Activity-based decision

An integrated framework for activity-based decision

Mike Partridge, Lew Perren. Management Decision. London: 1998. Vol. 36, Iss. 9; pg. 580

Abstract (Summary)
Activity-based costing (ABC) data have the potential to inform a wide range of management decisions. Recently the term Activity-based management (ABM) has emerged to describe any application of ABC data, but there is currently no framework which pulls together the disparate strands of potential uses. Partridge and Perren have conducted a systematic review of internationally reported ABC applications to produce an integrated and comprehensive ABM framework for management decision making. Traditional methods of costing are critically evaluated as a precursor to explaining the ABC model. Partridge and Perren go on to illustrate their ABM framework with case examples.

Non-financial managers have a key role to play in the successful implementation of activity based systems (e.g. Friedman and Lyne, 1995). When deeply involved as members of the activity-based costing (ABC) project team, such managers have been identified as enthusiastic users of and propagandists for the new information systems. Friedman and Lyne (1995), in a study of activity-based techniques in 11 companies, found evidence of an improvement in the relationship between accountants and non-financial managers. Innes and Mitchell (1991), in a longitudinal study of an activity-based system in the Cummings Engines Company's Daventry plant, found that "managerial enthusiasm for ABC was extremely high and became manifest in requests for more and more information based on it".

These findings reinforce the authors' belief that activity-based techniques offer managers a user-friendly, understandable information system whose ownership they can share with accountants. Indeed it is probably essential that accountants take the (for them) revolutionary step of deliberately sharing the system with its users.

ABC data have the potential to underpin a wide range of decision-making processes. It is unfortunate that the design and development of ABC systems has often remained largely within the domain of accountants. This is perhaps understandable: Less justifiable is that the use of ABC systems has in most cases been focused on a narrow range of applications. Recently the term activity-based management (ABM) has emerged to describe any application of ABC data to management decisions. However, the research for this article found an absence of a framework which pulls together the disparate strands of potential uses.

Aim and method

Our goal was to eliminate the confusion which surrounds ABM. We conducted a systematic search of the management accounting literature over the past ten years in order to identify articles and texts which appeared to focus on ABM or associated areas. The sources consulted intentionally span academic and professional literature, as it is uncertain whether academics or professionals were the driving force behind the ABM concept. Each article or text was methodically analysed to discover applications of ABC data which the authors had identified as being a part of ABM. Each application was coded to allow traceability back to its original source (see Appendix). This allowed a flexible approach to textual analysis and categorisation, with the applications identified being iteratively grouped until patterns emerged. These patterns formed the foundation for the integrated framework we present in a later section.

It is our view that managers are insufficiently aware of the potential of activity-based techniques; accountants may also be possessive about the operation and use of information systems that have traditionally fallen within their domain. We are entering an era when collaborative, multi-discipline management is a pre-requisite for competitive advantage. Our belief is that given the will, and given the understanding, a new tool is available for use by managers.

Review of traditional and activity-based costing systems

This section provides an overview of costing systems and readers familiar with these concepts may wish to go straight to the next. The allocation of costs to products and other cost objects has occupied the minds of managers and accountants for more than a century. Whether it be to price stock, and hence to measure profit, to support pricing decisions or to provide factual underpinning forproduct/market decisions, the identification of product costs is a vital component in the process of management.

The accurate allocation of so-called direct costs (costs which are clearly linked with a specific unit of product) presents few problems. The treatment of indirect costs, overhead, provides a more contentious area.

Many firms use a blanket rate approach, where a convenient proxy for resource consumption is used to share overheads in a way which is superficially equitable. The "two-stage model" is a refinement of this approach. In the first stage overheads are allocated to cost centres using proxy measures for usage such as floor area or headcount. In the second stage a cost centre's overheads are shared between products using appropriate proxy measures (e.g. direct labour hours for labour intensive cost centres, machine running hours for mechanised ones).

These mechanisms, while they may produce adequate figures for stock valuation and so satisfy audit requirements, suffer from two defects if their output is to be used for decision making. First, the proxies used are not measures of resources consumed; they are simply a convenient way of allocating what appears to be a fair share of overhead to products. The result is a degree of averaging of unit product costs with all the inaccuracy that this may cause. Second, the cost rates generated are generally based on budgeted output volumes. Fluctuations in these volumes will mean that, because a proportion of all overhead is fixed in nature, the actual cost rates will be higher or lower than those originally calculated. This may lead managers to make inappropriate decisions.

