Book Description

Aggregate productivity statistics succinctly and conveniently measure the efficiency with which resources are being used in a country or industry, but problems of measurement and aggregation in the inevitable presence of heterogeneity require that these statistics be interpreted carefully to avoid misleading results. This paper exploits an unusual database to explore the differences between productivity trends as they appear at the aggregate level and as they may be experienced at the firm level. The Mine Safety and Health Administration (MSHA), as part of its mandated regulatory effort, has collected labor input and coal output information for every mine in the United States since 1972, along with data on site locations, operator identity, and mining techniques (U.S. Department of Labor, MSHA, "Part 50 Coal Mining Address/Employment and Accident/Injury Files"; henceforth USDOL, MSHA, Part 50). Thus, labor productivity can be observed for this industry at the lowest practicable level; and, based on this microdata, a national aggregate, as well as any number of sub-aggregates, can be formed from the bottom up. Working from microdata all the way up to the aggregate industry level not only supplements industry aggregate statistics but also permits an examination of the root causes of aggregate productivity change with greater clarity than is usually the case.