THE ABC BANDWAGON AND THE JUGGERNAUT OF MODERNITY T [PDF]

THE ABC BANDWAGON AND THE JUGGERNAUT OF MODERNITY T Colwyn Jones Faculty of Economics & Social Science

David Dugdale Bristol Business School

University of the West of England

Abstract: ABC is a quintessential phenomenon of modernity. It begins with persons changing their practices in specific local contexts which are penetrated by global change. It is provoked by a reflexivity in which a disruption of practical consciousness leads to a new discursive consciousness. These practices and discourses are then disembedded in the creation of an abstract system of symbols and expertise. This changes the nature of the disembedded practices/discourses by translating them into a standardized (global) form. It is this global form which is then reembedded in new localities but this process is reflexively mediated in local contexts so that, again, its form changes. Whilst this is happening epistemological reflexivity works away at the globalized level so that the original expertise is undermined and another globalized discourse emerges. What makes all of this so distinctively modern is the pace at which this happens and the territory it covers. As the binding together of time and space at global and local levels becomes more powerful, so it becomes more fragile. This is the history of ABC we explore in this paper. Accounting Stream, Critical Management Studies Conference, Manchester, 14-16 July 1999. Request: This is a working paper. Please do not quote.

Address for correspondence:

David Dugdale, Bristol Business School, University of the West of England, Frenchay Campus, Bristol, BS16 1QY. Telephone: +44 (0)117 9656261 x3418.

E-mail: [email protected].

The ABC Bandwagon and the Juggernaut of Modernity INTRODUCTION This paper is concerned with the development of management accounting theories and practices and their inter-relationships. Specifically it deals with the rise of ‘activity-based’ accounting and management. We trace this development from US manufacturing in the early 1980s, through its first theoretical formulation, successive extensions and amendments to this theory, to its application in UK business in the 1990s. In the course of this journey various activity-based costing practices were identified and branded as ABC. As its originators reflected on this new accounting, and as more and more people became interested in it, ABC spawned a number of other theories and practices - such as ABB, ABC/M, ABM, Activity Accounting, Total Cost Accounting. This is “the ABC bandwagon”. The bandwagon was constructed in a world which Giddens (1990) likens to a runaway engine of enormous power that we try to drive but fear may rush out of control at any moment. This is “the juggernaut of modernity”. Faced with the complexity and risks of the modern world people must trust in abstract systems of knowledge, whilst at the same time recognizing that knowledge can never be guaranteed, and thus is not ultimately trustworthy. The technologies of accounting perpetually undermine themselves and require constant support and reaffirmation. ABC was a response to global challenges to US manufacturing, and to a crisis of confidence in “traditional accounting” as means of informing managers’ actions. It was advanced as a more trustworthy form of knowledge relevant to modern conditions. However, even as this new form of accounting was being implemented in practice, the basic principles of ABC underwent transformation which undermined its early claims. Yet, at the same time, these claims were being used to promote its implementation in new sites. Our discussion is structured as follows. We begin by laying out an analytical framework for examining the ABC phenomenon drawing on the work of Anthony Giddens. We then explore the origins of ABC in case studies of mid-1980s US manufacturing practices, their exposition as ABC by Cooper, Kaplan and Johnson in 1987-89, and the subsequent popularization and dissemination of ABC. Finally we trace the development of second-wave ABC ideas in 198992, demonstrating their rapid and fundamental transformation. We conclude that the development of ABC is classic example of the power and simultaneous fragility of expert systems in modernity. ANALYTICAL FRAMEWORK Our exploration of the development of ABC draws on Giddens’ (1990) analysis of modernity. In accounting research interest in Giddens’ work has generally been in his development of ‘structuration theory’ which attempts to replace the dualisms of agency and structure with the ‘duality of structure’(Macintosh & Scapens, 1990; Boland, 1993; Dirsmith et al. 1997). Macintosh offers a succinct summary of the main points:

structuration theory is concerned with the interplay of agents’ actions and social structures 1

in the production, reproduction, and regulation of any social order. Structures, existing in virtual time and space, and drawn upon by agents as they act and interact in specific timespace settings are themselves the outcome of those actions and interactions (1994:172). Giddens does not accord priority to either structures or action. Instead he suggests we should study social practices that both socialize (constitute) individuals as actors, and realize (embody) structures. The production and reproduction of society are seen as a skilled performance of its members who draw upon both practical and discursive consciousness. The former refers to our knowledgability about the world in a taken-for-granted way: the latter to our reflexive monitoring of action. ‘Systems’ are patterns of relationships that are structured and restructured in social practices: therefore systems have structures, but are not structures themselves. The three central structures in systems are ‘signification’, ‘legitimation’, and ‘domination’. Signification creates meaning in social interaction, domination produces power, and legitimation provides for a system’s morality. These three layers, while separable in the abstract for analytical purposes, are intimately entwined in reality (Macintosh, 1994:172). This general theoretical framework is the foundation for Giddens’ substantive sociology of modernity. It is this latter sociology on which we draw for this paper. The Nature of Modernity Giddens argues that the: dynamism of modernity derives from the separation of time and space and their recombination ... the disembedding of social systems ... and the reflexive ordering and reordering of social relations’(1990:16-7, emphasis in original). Let us consider each of these in turn before constructing a framework for research. Time and Space. For the majority of people in the pre-modern world time and place were firmly linked: we might characterize this as a concrete when-where. In modernity, time and space become separated and abstract. Time becomes standardized through the mechanical clock and universal calendar: space by uniform measurements and official maps. The separation of time and space involves standardization which ‘empties’them of the necessity for a particular local physical setting for time-space that now becomes global. This ‘distantiated’time and space can now be recombined through ‘time-space ordering devices’(such as railway timetables). Giddens refers to this process as ‘disembedding’ which means ‘the "lifting out" of social relations from local contexts and their restructuring across almost indefinite spans of time-space’(1990:21). This is particularly evident in the way modern rationalised organisations connect the local and the global. Disembedding Institutions. Giddens argues that the scope of ‘time-space distantiation’ is greatly extended by the development of disembedding institutions. He identifies two key mechanisms in this process. First there is the development of symbolic tokens, in particular - money. Modern money exists both as cash and coins in everyday transactions and as disembedded symbolic token. It is a mode of deferral in time and implies a space between individuals and their possessions. Thus it is ‘a means of bracketing time-space by coupling instaneity and deferral, presence and absence’ (1990:25). The second important disembedding mechanism is the development of expert systems, described as ‘systems of technical accomplishment or professional expertise that organise large areas of the material and social environments in which we live today ... An expert system disembeds by ... providing "guarantees" of expectations across distantiated time-space’ (1990:27). The layperson has little direct access to the knowledge that is inherent in such abstract systems and so the 2