ABC responds to the earlier criticisms by allowing activity based overheads to be allocated to activity based cost pools according to their genuine resource consumption. This is a great improvement over the previous methods although it still leaves a rump of overhead known as facility sustaining costs which are not susceptible to ABC analysis. We shall return to them in a moment. The key stages of ABC are shown as the shaded sections of our integrated framework (see Figure 1):

1 The division of the firm into activities (e.g. set up, procurement, quality etc.) (position 1 in Figure 1).

2 The allocation of overhead resources (position 2 in Figure 1) to activities excluding those indirect facility sustaining costs which are independent of activity levels.

3 The identification and quantification of activity drivers (position 3 in Figure 1) (e.g. Pounds per set up, Pounds per order, Pounds per inspection), which measure the use of an activity by cost objects (position 4 in Figure 1) (e.g. products, customers etc.).

4 The calculation of an activity driver rate for each activity.

5 The charging of an appropriate amount of each activity to each cost object (position 4 in Figure 1).

If total cost data are required, then facility sustaining costs can be allocated to cost objects using any suitable (though arbitrary) basis of the sort that was used in the traditional approaches.

ABC has the advantage of moving away from the simplistic fixed/variable cost model to one in which activity-based overheads may be allocated to cost objects in direct response to their consumption.

We have now described the vertical logic of the ABC system. We will next explain the horizontal axis of the model: from the left, cost drivers (position 5, Figure 1) are those factors such as economies of scale and use of technology which affect the cost efficiency with which activities can be performed; on the right, performance measurement (position 6, Figure 1) can be reflected by activity driver volumes and rates.

So the basic ABC model offers a new approach to product cost measurement, one which generates information that is credible, because its derivation is logical and visible, to both financial and non-financial managers, and which also offers a new visibility to resource management. What then are the uses of this new information? This is the point at which ABC extends into ABM and is the focus of the rest of the article.

An integrated framework for ABM

We have used the results from our meta-level review to extend the basic "cross" diagram (Raffish and Turney, 1991) into a framework which integrates ABC information with its possible applications (see Figure 1). We believe that this provides a coherent agenda for general managers who may wish to apply ABC derived information to a variety of decision making situations. We will now explain the components of the framework, which extends the basic "cross diagram" at its core.

This ABC system generates decision relevant information which may be used in three ways:

1to stimulate resource allocation decisions (position 1, Figure 1), i.e. changing the balance of resources allocated to activities in response to changing activity demand patterns and to changing activityefficiencies;

2 to provide cost object (position 4, Figure 1) information on which to base market interface decisions;

3 to generate performance measures (position 6, Figure 1) with which to monitor activity consumption and efficiency.

Early adopters of ABC saw its cost management role as the most valuable outcome of its implementation and the term activity-based cost management (ABCM) creeps into the literature from quite an early date. ABCM offers a new visibility to an area of hitherto near total mystery. For the first time, managers have a logical, quantified input: output model of overhead consumption in which actions to alter demand, efficiencies or waste can have a measurable effect on levels of resourcing.

Thereafter, uses of the core ABC model proliferated. The availability of believable metrics, based on causal logic, stimulated new applications and breathed fresh life into a range of established management practices.

Cost objects and related applications

If initially we pursue the vertical strand of the framework, then it can be seen that cost objects (position 4, Figure 1) sub-divide into products (position 7), channels (position 8) and customers (position 9). That is, it is technically possible separately to identify the costs of producing products, of using alternative distribution channels and of servicing individual customers. Cost object (position 4) data can be further divided to help with the following decision areas:

Stock valuation (position 10)

While the strategic relevance of stock valuation is debatable, ABC is certainly capable of performing this task. Indeed, Innes and Mitchell (1995) found in their survey of 251 of the UK's largest companies that 40 percent of manufacturing companies used ABC costs for this purpose.

Pricing decisions (position 11)

The determination of selling prices can rarely be effected without an awareness of unit costs, even if the decision becomes one of whether to trade in a particular market or not. There is considerable evidence to demonstrate that ABC unit costs can frequently reveal alarming profit variations between products as a result of prices developed from traditional costing systems. For example, Volkswagen Canada has applied ABC to its quotation pricing process (Gurowka, 1996). A detailed activity costing sheet highlights areas that increase product costs and areas where costs need to be cut. The costing sheet has found uses, both for pricing current products, and also pricing future business.

