In a rebuttal to John Seddon‘s latest paper, An Exploration into the Failure of Lean, Bob Emiliani asserts that the original purpose of TPS was to reduce cost. He quotes both Taiichi Ohno and Yasuhiro Monden saying so, and chides Seddon for not reading their works carefully enough. In the context of these documents, however, I think the quotes are misleading. Neither Ohno’s and Monden’s books, nor any other Japanese publication about manufacturing systems that I have seen, contain a discussion of what costs actually are.
“Cost” is assumed to be self-explanatory, as when we discuss the cost of a bottle of milk or a subway ticket. In manufacturing, with resources being shared over time among multiple uses, assigning a cost to any object, decision, or event, is anything but obvious, and is the object of spirited discussion in the world of cost accounting. You would reasonably expect the explanation of a system developed for the purpose of reducing costs to start with an explanation of what it means.
With TPS, it doesn’t. In Workplace Management, on p. 19, Ohno says “Costs do not exist to be calculated. Costs exist to be reduced.” On the face of it, it is like saying that, to lose weight, you don’t need to step on a scale. But there is no argument as to what weight is. In manufacturing, costs are whatever the accountants say they are, based on a model that takes into consideration some of the parameters of the activity, and applies rules for the depreciation of assets of time as well as their allocation among products. How well these models map to economic reality is a question that most managers don’t ask but that has been raised by academics like Robert Kaplan, and consultants like Eli Goldratt, Brian Maskell, or Orrie Fiume.
Right after saying that the purpose of the system is to reduce costs, Ohno and Monden treat the subject of costs like a can of worms that is best left unopened, and move on to topics like increasing productivity and improving quality, with the implication that it will drive costs down, regardless of what they mean and how they are calculated. They certainly do not advocate any of the actions that, everywhere else, are understood under the label of “cost reduction,” such as cutting every department’s budget by 5%, reducing employee benefits, brow-beating suppliers into cutting prices, and outsourcing to cheap labor countries.
Whatever Ohno’s motivation may have been to declare cost reduction as the primary purpose of TPS, I don’t find it helpful in understanding the system or communicating about it. What, on the other hand, strikes me as essential is the notion that improvement consists in enhancing one or more dimensions of performance without making any other one worse, and that it is always possible. Increasing productivity does not mean cutting corners in quality, eliminating defects does not mean adding more inspectors, and order fulfillment lead times can be shortened without increasing inventory… It is the end of management whack-a-mole.