plausibility of these guarantees rests upon faith in ‘the authenticity of the expert knowledge which they apply’(1990:28). Giddens argues that: In conditions of modernity, larger and larger numbers of people live in circumstances in which disembedded institutions, linking local practices with globalised social relationships, organise major aspects of day-to-day life (1990:79). This everyday organization involves reembedding which is ‘the reappropriation of disembedded social relations so as to pin them down (however partially or transitorily) to local conditions of time and space’(1990:79). Reflexivity. Giddens argues that modernity generates new forms of reflexivity. In early modernity social change appeared as a move from tradition to reason. So, for example, Weber (1922) identified rationality as the defining feature of modern society. At first this seemed to promise the replacement of the certainties of traditional knowledge with new certainties of science. However, as modernity developed, it became clear that modern reflexivity subverts reason: the rational questioning of all things leads to questioning the assumptions on which the questions rest. In the constant monitoring of behaviour and contexts, of thought and action, all knowledge becomes unstable. Whereas some have identified this as the post-modern move, for Giddens it is the working through of the modern (‘radicalised modernity’) as it comes to understand itself. In modernity knowledge is always provisional, contestable, and open to differing understandings. Since reflexivity becomes the basis for system reproduction, the instability of knowledge produces an unstable world. Given that the knowledge of expert systems cannot be guaranteed, why then do people continue to trust in them? Giddens argues that we make a ‘bargain with modernity’ (1990:90). We are socialized to trust so that as we learn science we also learn a respect for science as reliable knowledge. This spills over into respect for other forms of technical knowledge. This respect exists in conjunction with scepticism so that attitudes are ambivalent - a balance of respect and scepticism. In the modern world there is no escape from expert systems and ‘attitudes of trust towards abstract systems are usually incorporated into the continuity of day-to-day activities and are to a large extent enforced by the intrinsic circumstances of daily life’ (1990:90). Thus we have a predisposition to trust, which is constantly being undermined, but lack of alternatives means we make pragmatic compromises that are woven into our routines. ounting and Modernity: an Analytical Framework Drawing together these aspects of modernity we now construct an analytical framework for study of the development of accounting theory and practice. Accounting can be seen as a social practice that constructs information: concrete things and events are recreated as abstract values and exchanges. In the modern corporation such accounting information conditions the possibilities of ‘acting at a distance’ through its ‘centres of calculation’ (Latour, 1987). Accounting provides the mechanisms by which information on the local (e.g. operating units) is communicated to the global (eg. head office) where it constitutes a means to act back into the local (Johnson & Kaplan, 1987). In this process, accounting information is emptied of local (tacit) knowledge and changes its meaning (Armstrong, 1987a). The resulting ‘management by the numbers’ involves particular recombinations of time and space. The power of accounting as a disembedding mechanism rests to a considerable extent on its ability to re-present other forms of data and calculation in terms of money as pure information - thus 3

linking itself to modernity’s most potent symbolic token. Miller argues: ‘Whilst accounting shares with other means of quantification such as statistics the ability to translate qualities into quantities, it does so largely by translating these qualitative differences into financial values which need no further referent’(1994:2). Accounting, then, claims to provide the ultimate translation - the bottom line. Accounting is not only a mechanism of disembedding: it is an institution which is itself disembedded and reembedded recursively. In the process of disembedding the nature of accounting changes. The act of disembedding involves a reflexivity which mediates between concrete practices and abstracted theory. Accounting also changes as it is reembedded in local contexts where there it may meet with consent, mediation and/or resistance as it encounters reflexive monitoring in local contexts. In part, this is signalled by the long-running discussion of ‘the gap between theory and practice’ (Edwards & Emmanuel, 1990). The usual assumption is that the “gap” is created by “theory” taking a lead which “practice” is slow to follow. In our analysis, what is taken to be leading and following (or what is prior and subsequent) is a function of the point at which we cut into the cycle. If the dissemination of new accounting techniques is studied then it appears that theory leads practice: if learning from field research or personal experience is investigated then it seems that practice informs theory. In our approach theories and practices are seen as socially constructed in recursive cycles of disembedding and reembedding in which each influences the other (see Figure 1). FIGURE 1 ABOUT HERE In our discussion of ABC which follows there is a major disembedding-reembedding cycle which takes activity-based theories and practices from the local contexts of particular US manufacturing plants, through their globalized abstraction as ABC, and inserts them in new and diverse local contexts. Within this are minor cycles in which reflexive monitoring of contexts and actions continually reforms these theories and practices at both global and local level. THE ORIGINS OF ABC There are two main strands in the development of accounting theories and practices. The first, which is more familiar to academics, flows through the work of Bob Kaplan, Tom Johnson, Robin Cooper, and their associates who brought “ABC” to global attention. For convenience (with some apology to Johnson) we label this “the Harvard approach”. The second, less well known, is “the CAM-I approach”, mostly represented through the work of Jim Brimson, which produced “Activity Accounting”. In this paper we follow the first of these strands1. Theoretical Roots The origins of what was to become “ABC” appear highly specific. It began in the practices of a small number of US manufacturing companies in 1985 as interpreted in Harvard Business School cases written by Robin Cooper in 19862. By 1988 these practices had been branded as

