Transition probabilities to dependency and death in Italy : evidence from IT SILC survey

The increasing of retirement age in order to maintain an active population able to compensate for the growth in the number of pensioners is a central issue in the European debate. Central to the discussion surrounding the extension of the active lifespan is the state of health of the elderly : on one hand the health condition of the population is clearly crucial to all hypotheses linked to the extension of active life, on the other hand, European governments are specifically interested in forecasting the health care needs and services which will be required by the elderly. The calculation of health expectancy in Italy is generally based on the so-called Sullivan method, while the availability of the longitudinal module of the European survey on income and living condition, and the reliability of the Minimum European Health Module, allows to measure health expectancy in Italy on the basis of incidences. The aim of this study is to estimate the health transition probabilities for Italian elderly and to study how they are influenced by socio-relational variables, in order to calculate differences in health expectancy linked to socio-relational factors. The analysis is carried out using the Italian longitudinal module of the European survey on income and living condition: it silc 2004-2007; the study is limited to the population aged 55 and over in 2004 (4645 interviews) living in private households. Transition probabilities between chronic illness, disability and death are estimated using several methodologies. In particular, using a Markov model, we evaluate the impact of individual characteristics, as well as social and environmental variables, on health transition, controlling for initial condition and sample attrition.