Working Paper: NBER ID: w2813
Authors: Wayne B. Cray; John T. Scholz
Abstract: This study test for effects of OSHA enforcement, using data on injuries and OSHA inspections for 6,842 manufacturing plants between 1979 and 1985. We use measures of general deterrence (expected inspections at plants like this one) and specific deterrence (actual inspections at this plant). Both measures of deterrence are found to affect accidents, with a 10% increase in inspections with penalties predicted to reduce accidents by 2%. The existence of specific deterrence effects, the importance of lagged effects, the asymmetrical effects of probability and amount of penalty on accidents, and the tendency of injury rates to self-correct over a few years support a behavioral model of the firm's response to enforcement rather than the traditional expected penalty' model of deterrence theory.
Keywords: OSHA enforcement; workplace accidents; deterrence theory; behavioral model
JEL Codes: J28; K32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| expected inspections (L68) | accident rates (J28) |
| actual inspections (L85) | accident rates (J28) |
| enforcement actions (G18) | accident rates (J28) |
| penalty probability (C11) | accident rates (J28) |
| penalty amount (H26) | accident rates (J28) |