This post is target at people who know Throughput Accounting well and are ready for an intellectual exercise. I think that the title Throughput Economics fits better than Throughput Accounting in describing a methodology that supports superior decisions, but has very little relevancy to accounting.
The usual setting is that organizations buy materials or goods they eventually intend to sell, then they use the capacity of internal resources and then they sell.
A key difference, important for making decisions using the Throughput Economics rules, is that the cost of materials behaves in an almost linear way. When 20% more materials are needed then the additional cost of materials is about 20% of the original cost. Transportation and other factors might add some costs, but for being “about right” treating the cost of materials as linear is good enough.
This linearity is quite important for simple, yet effective, calculation of the impact of increased sales resulting from promotions, export or special deals.
For instance, suppose 1,000 items of Product-X are sold for $10 and the cost of material per unit of X is $6. The resulting total T is $4K. If, there is a demand from a large client for 3,000 additional X for only $8 per unit, and that demand does not affect the regular demand, then the additional delta(T) is 3,000*($8-$6) = $6K. Of course, this simple calculation is not enough to fully justify the decision. At the very least there is a need to validate that the available capacity is enough for the total of 4,000 units.
The cost of capacity behaves in a very different way than the cost of materials. Suppose the most loaded resource has enough capacity for up to 3,000 units of X. This means that by increasing the required quantity of X from 1,000 to 3,000 – there is no extra cost of capacity.
However, there is a need to consider the cost of capacity for processing the extra 1,000 units.
Do we know the cost? We cannot automatically assume it is possible to produce the additional 1,000 units. Sometimes there is no valid way to get the extra capacity needed for such extra demand. In such a case the only option is to free capacity by not producing something else. For instance, by preferring to sell 3,000 units to the client that pays less per unit, but generates overall higher T than the regular market. In other cases it is possible to use overtime or outsourcing, but then it is mandatory to calculate the extra cost, delta(OE), required for the extra capacity that is beyond the regular available capacity. Thus, the cost of capacity behaves in a non-linear way, while the cost of materials is linear.
What if the organization owns part of the materials?
This is a typical case with manufacturers based on agriculture, like meat, wine, juice or other types of food processing when those organizations also own farms that raise the key raw materials. Usually those manufacturers are able to buy more quantity of those materials from other suppliers.
The farms plan, ahead of time, the amount of materials required for manufacturing. However, the lead-time of agriculture cannot be shortened. Thus, the initial decision on the quantity of every specific item cannot be altered within the given lead-time. In most cases the frequency of such decisions is highly limited, mostly only once a year.
There are two ways to model such a case:
- Define two different alternative products. The primary products use the self-owned materials. These products have high T per unit because the cost of those materials is not included in the TVC. However the primary products are limited by the available amount of materials. When the demand is much higher than that limit, the alternative products, actually exactly the same from the client perspective, can be produced and offered to the market, but with higher TVC, hence lower T per unit.
- Treating the owned materials as if they are resources!
The idea is that the resources behave anyway in a similar way to those materials. There is cost to maintain a certain amount of available capacity. The cost of utilizing the resource for 20%, 60% or 99% of its availability is the same. But, when there is a need for additional capacity the cost jumps.
This basic non-linearity of the cost of resources forces Throughput Economics to use an algorithm that is able to consider such behavior in order to support better decisions. In itself this ability is still pretty simple – just noting when a resource starts to be overloaded. Modeling certain materials as resources does not add any new difficulty.
The initial available quantity of materials, owned by the organization, carries a fixed cost no matter whether all the quantity is used or not. So, modeling such materials as resources, with the appropriate level of available quantity and a cost figure for adding materials above that level, is in line with the Throughput Economics algorithm for considering the resources capacity and temporary elevation of capacity.
The one difference in the Throughput Economics methodology is that purchasing more materials from suppliers should be treated as delta(OE), rather than part of delta(T). The key supporting information for a decision is delta(T) minus delta(OE) and that result is not impacted.
My general insight:
We like linear behavior because it is simple and in most cases good enough for supporting decisions. We are all aware that reality does not truly behave in a linear fashion. There is a common temptation to try to find more precise solutions by considering the complexity. Falling in this trap yields confusion and far inferior level of performance.
The insight is that even when we face true non-linear behavior, it is still relatively simple behavior, where all we have to do is to consider the sudden jumps at certain values.
Eventually we are able to reasonably predict the impact of decisions and changes within a certain range of possible outcomes. What we get is much better than assuming complexity.
Eli Schragenheim 