1

The analysis of the second strand is not yet complete. Already we regard it as an important area for understanding the rise of activity-based accounting and management, and it will be included in later drafts of the paper. 2

There are occasional references to earlier literature on activity-based accounting approaches (Staubus, 1971; cited in Johnson, 1988) and vague references to practices in General Electric in the early 1960s (Johnson,

4

‘a new approach’ that produced ‘activity-based information’ and heralded as ‘a blueprint for world-class management accounting’ (Johnson, 1988) and a year later they were brand-named as ABC (Cooper, 1989). From this account the origins of ABC seem clear enough - it was ‘discovered’ in emergent accounting practices in US manufacturing in the mid-1980s. What were the circumstances of this discovery? Perspective on Manufacturing. US manufacturing in the early 1980s was seen as experiencing a wave of new opportunities and threats. There was the development of forms of advanced manufacturing technology (AMT) - for example, through successive generations of numerically-controlled machine tools (NC, CNC, DNC); flexible manufacturing systems (FMS); computer-aided design (CAD) which could be linked to the computer-aided manufacturing systems (CAD-CAM); all of which offered the prospect of computer-integrated manufacturing (CIM), There was increasing international competition which threatened to relative security of American home markets. In particular, there was the challenge from Japanese manufacturing, which was believed not only to be making superior use of AMT, but also to have devised more effective management techniques - with just-in-time (JIT) and total quality management (TQM) beginning to recognized as central to Japanese success.. There was also a concern that consumer markets might be changing so that the American dominance in mass consumption/mass production business could now be destabilized. The long-run, largebatch, long-term production in which the US had previously excelled was might now be obsolete in face of demand for short-run, small-batch, short-lined products. In short, the US businesses were seen as facing major issues in a ‘new manufacturing environment’. This, at any rate, was the diagnosis advanced by many Harvard University academics, with much of their work published in the Harvard Business Review. The titles of these works are evocative - ‘The focused factory’ (Skinner, 1974); ‘Managing our way to economic decline’ (Hayes & Abernathy, 1980); ‘The hidden factory’ (Miller & Vollman, 1985), ‘Postindustrial manufacturing’ (Jaikumar, 1986). These, with other texts such as ‘Japanese management techniques’ (Schonberger, 1982) and ‘The Japanese corporation’ (Abegglen & Stalk, 1985), informed the Harvard approach to manufacturing and all are cited in later ABC articles. Critique of Traditional Accounting. Alongside this general perceptive on manufacturing, the Harvard approach was influenced by Kaplan’s pre-ABC critique of existing accounting systems. In `Yesterday’s accounting undermines production’ (1984) he lays out the themes which will drive his work over the next 15 years. He identifies three problem areas: inadequacies of traditional costing systems; the use of the ROI measure; and the dominance of the financial accounting mentality in enterprises. At this stage there is no sign of any activitybased approach but there is an interest in the use of nonfinancial indicators (which would eventually emerge in the ‘balanced scorecard’ literature of the early 1990s). The paper also reflects Kaplan’s commitment to field research (Kaplan, 1986, 1988b) as a means of developing accounting theory - a method which continues to be important over the coming years (Kaplan, 1998). The Initial Case Studies.

1992b). However, these are recollected retrospectively after the “brand-named” product ABC had been produced.

5

Field Research “Failure”. Kaplan (1985) reports the findings of the first of his excursions into the realm of management accounting practice. He identified a ‘select set’ of innovative companies (including Hewlett-Packard, Intel, IBM, Westinghouse, and Omark Industries) on the assumption that since they were leaders in developing JIT, TQM and CIM, they were likely to be implementing innovative accounting systems. He was disappointed with his findings: I had hoped to be able to document the incidence and value of innovative accounting and control systems for the new industrial competition; to learn how firms making major changes in their manufacturing operations were developing and using measures of quality, inventory reductions, manufacturing flexibility, employee morale and abilities, productivity, and new product effectiveness. Instead I found that changes in accounting procedures lag far behind changes in the real phenomena they are supposed to represent (p.75). Kaplan speculates on the reasons for this lack of innovation in accounting and suggests four: lack of role models (in contrast to the Japanese techniques informing manufacturing change; the prevalence of computer-based systems with extensive traditional accounting programs; the emphasis on financial accounting even among managerial accountants; and the: most important explanation for accounting lag ... that senior company management have not emphasized the need to improve the relevance and responsiveness of their management accounting systems (p.78). Thus there is no sign of innovative management accounting - let alone activity-based costing in Kaplan’s survey of innovative US manufacturing companies in 1985. Cooper’s Discovery. A discovery was, however, made in the same year by one of Kaplan’s colleagues - Robin Cooper. There is no indication that Cooper had self-consciously sought innovative companies in the same manner as Kaplan. Perhaps he found interesting sites not because they were innovative (and successful) but because they had urgent problems that had to be addressed. Two cases were of key importance: the Schrader Bellows and John Deere companies. In a series of Harvard Case Studies Cooper, with various research assistants (Cooper et al.,1985), tells the story Schrader Bellows Group - a division of Scovill Inc. The case opens with the views of William F. Boone, Vice President of Planning and Development at Scovill, who notes that, in a diversified company, it is not uncommon for a few products to subsidise the others and that: traditional cost accounting systems systematically mask the damage caused by losing divisions, product lines and products (p.321). In 1983, Boone analysed the product profitability of the Schrader Bellows Group. One plant produced over 2,700 different final products and stocked up to 20,000 parts: ‘Over the years, the firm’s strategy of providing a full product line had resulted in a considerable proliferation of variations for each product.’ The firm had a traditional standard costing system but with several overhead pools rather than a single plant-wide pool. Product costs were calculated by the firm’s computerized standard cost system as the sum of material, direct labor, and overhead costs. [To calculate overhead costs the] total costs of the setup and quality control departments were allocated to the manufacturing departments... The allocated costs from each of the support departments were then added to the overhead costs of the manufacturing departments, creating a separate overhead pool for each manufacturing department. Departmental burden rates were determined by dividing each overhead pool by the direct labor hours worked in the department (p.3245).. 6

