Measuring labor productivity in mining

Abstract

Labor productivity in mining is often measured by output per mining company employee over a given time period such as a year, where output typically is the content of the main metal product contained in the ore. The necessary data for this measure are readily available, an important advantage, but it does suffer from a number of known shortcomings. In particular, it fails to account for changes in the average number of hours mining company employees work annually, the growing use of outsourcing and contract employees, fluctuations in the quantities of byproducts produced, and changes in the extent to which ore is processed domestically. This study assesses the importance of these shortcomings by examining labor productivity in the Chilean copper industry over the 1978–1997 period. It finds the readily available measure of output per company employee closely corresponds with a measure of labor productivity that corrects for the noted deficiencies. This means the almost four‐fold increase in labor productivity over this period in Chile is real, and not simply the result of the rise in outsourcing and the tendency to export more copper as concentrate rather than blister or refined metal. The errors introduced by these developments are large, but they are persistently offset by the error introduced by the decline over time in the average number of hours worked annually per company employee. This fortuitous offsetting of substantial errors, however, cautions against generalizing the findings and assuming output per company employee is a reliable measure of labor productivity for other commodities, countries, or time periods.