I have done some testing on the method and it reveals some interesting results. There is also no need to distinguish between resources, drivers, activities etc. as the relationships are mathematically constructed.

There seems to be a trend – converging of various disciplines from Management Accounting, Econometrics, Business Intelligence, Predictive Analytics towards real life predictive cause-effect models.

Gary did predict this with his concept of a Chief Performance Officer.

Shouldn’t we be recognising that the composition of a unit price is not just Fixed + Variable…. and that there can sometimes be a large grey patch in between. “Variable” is not always (rarely) straight-line, and “Fixed” can be a function of time (we can’t get rid of the “fixed” costs immediately, but can reduce it over time… ) ?

I think Goldratt is being unfair in his criticism of management accounting and, to me, as I said throughput accounting seems very much like marginal costing. I think the waters become muddied at the point where short term becomes long term, and all costs become variable.

My argument is that yes the product demand consumes the resource of, in this case, direct labour but given the contradictions highlighted in my hypothesis isn’t it, at best, misleading to attempt an allocation of costs per unit? Wouldn’t those costs be more appropriately entered into FG stock as period conversion costs?

Sorry. I am not buying into your logic. You will not get me take “the leap of faith” you reference. Labor payroll expenses and their capacity may “belong” to operations to “supply” to work, but it is product demand and mix that causes and consumes that capacity.

Resource capacity is adjustable, especially different types of skill sets. Longer-term product mix demand changes can mean you may need more of some and less of another. TOC is OK for short-term prodction mix optimization, but beyond that needs marginal expense analysis, etc.

Also, Goldratt’s TOC “throughput accounting” the way he defined it requires the very special condition of a physical bottleneck 24 x7 x 365 days. Without that in place, then good old marginal / incremental expense analysis applies. (But few organizations do that well.)

After doing several ABC Studies or Processes based cost studies I was amazed at how many CEO’s and C level executives did not know what was going on within these activities or processes to get their products to the final consumer. These studies where the pre-driver styles in order to get to the drivers of the cost (cause and effect style). They seem to rely on spreadsheets, graphs and ROI presentations. Yet, when faced with possible OSHA violations within the processes, no one wanted to change and look at what was going on.

Maybe with the show (Undercover Boss) more Chief Executives will get the message that it is not ROI but the knowledge of what it actually cost to get that product (or service) to the final consumer.

Despite his seeming downer on management accounting (after all Goldratt’s throughput accounting is not very different to marginal or relevant costing) he is right about the arbitrary allocations to arrive at unit cost.

Consider two plants making the same item, material cost per unit is £100, conversion costs for the period are £1000.

Plant A produces 1 unit Plant B produces an infinite number of units.

Costs per unit will be:

Plant 1 Material cost + conversion cost/unit = £1100

Plant 2 Material cost + conversion cost/unit = £100

I think that for processed items FG Stock should be the sum of material costs per unit x number of units plus conversion costs for the period (identified seperately).

Attempting to compute total costs per unit has no value as far as I can see, but it could lead to invalid pricing decisions (especially where cost + pricing is used) or invalid make/buy decisions.

To take this a step further one might argue that variable costs at least should be included in the cost/unit, not so. Let us consider direct labour, this should be fairly straightforward to assign to products. Total direct labour/total output = direct labour cost per unit. But:

What happens if I improve the efficiency of my workforce? Perhaps we have a culture of continuous improvement, then over time the labour required to manufacture x number of items decreases. This then is not a fundamental property of the product, “direct” labour is a property of the process, and once you make that leap of faith you realise that there are no truly variable costs because everything can change irrespective of the product or quantity being manufactured.

I believe one of the problems that the Theory of Constraints (Eli Goldratt and the book “The Goal”) have with ABC is their assumption is “We own the capacity which is fixed (unadjustable), so let’s fill it.” Hence their simplified fixation with “Any product cost is meaningless, and there is no calculated product cost. Only purchased material expenses are variable.”

That resource capacity is adjustable (temps vs. full-time employees, lease vs. buy) is a separate discussion for another time.

What I like about Mike’s point is the focus on filling freed up and available capacity that results from productivity improvements. This begs the question, “Who owns idle capacity? Operations or Sales?” I believe it is the latter. Operations’ goal is create more with improvements. The basic options are either fill unused capacity with sales (to make more money) or remove it (layoffs?). The Sales function should be involved and understand unused capacity management.

It is what we do with the increased capacity that counts. If we do nothing then results will be pretty much the same. If we have a strategy for utilising the resources we have freed up, then results will improve.

For me one of the major benefits of ABC is actually analysing what it is we do. From there it’s a small step to ask why do we do it? And could we do it better? Do we need to do it?

Once people start thinking about the processes they are involved in then great things can happen.

I believe the challenge is with respect to the initial investment, since you can’t get that sort of analytics without the toolset in place. At the beginning I think it is necessary to prove that you are providing the information as cost effectively as possible, and as Gary points out, have business decision makers who see the potential business benefit and are ready to make decisions on that basis.