Excerpt
Table of contents
1. Introduction
2. Main arguments of Johnson and Kaplan in “Relevance Lost”
3. Subsequent treatment of the topic in academic publications
A. The first enhancements
B. The upcoming of activity-based management (ABM)
C. The development of activity-based budgeting (ABB)
D. Recent trend: time-driven ABC
E. Criticism
4. Impact of activity-based costing on companies
5. Conclusion
6. References
1. Introduction
Every accounting student of the past sixty years has learned about inventory costinga bookkeeping procedure that manufacturing accountants follow to separate the production expense of an accounting period from the cost of manufactured product inventories at the end of the period.
(Johnson and Kaplan, 1991, p. 130)
This technique of valuing inventory should, although often practiced, not be used for managerial decision making though. It oversimplifies the consumption of overhead costs by products, services and customers and therefore leads to distorted cost information.
Activity-based costing (ABC), developed by single manufacturing firms in the early 1980s, seems to provide more reliable information. The second part of this work describes the concept of ABC by summarizing the arguments of two pioneers in this field. In their book “Relevance Lost: The Rise and Fall of Management Accounting”, first published in 1987, H. Thomas Johnson and Robert S. Kaplan (1991) examine the traditions of management accountting and describe possible improvements. In part three the developments of ABC in the last 20 years are described by reviewing a choice of important literature. Part four then shows the impact that ABC had on implementing companies. The conclusion, part five, contains an assessment of the used literature and an evaluation of whether the critic of traditional management accounting has been overcome by ABC.
2. Main arguments of Johnson and Kaplan in “Relevance Lost”
Johnson and Kaplan argue that a good cost system has to fulfil four functions (1991, p.228, citing Dearden 1967):
i. providing information for the preparation of financial statements
ii. allow managers to control the processes within a company
iii. calculate longand short-term product costs
iv. produce data that can be used for special studies[1]
Reviewing business practice and the management accounting research of the 19th and 20th century the authors say that the design of cost systems in the late 1980’s first and foremost leads to the accomplishment of the first function. This is all the more astonishing since there have been cost systems at the beginning of the 20th century suiting the demands for information of process control and product cost calculation very well.[2] Reasons for the disappearance of those systems, according to the authors, are an insufficient cost-benefit-relationship at this particular time, the rising of financial reporting and the missing scrutiny of academic researches. (Johnson & Kaplan, 1991, pp. 125-145)
The main problem with the current situation, so the authors say, is that the companies are using a method for decision making that once was put in place by the first auditors to separate occurred cost between sales and inventory. This method uses a two stage process:
- First all overhead costs are allocated to a few cost centres (e.g. departments).
- Then they are assigned to products using one recovery base: direct labour
Direct labour, however, is declining and often only counts for ten percent of total costs. In addition overhead costs grow steadily in most companies. Basing managerial decision on absorption costing therefore leads to wrong conclusions. Especially in a firm with a large range of products and a high proportion of overheads, products with a relatively high amount of direct labour appear too expensive. That is because other products that require less direct labour (but eventually therefore more set-ups or quality control) shift their costs to the former group. (Johnson & Kaplan, 1991)
A better way to set up a cost system for decision making, according to the authors, is to figure out what actually drives the overhead costs. They therefore describe a changed two stage allocation system:
- The first step again is to specify cost centres and assign the overhead costs to them. But in distinction to traditional systems the cost centres are characterized by having one special driver rather than an organisational function.
- The second step is to determine the cost drivers that really cause changes in the cost centres and base the overhead allocation on them (e.g. numbers of orders in receiving department). (Johnson & Kaplan, 1991) However, Johnson and Kaplan (1991, pp. 229-240) argue that there are a few differences depending on the purpose of the cost system:
a. Control of processes
For an effective process control system it is necessary to consider the time span in which the process gets done. The authors illustrate that as follows:
“In departments producing many items per day, the per-unit material, labor, machine time, and utility consumption could be controllable daily, hourly, or even more frequently. (…) For departments such as a research laboratory (…) hourly or daily cost control is nonsensical.” (Johnson & Kaplan, 1991, p. 229- 230)
Furthermore it is advisable to include only costs that occur within the cost centre and only those which are reliably traceable and measurable.
b. Calculation of short-term product costs
Determining short-term product costs the non-traceable costs have to be included. It is also important to measure the use of scarce resources, so the authors. Most product related decisions, however, have to be made based on long-term product costs.
c. Calculate of long-term product costs
When it comes to the calculation of long-term product costs it is helpful to understand all costs as variable since they all are due to managerial decisions. The product costs then can be computed as follows[3]:
- measure or estimate the use of cost drivers by every product
- sum up the figures to get the total number of demands for every cost driver
- divide the total costs of the cost centre by the number of drivers (i.e. costs per driver)
- measure or estimate the demand for all different cost drivers of a product
- multiply the different costs per driver and the corresponding number of cost drivers to get the product costs of the factory
[...]
[1] Johnson and Kaplan (1991, p. 228) argue that those special studies nearly always require special data. Building a standardised cost system on those situational requirements does not seem appropriate. The authors therefore do not go into further detail in this respect.
[2] Johnson and Kaplan (1991, p. 222) state, however, that the comparability is limited as less overhead costs, more direct labour and only little product diversity occurred at the beginning of the 20th century.
[3] Johnson and Kaplan make a differentiation. They compute the factory product cost first and then add up the results of a similar calculation of non-factory costs (Johnson & Kaplan, 1991, p. 244). For explaining the general nature of the technique this differentiation can be left out.