Product design (position 12)

ABC cost analysis can identify those activities which add to or create product value and those which are unnecessary, wasteful or used in too great a quantity. Turney (1992) suggests that the traditional approach of product design followed by process design can leave the process designers with a time-consuming and costly task. He supports concurrent engineering, where product and process design are performed in parallel, in full knowledge of the activities (and their costs) that will be required.

Transfer pricing (position 13)

There is as yet limited evidence of the use of ABC to generate transfer prices (Innes and Mitchell, 1995; Morrow and Ashworth, 1994) although the benefits of using a visible and equitable cost base must be considerable, apart from the avoidance of argument and negotiation. One can envisage internal customers for products or services being motivated to modify the nature of their consumption patterns and specifications. We are aware of one financial services company where a central data processing facility uses ABC cost measures to charge out transaction processing costs to branches.

Channel costs and channel decisions (position 14)

The original ABC model is capable of generating cost data for a variety of cost objects (position 4), most commonly products (position 7) and customers (position 9). In certain organisations it may be useful to generate costinformation for alternative distribution channels (position 8) (e.g. Booth, 1996) so as to assist management in making informed choices in this area. Cooper et al. (1992) "illustrate the case of a metal fabrication and distribution company whose ABC analysis revealed that the most profitable products were those shipped direct from the mill. Such shipments incurred almost no inventory and material handling costs". The findings caused management to question the competitive advantage of their distribution business.

Customer profitability analysis(position 15)

After product cost analysis, customer profitability analysis provides one of the most frequently cited ABM applications, partly, it could be assumed, because of the dramatic nature of the information that it tends to produce. Bellis-Jones (1989) first put forward the message that at least 20 percent, possibly 30 percent, of customers are actually eroding profits when the full costs of servicing them are accurately attributed. Sweeney and Mays (1997) reference the First Tennessee National Corporation, a regional US bank, where 30 percent of customers were found to be providing 88 percent of the company's profit, while another 30 percent generated a loss of 7 percent. The situation "was corrected through a combination of higher minimum bank balances, new products and process redesign".

Output decisions (position 16)

Decisions as to what products to sell, to which market segments or customers, at what prices and in what volumes, are fundamental to the formulation of competitive strategy. ABC offers medium-term cost data in which cost variability is more realistically measured and which consequently provide bases for product/mix/market and out-sourcing decisions. This is essentially the basis for activity-based budgeting.

Activity-based budgeting (ABB)(position 17)

Budgeting for overhead has long presented insoluble problems for management. Incremental budgeting, characterised as "last year plus 5 percent", remains a widely used mechanism. Zero-based budgeting (ZBB), in various manifestations and for various reasons, has never achieved widespread use. The ABC model, with its quantitative connection between outputs and resource inputs, can meet a long-felt need. This application of ABC methodology is in essence an example of output analysis (position 14), when one of the output variants is selected as the operating plan for the coming budget year.

Activities and activity-related applications

If we now return to activities (position 2) and activity analysis (position 3) in the upper part of Figure 1, there are a number of valuable spin-offs at this level.

Value-adding/non-value adding analysis (position 18)

Cited by most commentators, the analysis of activities into value-adding and non-value adding appears to be fundamental to the cost management capability of ABC analysis. Turney (1992) suggests that an activity has value if it is essential to the customer or it is essential to the functioning of the organisation. Hixon (1995) complements this by defining non-value adding activity as "anything that can be eliminated without detriment to the final product or service". Antos (1992), in an examination of ABM in an Alaskan oilfield water treatment facility, illustrates a number of NVA activities including drilling a hole incorrectly, engaging in litigation because of spills and violations, and dealing with accidents. The identification and cost analysis of NVA activities act as a trigger to stimulate their reduction or elimination.

Value chain analysis (position 19)

This is a particular form of activity analysis which derives directly from Michael Porter's (1985) value chain concept - the overriding requirement for value-creating product attributes to produce a surplus of income over the cost of the activities which generate them. Morrow and Ashworth (1994) summarise the concept as "a business is a series of linked activities which ultimately add value to the customer. Activity based management recognises this fact and therefore helps management to view the organisation by understanding the activities, their cost and how they link together to form a simple chain of value-creating activities for a business."

