Abstract
This study provides a summary of the capacity of the Medical Expenditure Panel Survey (MEPS) to produce national estimates of the continuously uninsured population over a four year time frame and the adopted estimation strategy. The paper reports on the employed methodology that permits more accurate estimates of the long term uninsured than heretofore derived. This is accomplished through the linkage of the survey to the National Health Interview Survey, allowing for a reconciliation of insurance coverage responses covering overlapping time periods. The production of national estimates of the long term uninsured for a four year interval is a new MEPS output, and this study identifies the characteristics of the uninsured. The utility of this information to improve behavioral models of health care expenditure patterns is also described.