"Results of studies that examine the relation between trade union presence and injury rates are often indeterminate and trade unions are sometimes apparently associated with danger not safety. The British WIRS data set has provided a unique resource whereby researchers may examine the relation between trade unions and injury rates as mediated through particular arrangements for health and safety. Yet here, too, most investigations have failed to find a negative relation. It is in this context that this article returns to the original data. Utilizing improved statistical techniques, it concludes that cases where trade unions have an input into health and safety committees and where there are representatives are to be preferred to those where there is no such trade union input or no representatives. It argues that considerable strengthening of regulatory provision is required on employee representation and consultation if health and safety is to be improved."

Support: Digital