Boone then appointed two people - the Head of Strategic Planning (PhD aeronautical engineer with an MBA in finance) and the Manager of Strategic Planning (degrees in economics and law and an MBA) to determine the profitability of products manufactured at the plant. When they visited the plant they noted that: the procedures used to allocate support department costs to the products were simplistic and, in our opinion, bore little relationship to the underlying economic reality (p.329).. They decided to interview the support department heads in order to find better bases for allocating support overhead. In a later case study Cooper reports a developing costing procedure at Schrader Bellows: Our first step was to explode the sales data. We took the quantity sold of each end product, determined the individual components contained in the end products, and then estimated the quantity of each component required in 1982. The second step, costing the components according to the selected allocation bases, was the most challenging. A great deal of judgement was required to allocate the support department costs down to individual components. We also spent a considerable amount of time insuring that the factory and sales, general and administration costs were appropriately allocated. In our opinion, very few of these costs were truly fixed. After we completed the allocation of costs to the components, we imploded these costs back to the product level by summing the costs of all components in a given product (p.336). A computer program was written to calculate product cost data for some 2,000 products and it took several months to debug the program and overcome the many problems that had to be solved. Although we are provided with little detail it seems that this exercise was a fore-runner of what was later to become known as activity-based costing. The aim of the Schrader Bellows exercise is clear. It was to allocate overhead costs more accurately to products - allocating not only manufacturing overhead but also sales, general and administration costs. Kaplan’s Discovery. Kaplan was soon to find his own example of activity-based costing. This is the case of John Deere Component Works (JDCW) again published as Harvard Case Studies (March & Kaplan, 1987) prepared by research associate Artemis March under Kaplan’s supervision. JDCW was part of a vertically integrated company, a captive producer for Deere’s equipment divisions with virtually all sales internal. In practice, equipment divisions did not always follow the corporate guidelines for internal sourcing, and JDCW lost a portion of the equipment factories’business to outside vendors (p.293). By early 1984 the Gear and Special Products Division of JDCW had much excess capacity and its managers set out to penetrate other markets (such as automobile OEMs). Also, in October 1984, when Deere invited bids for the manufacture of 635 parts Gear and Special Products bid for 275 of them, but was awarded only 58. The bidding experience generated a good deal of ferment at Gear and Special Products and confirmed the feeling of many that “we didn’t even know our costs.” Some of us were quite alarmed. We had been saying , “Let’s go outside,” [to gain orders] but we couldn’t even succeed inside...And when we looked at the results, we knew we were not

7

costing things right. It was backwards to think we could do better in low-volume than high-volume parts, but that’s what the cost system said (p.295). JDCW employed a standard cost accounting system which, for many years, had used direct labour run time as the base for overhead absorption; in 1984, machine hours was introduced as well. Keith Williams and Nick Vintila (a manufacturing supervisor) were chosen to undertake a special study in Gear and Special Products and, as a first step, they studied a sample of 44 of the 275 bid parts. Vintila spent the first half of 1985 working full-time on what became known as the ABC (Activity-Based Costing) study... he and Williams learned that use of overhead resources could be explained by seven different types of support activities: direct labor support, machine operation, setup hours, production order activity, materials handling, parts administration and general overhead. Vintila then analysed all the cost categories in the general ledger, estimating what percentage of each cost was ‘caused’ by each of the seven ‘overhead driving activities’(p.297). When the activity-based method was used to allocate overhead, 41% of the overhead shifted from the original two pools (labour; machine support) to five new pools (machine setup; production orders; materials handling; part administration; and general & administration). A Lotus 1-2-3 spreadsheet was used to calculate activity-based costs for individual parts and Williams and Vintila were able to demonstrate the change in estimated costs for the sample of 44 parts. Further studies were undertaken and ‘ABC was widely embraced for decision making in the machined parts business’. ABC was used to prepare bids for both Deere and outside customers; to re-evaluate the economics of long production runs; to inform decisions about which parts should not be produced on automatic machines; and to help division managers in deciding how to re-arrange the machining departments. When a new division manager was appointed to Gear and Special Products in September 1986 he described his division’s response to activity-based costing: Few things have generated more excitement. Even though its still an allocation, it’s such an improvement. Parts we suspected we were undercosting have turned out to be even more expensive than we had thought. It’s proven what we suspected about the costs of material handling and transport distances, and triggered our making layout changes. When it showed us the costs added by secondary operations, we brought them back onto the main floor (p.304). The John Deere case shows clearly how activity-based costing was seen as an improvement on the traditional standard cost system. Instead of two overhead cost pools there were seven and the estimated costs were much more reliable. The system was very obviously seen as a means of better allocating overhead to product (“even though it’s still an allocation it’s such an improvement”). The consequences of ABC analysis can, again, be seen clearly, pricing, outsourcing and method changes were areas that immediately received attention in the light of the ABC analysis. The genesis of ABC seems to lie in the Schrader Bellows (1985) and John Deere (1987) cases. Both concentrated on recosting manufactured products so as to better inform management as to the economic consequences of particular activities. In both cases product costs were calculated that included revised allocations of manufacturing overhead. In the Schrader Bellows case the allocation of non-manufacturing overhead to product also featured. The difference between the “old” standard costs and the “new” activity-based costs was a key

