Behavior vs. Sorting into Jobs
| Sample: | All | All | All | New Hires |
|---|---|---|---|---|
| Outcome: Sickness Absence | Present | Present | Present | Pre-Hire |
| Mechanism: | Baseline | Behavior | Behavior | Selection |
| (1) | (2) | (3) | (4) | |
| Low substitutability | −0.0131*** (0.0006) | −0.0058*** (0.0007) | −0.0066*** (0.0010) | −0.0043*** (0.0014) |
| Number of observations | 5,863,497 | 5,863,497 | 5,863,497 | 387,901 |
| Mean of dependent variable | 0.109 | 0.109 | 0.109 | 0.116 |
| Background controls | Yes | Yes | Yes | Yes |
| Year fixed effects | Yes | Yes | Yes | Yes |
| Occupation fixed effects | Yes | Yes | No | Yes |
| Establishment fixed effects | Yes | Yes | No | Yes |
| Worker fixed effect | No | Yes | No | No |
| Worker × job fixed effects | No | No | Yes | No |
Notes: The standard errors are clustered on the establishment level in Columns 1 and 3 and on the worker level in Column 2. Job is defined as the combination of occupation and establishment. Low substitutability is defined as having zero to five substitutes (the reference category is employees with more than five substitutes). The background controls are gender, age, education, birth country, having children aged 0–3, and establishment size. In Column 4 we also control for the pre-hire employment status of the new hire. *p < 0.10, **p < 0.05,
↵*** p < 0.01.