Organisational re-design (position 20)

Activity analysis will frequently reveal that current organisational structures do not reflect the way in which the organisation operates. Further, as Turney (1993) points out, "the artificial dividing of work into vertical functions creates communication barriers and results in excess cost and time and in poor quality". Hixon (1995) emphasises the point, suggesting that while departments may optimise individually, they may also compete, and there may be little or no cross-functional coordination.

Process re-engineering (position 21)

Business processes are sequentially related sets of activities. Hence activity analysis and activity cost measurement offer, whenwith benchmarking, a springboard for process re-engineering. As Evans and Ashworth (1995) assert "the recent drive to re-engineer business processes is often pursued with no regard to, or real understanding of, before-and after-process costs of the underlying product or the customer profitability realised from performing those processes ... There is often no real framework to monitor the return on investment...". Indeed , without ABC, there is often little to no measurement.

Benchmarking (position 22)

Having already identified that an early product of the ABC model was the generation of performance measures, it is a logical progression to use ABC metrics to conduct benchmarking studies. For example, Coburn et al. (1995) describe how ABC benchmarking was successfully performed for the accounting department at the Marketing ResourceGroup of US West, a publisher of telephone directories.

Continuous improvements (position 23)

Continuous improvement is implicitly an outcome of several of the foregoing paragraphs. Gurowka (1996) describes Volkswagen's process improvement tool called KVP (Kontinuierliche Verbesserungsprozess), where the cost data provided by ABC ensure that the process improvement team concentrate on what is important and so prioritizes its targets.

Cost modelling (position 24)

Cost modelling, cost forecasting, cost simulation are different expressions which mean broadly the same thing: an ability to build future orientated cost models of products, events and processes. Innes and Mitchell (1995) found 61 percent of their respondents used ABC to model costs. The most common applications were to model individual product costs and for decision situations such as capital investment evaluations.

Quality costing (position 25)

Glad and Becker (1996) argue that the costs of wasted resources (e.g. product defects, raw materials, capacity, labour input, energy) should be separately measured and excluded from product costs. They propose that the four classic quality costs (prevention, appraisal, internal and external failure) can and should be measured and reported. It is the authors' experience that external failure costs, when accurately measured under ABC analysis, can provide management with alarming insights. The management of un-used capacity relates closely to the measurement of waste. If the resource is under-used and is not "stockable", then under-utilization is waste. As Glad and Becker (1996) say: "Capacity costs should be charged to the cost objects based on a realistic or practical capacity. Surplus capacity costs should be reported as wastage.". Carolfi (1996) describes an ABC-based quality management system and illustrates this with a case study based on a market research organisation. The study revealed that 50 percent of the analysts' working week was dedicated to rework, and 1,546 annual hours of computer time were required as a result; total cost approximated to US$660,000. The net result of the study was a two thirds reduction in staff numbers.

ABM in action at Volkswagen in Canada

Volkswagen (VW) Canada's plant in Ontario provides an encouraging example of many features of our integrated framework (which we have cross referenced to Figure 1). As Jim Gurowka (1996), a former senior finance officer at VW Canada states, "ABM information will help VW achieve its goal of becoming the ultimate, cost conscious, world-class, customer-focused supplier". The plant manufactures aluminium wheels, catalytic converters and die cast engine parts which are sold to VW and third party customers throughout the world. From being a captive customer, the parent company was encouraged to source product from the cheapest producers, not just sister companies. This led to a 25-30 per cent price fall at VW Canada, as well as a need to find new third party customers to fill the gap left by declining parent company orders.

A pilot ABC project in 1991 revealed that, in the die casting areas, 80 per cent of products were either losing money or were only marginally profitable. Profitability was sustained by a couple of high earning parts, a classic Pareto effect which had been hidden by the old costing system. Management's attention at this time was focused on growth andnothing much was done about the new ABC information. The ABC project, without a champion, "went to sleep".

By 1993, however, massive changes in technology and competition caused the company to take drastic action to maintain profitability. ABC was re-activated and helped management to take tough decisions based on sound information. For example, ABC information was central to a decision to re-engineer the major business processes (position 21).

In evaluating what proved to become a highly successful ABC implementation, a number of factors were felt to be important. For example, the joint project leaders brought strong backgrounds in purchasing and production, as well as finance, to the task; in addition, they made a strong commitment to share the ABC results not just with managers but all employees throughout the plant.