8

feature in both cases and typical managerial actions included dropping products, repricing and changing processes and layouts. The term ‘activity-based costing’ appears for the first time in the Deere case, as is recognized later by Kaplan (1990). However, this does not immediately grab the interest of the researchers: instead the subject matter of the two studies is described as ‘designing and implementing a transaction-based product cost system’ (Kaplan, 1988b, p.7.2, emphasis added). Supporting Cases. Having now discovered companies with innovative costing practices, Cooper and Kaplan were soon to find more cases. The most important of these was presented by Tom Johnson at a colloquium in June 1986 at the Harvard Business School. Cooper & Kaplan presented the Schrader Bellows case and, as Kaplan was to later to report, Johnson described ‘a similar system developed at the Weyerhaeuser Company to measure and control corporate overhead costs’ (Kaplan, 1994, p.248). Soon other companies were added to the list: After these initial observations on the new transaction-costing approach, Robin Cooper and I saw essentially identical systems that had been developed and implemented at several other manufacturing companies: John Deere, Hewlett-Packard and Tektronix in the U.S., Siemens in Germany; and Ericsson and Kanthal in Sweden (Kaplan, 1994, p.248, emphasis added). Kaplan was later to describe learning from these innovating companies: Cooper’s and my initial rate of knowledge creation were highest in the period 1986-1989 when we worked directly with our initial population of companies (1998, p.109). It may be seen as something of a puzzle why a number of manufacturing companies, in different countries, suddenly appeared to be spontaneously generating not only innovative forms of management accounting, but ‘essentially identical’ forms. Perhaps even more surprising, companies in other industrial sectors were also found to display the same phenomenon: service organizations, such as banks, transportation companies and health care institutions, were also beginning to use costing systems with the same underlying structure as those we had observed in innovating manufacturing companies (Kaplan, 1994, p.248, emphasis added). Only a year or so after Kaplan had reported the failure of his first field trip to uncover any new practices, activity-based costing appeared to be erupting all over the place (including HewlettPackard - previously a non-innovative company). The Rise of Activity-based Accounting. We find two possibilities for explaining this phenomenon. First, following the analytical framework we outlined above, the form of reflexivity which influenced the researchers’ approach to their case studies may have played a significant role in shaping their interpretations. They already had a general perspective on the state of US manufacturing constructed from themes of the implementation of new technology, increasing international competition, and change in consumer markets impacting on production. They also had a general critique of existing accounting practices - which was soon to advanced in Relevance Lost (Johnson & Kaplan, 1987). This may have inclined them to emphasize what was general in the cases and to pay less attention to the detailed differences. The view that activity-based costing began in a handful of sites, before it was generalized and popularized by academics (see below) appears to gain support from Cooper’s (later) comment that: 9

Activity-based costing - ABC - is a relatively new concept, the oldest known system only having existed for a few years. But activity-based costing systems are gaining lots of attention’(1989). There were differences, even in the two initial cases. The term ‘activity-based costing’ may have been coined at John Deere, but it is the Schrader Bellows case to which Cooper and Kaplan repeatedly return over the next few years (while John Deere is only accorded occasional references). Despite Kaplan’s insistence that ‘cost systems are context-specific’, that there ‘are no generic cost systems that work well in all environments’, and that systems must be ‘tailor-made’ (1988b, pp. 7.3-7.4), what was to emerge was ‘a blue-print for worldclass management accounting’ (Johnson, 1988). This would suggest that, in disembedding specific local practices so that they could become global, abstract principles, the researchers acted upon their materials to shape them into a coherent movement. The second possibility is that there were widespread changes in US manufacturing (and perhaps elsewhere) which were driving company management accounting systems along new, diverse, but eventually convergent paths. This is suggested by Johnson’s (1992b) history of activity-based initiatives beginning in the US company General Electric in the 1960s, developing through the ‘cost-driver’ approaches of consultancy firms such as Bain & Co., and Boston Consultancy Group in the 1970s and early 1980s, and being implemented in some US companies (Johnson adds Union Pacific and, possibly, Caterpillar and Hewlett-Packard to the two initial Cooper & Kaplan cases). The contribution of academics Gordon Shillinglaw (at Columbia University) and George Staubus (at Berkeley) are noted but not accorded much influence: the activity concepts they enunciated seem not to have influenced other academic thinking (until very recently), nor do they appear to have influenced ... activity-based developments in business (Johnson, 1992b, p.27). All of this constituted the business development of: cost-driver activity-based costing (ABC), [which] was eventually codified by Harvard Business School professor Robin Cooper (p.29, emphasis added). This interpretation places little weight on the role of academics in constructing activity-based costing - they merely tidy-up and package pre-existing developments in industry3. THE EXPOSITION OF ABC In the mid-1980s the materials gathered by Cooper, Kaplan and Johnson began to crystallize as ‘activity-based costing’. Although the term is not actually used in Relevance Lost (Johnson & Kaplan, 1987), the book draws together the themes of the new manufacturing environment and the inadequacies of traditional management accounting that we noted earlier, and begins to set out prescriptions for the future. Relevance Lost The discussion of process control emphasizes the need for clearly defined cost centres and the identification of appropriate measures, which need not be direct labour hours: For instance, machine hours may be relevant for highly automated departments, number of orders received or processed for the receiving department, number or some physical 3

The social construction of accounting pictured in this view is perhaps more clearly evident in the second strand of activity-based accounting and management - the CAM-I approach.

