The effect of HMOs on fee-for-service health care expenditures: evidence from Medicare revisited.

This paper implements a new method for calculating the extent of selection in the aged Medicare HMO market. Selection is measured as the difference in average costs between new Medicare HMO enrollees and Medicare fee-for-service stayers with data from 1990 to 1994. Results suggest Medicare HMO enrollees were 1030 US dollars cheaper in their first year of enrollment. The effect is found entirely in Part A (hospital) expenditures, confirming selection is based on inpatient rather than outpatient or preventive care. These results are consistent with previous work.