A problem with conventional output measures is that they do not capture such "outputs" as cleaner air, cleaner water, and a more accident-free work environment, which result from the installation of equipment that abates pollution or promotes health and safety on the job. Conventional input meas ures, on the other hand, do include inputs which perform such functions. In this study, we test different specifications of the relationship between manufac turing capacity and the manufacturing capital stock. Not surprisingly, the empirical results indicate that the drive to control pollution and provide a safer and healthier work environment has caused a decline in the elasticity of capacity with respect to total capital in manufacturing, because pollution abatement investment (PAI) and occupational safety and health investment (OSHI) generally do not augment conventionally-defined productivity. Ad ditionally, we discover that some indexes of manufacturing capacity are con structed by means of methodologies which assume an unchanging historical relationship between capacity and capital, and thus are unable to detect the declining capital elasticity of capacity.