10

measure (pounds, gallons, square meters) of orders shipped for the shipping department, number of set-ups and pounds of material moved for an indirect labor department. Our goal should be to do the best we can in explaining the short-term variation in costs within each cost center. It is argued that there is a danger in using short-term variable costs for pricing, product introduction/deletion and make versus buy decisions: These decisions turn out to involve the commitment of the firm’s capacity resources and should be made in the light of long-term, not the short-term, variability of costs ... A good product cost system measures the long-run costs of each product.’(pp.233-4). They argue that we can treat: ‘virtually all costs as variable.’ (p.234) and: ‘Assuming that the current size and costs of the overhead departments is defensible ...our task is to identify the cost drivers for these departments.’ (p235). Typical cost drivers are identified and reference made to Miller and Vollmann’s (1985) taxonomy of transaction types. Finally a (rather vague) explanation is given of the way revised product costs can be derived form the ‘cost per driver’ data generated. This is seen as producing surprising information that can be generated by such an exercise and the actions that might be taken. A further section deals with the need to extend the analysis to ‘costs outside the factory’ (p.244) and attributes the neglect of nonmanufacturing costs as ‘likely [to] be attributed to a financial reporting story somewhere in the distant past.’Implicit in all of this are the ideas of activity-based costing. Early ABC Papers Once the idea of activity-based costing had crystallised it did not take long to disseminate and popularize it. Practitioner-oriented journals - Harvard Business Review and Management Accounting (US) - were important publication vehicles and several articles were published in 1988 and 1989. These instructed accountants and managers that ‘One cost system isn’t enough’ (Kaplan, February 1988a), that ‘[traditional] cost accounting distorts product costs’ (Cooper & Kaplan, June 1988), that ‘a solution is at hand. Known as activity-based costing systems’ (Johnson, June 1988), and that this will enable managers to ‘Measure costs right: make the right decisions’(Cooper & Kaplan, October 1988), so that ABC is the ‘key to future costs’ (Cooper, 1989). In the course of these few months the term ‘transaction costing’ was replaced with ‘activity-based costing’ as the preferred label, and then graced with the acronym ABC - putting it in line with other management techniques of the period. What was the nature of this early ABC? Cooper and Kaplan’s ABC Kaplan (1988a) identified three purposes for cost systems: inventory valuation, operational control, and individual product cost measurement, and argued that the cost system for external reporting did not give managers relevant information for performance measurement and product cost purposes. In particular standard cost systems are criticized for failing to measure product costs ‘accurately’, and compared with ‘a more accurate system’ (not named, but presumably the Schrader Bellows case). Thus companies need an alternative to their standard costing system to discover how their product costs are being distorted. To construct this information: Analysts, attempting to understand the demands a product makes on the company’s resources, can start by interviewing the supervisors of production, support, logistics, and marketing departments... Product cost estimates will not have the five- and six-digit precision reported by a standard cost system. They will also more subjective and less precise than the measurements in an operational control system ... But the estimates will realistically approximate the long-run demands each product makes on the organization’s resources” (p.64, emphasis in original). 11

What is on offer then is not a cost system which is more accurate because it is more precise, nor because is it less subjective: rather it claims to be more accurate by being more realistic The theme is pursued in Cooper & Kaplan 1988a. Again the importance of product costs and managers’ rejection of short-term perspectives is emphasized. Managers are reported to be unconvinced that their full-cost systems are adequate for product related decisions. Cases are cited (Mayers Tap, Rockford and Schrader Bellows) as evidence that product costs can be systematically distorted, and that this impacts on important pricing and product decisions. Sections of the paper deal with the failure of the two-stage allocation system, the failure of marginal costing, and the failure of fixed-cost allocations based on labour cost or hours. Having set the scene with a fictional case (the ubiquitous “pen example”) the article then discusses the importance of ‘transaction costing’ utilizing the Schrader Bellows case. The company is reported as having recently: performed a “strategic cost analysis” that significantly increased the number of bases used to allocate costs to products ... The effect of changing these second-stage allocations was dramatic ... the change in reported product costs ranged from minus10% to plus 1,000% (p.24). The changes are reported in table which shows seven products’ ‘unit cost’ and ‘unit gross margin’ under the existing cost system and the ‘transaction-based system’. The differences between the two sets of figures are then used to demonstrate the distortion that traditional systems generate: this distortion is ‘serious’, ‘systematic’, and cannot (in general) be overcome without a ‘different type of allocation base ... for overhead costs that vary with the number of transactions performed, as opposed to the volume of product produced’(p.25). In Cooper & Kaplan 1988b we find a confident exposition of the alternative to traditional cost systems - this now being referred to as ‘activity-based costing’. After a cursory dismissal of the old ‘simplistic’approaches they introduce their own: The theory behind our methods is simple. Virtually all of a company’s activities exist to support the production and delivery of goods and services - and they should therefore be considered to be product costs (p.96). Thus they begin by re-asserting the point that activity-based costing about all costs (not just factory costs) and suggesting that (virtually) all costs can be related to product. We believe that only two types of costs should be excluded from a system of activitybased costing. First, the costs of excess capacity should not be charged to individual products....The cost of excess or idle capacity should be treated as a separate line item the cost of the period, not of individual products ... The second exclusion from an activity-based cost system is research and development for entirely new product lines (pp.101,102). Finally, Cooper & Kaplan consider the strategic implications for executives now that they are ‘armed with more reliable cost information’ that an ABC analysis might generate. They are careful to note that dropping unprofitable products is not the only option, prices can be raised and/or processes redesigned. They conclude: Activity-based costing is not designed to trigger automatic decisions. It is designed to provide more accurate information about production and support activities and product costs so that management can focus its attention on the products and processes with the most leverage for increasing profits. It helps managers make better decisions about product, design, pricing, marketing, and mix, and encourages continual operating improvements (p.103).

