Late Night Discussions
Number 12

by Dr. Eliyahu M. Goldratt

"Alex, it's been a long time since I've seen you so grumpy. Any particular reason?"

"Oh, it's just the same old story. I wasted all afternoon at corporate, fighting an attempt to trim one of my product lines. It's the fourth time this year this has happened. Some whiz kid does a cost analysis, and finds out that on one of our product lines our average selling price is lower than what he calculates to be our 'product cost'. He immediately jumps to the conclusion that that product line must be trimmed. The fact that if it is trimmed the company will swing from profit to loss, since the resulting reduction in operating expenses is almost zilch compared to the lost throughput, doesn't even cross his mind. I don't have anything against these whiz kids - they're just doing what they learned at school. The thing that really drives me crazy is the quarrels with top management - from them I do expect focusing on the bottom line."

"Did you succeed in persuading them not to trim the product?"

"Yes, I did. But frankly, I think that what really persuaded them was not my arguments, but the fact that they're reluctant to mess around with my division because it's so profitable."

Jonah smiles at me and says, "Why don't we devote the evening to finding out the reason for your constant fights? I think that it's about time that we conducted a TOC analysis on this subject."

"Good idea! I'd like to understand how come that even when I'm dealing with prudent, experienced managers, so often I find that cost accounting considerations obscure business decisions."

"Fine," says Jonah. "Let's take it as the first effect on our list of undesirable effects; managers' decisions are obscured by cost accounting considerations. What else would you suggest we add to the list?"

"The fact that the concept of 'product cost' is widely accepted," I say without hesitation.

"I don't have any argument with this statement, but just out of curiosity, why do you relate to this effect as undesirable?"

"For me," I start to explain, "the cost of running the operation is real money; we pay money to the workers, we pay money to engineers, we pay money to vendors; we never pay money to a product. 'Product cost' exists only as the result of some calculations. Nevertheless, my peers sometimes behave as if 'product cost' was more real then the actual money that we pay."

Smiling, he asks "What else?"

"The concept of 'product profit' is also widely accepted," I say, and without waiting for Jonah's question, I explain, "Once again, the same phenomena. A product doesn't have a bank account; it's the company that makes the profit, not the product. 'Product profit' is a mathematical entity. Take what happened today as an example, in the name of insufficient 'product profit' they were willing to sacrifice the company's profitability."

"Shall we go to the next step and try to connect the effects?" Jonah asks. "What do we have? The fact that the concept of 'product cost' and the concept of 'product price' are widely accepted, and the fact that these concepts obliterate business decisions. What we've seen so many times in the past is that if people are accepting a damaging concept it must be that this concept was once very valid, and it only became damaging when people continued to use it after circumstances had changed."

"Right," I say. "So if we want to understand why 'product cost' and 'product profit' concepts were widely accepted in the first place we have to discuss the industrial environment at the time they were invented. If I'm not mistaken, it happened around the beginning of this century."

"Yes," Jonah says. "At that time the vast majority of industry expenses were for raw materials and labor; these two categories accounted for more than 90% of the cost. I think the major factor that's important for our discussion is the fact that the common method to compensate workers in industry was piece payment-- one piece less produced, less money was paid to the worker, one piece more and more money was paid. Compare it to today, when we pay workers by the hour, we pay Social Security, and pension - one piece more or less produced, and the same amount of money is paid to the workers."

"Why do you stress this point?" I ask.

"Just so that we recognize that at the beginning of the century the vast majority of the costs were totally variable. You see, I don't think that differentiating between variable and fixed costs is useful. At the end this difference is based on arbitrary decision. Even the building is not fixed cost, it can be sold. "

"Jonah, you've confused me. I don't understand what you mean."

"Look at industry in the past ten years," he says. "Overall, what we call variable costs have stayed the same, while during the same time, fixed costs have doubled. So variable is fixed, and fixed is changing."

I laugh loudly. "I see your point, why we shouldn't use variable and fixed, but what do you mean by totally variable?"

"I call costs totally variable if the smallest change in operation causes a change in the money we have to pay. The smallest change is a decision to produce and sell one more or one less unit. This is a much more objective criteria. For example, raw materials are usually totally variable - if you decide to produce one unit more, you have to buy more material from the vendor. Workers salaries were at the beginning of the century, totally variable. Today they are non-totally variable. I think that rather than using the categories of variable and fixed, a much more useful classification is totally variable and non-totally variable."

"So, at the beginning of the century, over 90% of the costs were totally variable. I still don't understand why it's so important for our discussion."

"Only because," he answers, "totally variable costs can be accurately allocated to the individual products, and thus, at the beginning of the century, allocating company expenses to products, what we call today 'product cost', was a very reasonable approximation."

"I see. It is much easier to accept a very reasonable approximation than to accept the convoluted calculations that are required today to determine 'product cost'. And what about 'product profit'?"

He thinks for a minute and then answers, "Sales dollars, like expenses at the beginning of the century, can be easily, and in most cases accurately, allocated to products. Since net profit is total sales minus total expenses, then at that time net profit of a company could be reasonably approximated by 'breaking it' into components of net profit of the individual products."

"Or in other words," I continue, "at the beginning of the century, 'product profit' was also a very reasonable approximation. But Jonah, what's the use in doing all these mental exercises? Even if 'product cost' and 'product profit' were very reasonable approximations, who needs these accounting phantoms?"

"Good question," he answers. "The only plausible reason I can think of is that these mental exercises, as you called them, helped solve a major problem that managers at that time were facing. I wonder what it was?"

For a few minutes, neither of us speak. Then it dawns on me. The problem was not facing managers just at the beginning of the century. It's the same problem facing managers today. I think this is my biggest problem. Loudly I say "How does a manager determine the impact of a local action or a local decision on the global objective of the company? At the end, all of our actions and the vast majority of our decisions are of a local nature, but what we are trying to achieve by all of them is is the objective of the company. How do we connect between the local and the global?

"Excellent," Jonah jumps to his feet and starts to pace the room. "Now it all starts to make sense. In the beginning of the century allocating expenses and profits to products enabled management to logically dissect the company into product-by-product slices. This certainly made their life much, much easier. It is much more convenient to consider one product, than to take into account all the intricacies of the company. This slicing by product provided management with a straightforward way to connect their actions and decisions to the global objective of the company."

I think about it some more. He is absolutely right. Slowly I say, "So, at the beginning of the century, allocating costs was a very powerful concept. It enabled managers to manage."

"Yes," says Jonah. "But concepts are not enough. In order to fully utilize a concept, procedures have to be developed. But since the concepts were so powerful it didn't take long before the procedures to calculate product cost and product margin - what we today call cost accounting - were fully developed. But since the concepts were so powerful it didn't take long before the procedures to calculate product cost and product margin - what we today call cost accounting - were fully developed."

"Now I see. If 'product cost' and 'product margin' were very reasonable approximations and at the same time they enabled managers to overcome one of their biggest problems, no wonder they became widely accepted. Now what we have to do is to connect to our first undesirable effect. Why is it that today they obliterate business decisions? I think that I can see the connection, but as we learned from experience, I'm sure that it will behoove us to connect systematically, it will clarify in what way business decisions are obliterated."

"Why not," he says. "Let's devote the next discussion to that."

This "Late Night Discussion" is Copyright © 1992 Dr. Eliyahu M. Goldratt