EVALUATING THE RELATIONSHIP BETWEEN INCOME, SURVIVAL AND LOSS OF AUTONOMY

We use both subjective and objective measures of the probability to survive to age 85, of the probability to have activities of daily living (ADL) limitations, and of the probability of entering a nursing home. [...] 4 Findings 4.1 Descriptive evidence Before moving to the econometric analyses, we present in this section correlations between our variables of interest, that is between income and the objective and subjective survival proba- bilities, and between income and the probabilities to become dependent (either estimated by the number of ADL limitations or by the probability to enter a nursing home). [...] Yet, the relationship between income and the objective probability of survival to 85 appears stronger than the relationship between income and the subjective probability (the regression line for the objective probability is steeper than the regression line for the subjective probability). [...] 4.2.4 Estimation of the probabilities of moving to a nursing home Table 4 (Panel C) confirms the negative relationship between the objective probability of being in a nursing home and income, and the positive relationship between the subjective probability and income, which we had found in Section 4.1.3. [...] The point estimates are larger when explaining the probability of survival (0.1007 compared to 0.0461 and 0.1585 compared to 0.0437 for columns 1 and 2, respectively, and similarly for columns 3 and 4), but the signs (positive) and the statistical significance (significant for the objective probabilities, not significant for the subjective ones) are the same.