12

These three papers lay out the main features of the Cooper & Kaplan version of activity-based costing, which Cooper (1989) summarized and illustrated with case studies. Activity based cost systems are more accurate than traditional volume-based systems ... they all appear to rely upon a two-stage allocation procedure to indirectly trace the costs of resources consumed in the production process to the end products (emphasis added) The new approach had been discovered in US manufacturing plants (most notably Schrader Bellows). It involved analysts gaining interview data on activities in overhead and service departments, identifying cost-drivers - from two in a very simple case to nine in a relatively complex case (Cooper, 1989) - and then tracing virtually all costs (except excess capacity and R&D) to products via a two-stage allocation procedure. Essentially then, ABC was seen as a superior version of full absorption costing. The enthusiasm demonstrated for it by Cooper & Kaplan was only tempered by their warning that the superior information it generated should not be used to make ‘automatic decision’. The vision for ABC was still relatively narrow: it should be used to supplement, rather than supplant, other cost systems. The central role of ABC was to supply managers with information - crucially ‘strategic cost management’ information. Johnson’s ABC. Some further dimensions to ABC were added by Johnson (1988). The scene is set by a first page headline: To be world-class competitors companies should manage activities - not costs (p. 23). Whilst recognizing that: ‘strategic cost information, enables managers to assess the long-term profitability of a company’s current mix of products’(p.24), Johnson concentrates on the ways in which activities might be managed as sources of competitive value. If Cooper & Kaplan’s central concern was the more accurate measurement of costs, Johnson was more concerned with attending to the activities which caused them, and advocated: four steps to managing waste in operating activities: chart the flow of activities throughout the organization; identify sources of customer value in every activity; and eliminate any activities that contribute no identifiable value to customers; identify causes of delay, excess and unevenness in all activities; and track indicators of waste ... businesses need information that it will make it possible for managers to identify and to eliminate generators of nonvalue activity (pp.28,29). He identifies a number of nonfinancial measures - elapsed time, distance moved, space occupied, number of part numbers - that might be important in managing activities, and argues that JIT and TQM companies need to reduce their emphasis on financial measures. With this approach Johnson sets a much wider agenda for ABC - now seen as the e technical core of a general management approach - and, in this, anticipates the development of activitybased management (ABM). POPULARIZATION OF ABC ABC had now been, in Giddens terms, disembedded from the specific local contexts of accounting practise and constructed as a discourse at the global-abstract level. This discourse was then to be disseminated and reembedded in a wide range of new local contexts. Interest in ABC ideas was both widespread and rapid. In part this interest was new - inspired by the articles themselves - and in part it confirmed other similar developments in activity-based approaches. 13

CAM-I Of the similar developments the most important seems to be the work of CAM-I. This organization - originally Computer-Aided Manufacturing, International, later renamed Consortium for Advanced Manufacturing, International - is a research and development group sponsored by a coalition of some of the largest industrial organizations, the big professional accountancy firms, and government agencies in the US (see Figure 2). In 1986 it set up a group composed of representatives of these organization (plus university academics) to create the CAM-I Cost Management System. FIGURE 2 ABOUT HERE The brief for this group was to condense ‘the cumulative experience of the sponsors who are some of the largest and most successful companies in the United States and Europe’ (Berliner & Brimson, 1988, p.xi) and construct from this a unified cost management system (CMS) available to CAM-I members. The project was planned in three phases: conceptual design (1986); systems design (1987); and implementation (1988), with Jim Brimson as CMS Project Director. The time period of the project closely overlaps with the development of ABC, and both Kapan and Cooper are listed university members and identified for special thanks ‘for the significant amount of time they have devoted to working with the CAM-I staff and the sponsor organizations’ (Berliner & Brimson, 1988, p.xv). Brimson (1988) presented a paper on the CAM-I cost project at the same conference Kaplan delivered ‘Regaining relevance’. There were clearly quite close connections between the Harvard group and CAM-I, and the outcome of their work was similar. In the case of CAM-I it was the construction of “Activity Accounting” which aims: ‘to measure the cost of resources consumed in performing significant activities of the business’ (Berliner & Brimson, 1988, p.85) and from this to provide information on product cost, pricing, life-cycle costs, and the non-value-added cost element. Management Consultancy and Practitioner Publications Others were also interested in the ABC approach. The Peat Marwick McLintock Cost Management Group approach to product costing in AMT environments was influenced by ABC (Jeans & Morrow, 1989). The attempt by Ernst and Young’s Cost Management and Manufacturing Industry consulting services to differentiate their own product as “total cost management” (TCM) did not disguise its close relationship to ABC (Ostrenga, 1990). Cooper & Kaplan were themselves involved with consultancy when, in the late 1980s: we made a strategic alliance with a large consulting organization (KPMG Peat Marwick), trained their U.S. and overseas consultants, and served as consultants and observers to their clients’activity-based costing projects (Kaplan, 1998, p.102). The outcomes of these, and other developments were publicized in the Management Accounting - the professional magazine of the Institute of Management Accountants in the UK. Interest grew rapidly reaching a high point in the mid-1990s (see Figure 3). Even though there was a decline in the number of articles after that ABC retained its claim on readers’ attention, and for the 1997-98 issue the gold and two joint silver Lybrand awards for best articles went to, respectively, case studies on: JIT, target costing, and ABC: ABC solely; and ABC and TOC (theory of constraints). FIGURE 3 ABOUT HERE

