The proposed project will help to clarify the process by which mortality and functional health combine to determine differences in total, active and inactive life expectancy by age, gender, race-ethnicity and education for the U.S. population 51 years of age and older. It will lead to increased knowledge regarding the length of life spent in different states of function and transitions between functional states which have important policy implications given pending increases in the older population, and particularly toward an increased understanding of the special needs of disadvantaged groups. Data from the Health and Retirement Study (HRS) and the Medical Expenditure Panel Survey (MEPS) will be applied. Sufficient data are available to examine disparities for Hispanics, African Americans, and non-Hispanic whites. The proposed project will make the following contributions. (1) Evaluate alternate measures of functional health across age, sex, race-ethnicity and education groups, and the impact of the type of measure on conclusions regarding disparities in functional health. (2) Calculate estimates of active life expectancy by age, sex, race-ethnicity and education with nationally representative data from 1996-2002. (3) Provide the first estimates of active life expectancy based on longitudinal data that include persons in middle ages and separate estimates for Hispanics. (4) Evaluate whether there is evidence of bias associated with the assumption of a single transition between interviews by comparing results of traditional panel and embedded Markov chain methods, and examine whether there are systematic differences in estimates of transitions across sex, race-ethnicity and education groups based on the two methods. (5) Examine the impact of interval duration and missed transitions in functional status on estimates of active life expectancy derived from panel data under the two methods listed above. (6) Apply microsimulation and bootstrapping techniques to estimate sampling variation and confidence intervals. With the sample including data for individuals at the tip of the Baby Boom generation (born 1947 and earlier), the information derived from the proposed study will help to plan for the pending increases in the older population. ? ?