The Application of Statistical Methods to the Control of Industrial Costs

Abstract

Quality control studies the variation from standard of tangible articles, whereas “cost control” studies the variation from standard of the workers' efficiency, and consequent variation in production cost. In essence, the system consists in setting standard times, and thus standard costs, for all jobs and for the elements making up the complete jobs. Percentage divergencies from those standards are recorded at regular intervals on charts, a separate chart being used for each class of work, both as regards time and cost. Control lines, calculated from previous results over a period, upon the same mathematical principles as for quality control, are inserted on the charts. When a reading falls outside the control line, action must be taken to ascertain the cause. However, unlike quality control in which articles covered by abnormal readings can be scrapped, abnormal time or cost must still figure in the final examination of production cost. A sufficiently fine subdivision of the elements of a job will prevent an occasional abnormal reading on one portion from giving abnormal reading for the complete job. The individual readings are combined to form a grand total chart; careful watching of the detail charts enables the grand total chart to be maintained in control and it is from the latter chart that the “cost index” (percentage divergence from standard of the cost of the complete product) is obtained, which enables the latest actual cost at a particular moment to be determined directly. A continuous check is thus kept on the production cost of all the principal details; any abnormal fluctuations of cost become quickly apparent, and can be readily traced back to their cause. Concurrently, a simple method is provided for obtaining the total actual cost of the complete product at any time.