Reestimating Osha's Effects: Have the Data Changed?*

Our analysis replicates earlier studies of OSHA's inspection effects to see if, in the face of recordkeeping changes and new inspection targeting procedures, the relationships previously found still hold. The sizes and patterns of coefficients that we obtain in our analyses, including the lack of evidence that OSHA inspections (or the threat of them) reduce the injury rate, are consistent with those found earlier. We also analyze whether the new inspection procedure resulted in the reduction of reported injury rates among both inspected and uninspected plants. Consistent with an earlier paper, we find some downward bias in reporting among uninspected firms affected by the records-check procedure; however, we do not find this reporting bias among inspected plants.