses among a group. Market characteristics also are significant in explaining physician turnover in HMOs. CONCLUSIONS: It appears that physicians accurately forecast how they will react to individual financial risk, although they dislike restrictions imposed by HMOs.