Educational differences in the effect of health status on personal income and its different components (all 16–64 years, excluding students, self-employed, unpaid work; 11 countries)
| Personal income (gross, per month) | Labour force participation (% of population 16–64 y) | Number of hours worked/week | Hourly income | |
| Summary measure of effect with 95% CI* | ||||
| Low educated | 1.280 (1.250 to 1.300) | 1.046 (1.041 to 1.051) | 1.019 (1.012 to 1.025) | 1.041 (1.031 to 1.052) |
| Mid educated | 1.140 (1.120 to 1.160) | 1.030 (1.024 to 1.037) | 1.007 (1.000 to 1.013) | 1.025 (1.014 to 1.037) |
| High educated | 1.140 (1.120 to 1.170) | 1.023 (1.014 to 1.031) | 1.009 (1.001 to 1.018) | 1.030 (1.014 to 1.046) |
| Test (p value) of interaction† | <0.0001 | <0.0001 | 0.0011 | <0.0001 |
↵* To be interpreted as the relative increase in the economic outcome per one unit increase in general health. Estimated with log-linear regression in which the dependent variable is health status measured at interval scale from 1 (for 'very bad') to 5 (for 'very good'). With control for age*gender, co-habitational status and country.
↵† Interaction between education (as 3-level nominal variable) and health (measured as the interval variable). Estimated with a log-linear regression model with control for age*gender, cohabitation status and country.