Working Paper: NBER ID: w1418
Authors: Henry Saffer
Abstract: This paper presents another extension of the approach initiated by Brown. As in Brown's work, the wage change specification is used to control for bias due to omitted ability data. Then, as in Duncan and Hoimlund's study, working conditions are measured using subjective self?reported data. However, in this paper, working conditions are measured by a single comprehensive variable. This approach eliminates omitted working conditions as a source of bias. The working conditions measure is then treated as an unobserved variable which limits measurement error to an unknown scale factor. The model is estimated using a technique derived by memiya (1978).
Keywords: compensating differences; working conditions; wages
JEL Codes: No JEL codes provided
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| working conditions (J81) | income (E25) |
| income (E25) | working conditions (J81) |
| working conditions (J81) | earnings (J31) |
| changes in income (E25) | changes in working conditions (J81) |