Full bibliography

This study examines the validity of injury statistics used to monitor workplace safety in the Canadian province of Alberta. These indicators were found to significantly under–report the rate of injury and to be vulnerable to gaming by both employers and the workers´compensation board. These threats to the validity of the measures should limit the inferences drawn from the measures. Injury–based statistics were also found to be inadequate proxies for the broader construct of workplace safety. The political feasibility of alternative measures is also discussed.