Table 6—
Correlation of Husband’s and Wife’s Innate Permanent Health
| (1) |
(2) |
(3) |
||||
|---|---|---|---|---|---|---|
| Permanent health (ηi) from: | Fixed Effects | System-GMM | Fixed Effects | System-GMM | Fixed Effects | System-GMM |
| Self-reported health | 0.374 | 0.302 | 0.347 | 0.295 | 0.318 | 0.279 |
| (0.019) | (0.025) | (0.020) | (0.025) | (0.019) | (0.025) | |
| Chronic conditions | 0.221 | 0.206 | 0.191 | |||
| (0.022) | (0.022) | (0.022) | ||||
| College and race | No | Yes | Yes | |||
| Permanent income | No | No | Yes | |||
Note: The table reports conditional correlation coefficients between husband and wife’s estimated innate permanent health (ηi from Equations (1) and (3)). The first row corresponds to the baseline regression,in which the self-reported measure of health is used in the regression. The second row is computed using the number of chronic conditions as a dependent variable, as in Figure 8. To control for college and race, we introduce dummies for individuals and spouses having a college degree and being black, as well as the corresponding interactions. For the permanent income, we include husband’s, wife’s, and interacted individual fixed effects obtained from a regression of taxable individual income on years of education, age, age squared, marriage, and year dummies. All correlations are estimated from the PSID. Weights are used in the estimation. Bootstrapped standard errors in parenthesis.