While initial outcomes centred on product (position 7) and customer profitability (positions 9 and 15), activity-based applications gradually moved on to quotation pricing (position 11), budgeting (position 17), quality costing (position 25), process improvement initiatives (position 21), product (design) rationalisation (position 12) and resource utilization (position 1). ABC analysis is now an integral part of VW Canada's continuous improvement drive (position 23). As Gurowka says "process improvement (position 21) projects which lack costing information tend to be unfocused exercises which concentrate on perceived problems. The cost and economic data (position 24) provided by ABC ensure the process improvement team concentrates on what is important."

Discussion and implications

On the surface, many firms will appear to be using some of the applications from the framework. Closer examination may reveal that decision making is either without measurement or is using traditional costing mechanisms with the inherent limitations that we highlighted earlier in this article. Even when managers do have access to ABC data, evidence suggests (e.g. Drury et al., 1993; Innes and Mitchell, 1995) that they tend not to apply it comprehensively. Without measurement, management decision making has to be based on intuition, impulse and hunch. Traditional information is based on departments, functions and nominal codes. Modern management thinking has championed the importance of cross departmental initiatives, of processes rather than functions, of core competencies which transcend the confines of strategic business units. Such notions seem to be the way forward, but without sound quantitative activity-based information, we may see a reversion to intuitive decision making.

Our framework (Figure 1) brings together the disparate strands in the literature to provide a clear agenda for general managers wishing to integrate ABC generated information into their decision making. For managers who already have access to activity-based information it provides a reference point against which to compare their current applications. We suggest such managers conduct a systematic audit of their current decision-making practices against our framework. For most organisations this will identify areas of ABC application which are missing (Innes and Mitchell, 1995). We suggest that the audit forms the foundation for a programme to extend the use of ABC information within the organisation. For managers who do not have access to activity-based information, our framework provides an insight into how such information could be employed in their organisations. We hope this will act as a stimulus for such managers to explore the feasibility and implications of ABM.

Friedman and Lyne's (1995) research has shown the importance of involving a multi-disciplinary team of financial and non-financial managers in the implementation of activity based systems. Our framework has the potential to interest a range of financial and non-financial managers from different areas of the organisation. It not only provides an overview of ABC, but also a coherent picture of how such data can influence different aspects of decision making within the organisation. It provides a common platform of understanding, a standard language and conceptual underpinning, which will help managers from different parts of the business to share ideas and potential applications of ABC information. This transparency of process should help managers maintain objectivity, especially when they are investigating current decision-making practices in politically sensitive areas.

There is a risk of the framework becoming a cage rather than a platform which enables objective analysis. We hope our framework will be applied flexibly, contributing an agenda for analysing current and future decision making approaches, while allowing specific issues to be explored within their organisational context. We do not expect that all features of our framework will be equally important to all organisations, rather we expect organisations to customize the features which fit with their individual circumstances. For example, the use of cost object (position 4) data may vary greatly between different types of organisation: for a small engineering consultancy a key strategic issue may be customer profitability (position 15), so emphasis will initially be placed there, whereas, for a large vertically integrated manufacturer, transfer pricing (position 13) may be a particularly problematic area. Research suggests that ABM is more likely to be successful if an incremental approach to its implementation is adopted (e.g. Player and Keys, 1995). We recommend a phased introduction of ABC applications, which in the early stages emphasizes areas which will provide worthwhile information for decision making and are not too politically sensitive. This should help implementation through the demonstration of its positive effects.

Conclusions

Our analysis has traced the development of ABM from its origin as a product costing technique through to the extended application of ABC metrics. The number of successful ABM applications reported in the source literature is encouraging. Case examples have been used in our discussion in an attempt to bring the framework to life and to show readers the positive outcomes that can be achieved through such approaches.We have great respect for two core books (Morrow, 1992; Turney, 1991) because they were instrumental in setting the embryonic ABM agenda. Indeed Turney's work was especially prescient.

Even ABM's most ardent supporters would hesitate to claim that ABC generates "absolute truth". Its metrics, though, are certainly better than none and better, too, than the flawed data generated by many traditional cost accounting and financial accounting approaches. Having said that, there is sadly evidence from surveys (e.g. Drury et al., 1993; Innes and Mitchell, 1995), that for many organisations ABM remains no more than wishful thinking. Our framework provides an agenda which may encourage managers to employ ABC information proactively in their organisations.