14

ABC in the UK ABC ideas quickly crossed the Atlantic into the UK. The spread of these ideas is demonstrated in the pages of Management Accounting (UK), the magazine of the Chartered Institute of Management Accountants (CIMA). Readers were introduced to the new thinking in September 1988 through an interview with Kaplan (Maskell,1988). In the next ten years there were over 100 articles which dealt with ABC (or ABC-related issues such as ABB and ABM) peaking in 1995 (see Figure 4). These consisted of expositions of ABC, general reviews of developments in ABC theory and practice, reports of CIMA meetings and research projects on the subject, surveys of ABC implementation, case studies - and the occasional critique. FIGURE 4 ABOUT HERE The case studies report implementation in manufacturing (Dugdale & Shrimpton, 1990; Laing, 1994; Brent, 1995; Lyne & Friedman, 1996) communications (Gwynne & Ashworth, 1993; Bussey, 1993), publishing (Mitchell & Wycherley, 1994), and brewing (Mason, 1996). Innes & Mitchell (1991) reported that almost half the CIMA members they surveyed in 1990 - a mere five years after its discovery, and only a couple after its formal exposition - were employed by organizations which had considered implementing ABC4. Brent (1992) found an astonishing 90% of sampled firms considering ABC with 10% implementing it: and by 1994 16% of sampled CIMA members were using ABC (Innes & Mitchell, 1995). Whilst we have reservations about the validity the data gained from mail questionnaire surveys of this type (see Dugdale & Jones, 1998) the pages of Management Accounting point to widespread awareness, and some implementation, of ABC ideas across a range of UK businesses. Most of the general expositions and reviews of ABC are faithful to its 1988-89 form, or incorporate new features into the absorption costing framework of that period. As late as 1997 Lucas commented that: textbook writers ... continue to present ABC as a product costing system - a more sophisticated absorption costing system, using a number of absorption bases. Thus it appears that the ABC, as disembedded from local manufacturing contexts in 1983-86, and recreated at the global-abstract level in 1988-89, was becoming reembedded in practitioner thinking, and sometimes action, in the UK of the mid-1990s. THE TRANSFORMATION OF ABC THEORY Whilst this reembedding was proceeding, the theoretical development of ABC at the globalabstract level moved apace. Two key moves by Cooper & Kaplan were the making of a distinction between resources supplied and resources used and the construction of the ABC cost hierarchy (Johnson moved in an entirely different direction). Second-wave ABC Resources Supplied and Used. The first development extends, rather than overturns, earlier ABC thinking. In Cooper & Kaplan (1991, 1992) the 1988 notion of ‘excess capacity’ is redefined as ‘unused capacity’, and is given far more weight in understanding cost structures. The distinction between ‘resources used’and ‘resources supplied’is now seen as crucial. 4

This survey appeared to be confirmed by work published elsewhere that, in 1990, 60% of large UK firms were claiming to be using or planning to use ABC (Bright et al., 1992).

15

Activity-based cost models estimate the cost of resources used in organizational processes to produce outputs .. the costs of resources supplied or available [is] reported by the organization’s periodic financial statements (1992, p.1, emphasis in original). Later Kaplan explained the difficulties that arose when this distinction had not been fully appreciated: ABC costs were, in general, greatly in excess of the amount which spending would change, leading these people [i.e. sceptical academic colleagues5] to question the value, in fact the validity, of the ABC cost concept (1994, p.253). The Cost Hierarchy. The second development was of much more fundamental significance the construction of “ABC cost hierarchy”. Cooper & Kaplan (1991) point out that activity analysis allows costs to be analysed into those that relate to unit level activities, batch level activities, product sustaining activities and facility sustaining activities. (Other authors have mentioned customer sustaining and distribution channel sustaining activities.) This may seem to be merely a useful analytical insight but, in Cooper & Kaplan’s hands, it has profound repercussions. Armed with the hierarchical analysis they claim (in a boxed headline): ABC is a powerful tool - but only if managers resist the instinct to view expenses at the unit level (p.130) and elaborate in the text, that: Managers must refrain from allocating all expenses to individual units and instead separate the expenses and match them to the level of activity that consumes resources (p.130). This is a theoretical re-statement of shattering proportions. Cooper & Kaplan admit that: Initially, managers viewed the ABC approach as a more accurate way of calculating product costs (p.130). As we have shown earlier, managers were not alone in this: Cooper & Kaplan’s ‘simple theory’ of 1988 states that virtually all costs should be treated as product costs - excluding only excess capacity and R&D. By 1991 the list of excluded costs has grown dramatically. Unit-level activities include consume only direct labour, materials, machine costs and energy. Above this level there are batch-level activities (consuming setups, material movements, purchase orders, and inspection) and product-sustaining activities (consuming process engineering, product specifications, engineering change notices, and product enhancement). These costs are not to be applied at unit-level: Allocating expenses to individual units sends signals that managers can easily misinterpret. When batch- and product- level costs are divided by the number of units produced, the mistaken impression is that the costs vary with the number of units (p.132). This does not mean that costs cannot be traced to products, but the way in which this is done is now very different. Product cost over a time period can be derived by summing unit, batch and product-sustaining support resources consumed: ‘Production of shaft A103 consumed $2,453 worth of resources last year’ (p.132). However stating the unit cost of production (‘This product costs $30-plus a unit’, p.132) is found to be ‘less useful’. The 1991 version moves further by arguing that some costs should not be allocated to products at all. We are now told that there are: ‘largely “fixed” costs of owning and running the plant’ (p.133). They are “fixed” because ‘product-related activities do not impact facilitylevel costs - plant management, building and grounds, heating and lighting. This means that:

5

Eli Goldratt is the only one named specifically.

16

Only unit, batch, and product-sustaining expenses should be assigned to products. Therefore, in ABC, facility-level expenses are kept at the plant method and not allocated to products (p.133) This is an extraordinary change. ABC was originally advanced as a superior costing system. This was based on the premise that traditional methods of “allocating” (in the UK “absorbing”) support costs to products was flawed and misleading. One of the key features of ABC was its ability much more accurately to perform this allocation and hence better inform managers in making key decisions - such as product start-up/deletion, pricing, outsourcing. The superiority of ABC was specifically demonstrated through the unit costs it generated. This was the point hammered home in repeated uses of the Schrader Bellows valve costing example. Not only was more accurate unit costs the promise that ABC held out: it was also the proof of its superiority over traditional systems. The Nature of Second-wave ABC In the 1991-92 version of ABC is not to be used to derive “better” unit-level product costs. Instead, activity analysis should be employed in order to better understand the hierarchy of costs in the organisation and then managers will have better information for decision making. They will be able to identify relevant revenues and costs that should be considered for particular decisions. This is, of course, standard textbook advice and this view of activity-based analysis is much closer to a marginal or contribution style approach than it is to “better allocation” of overhead support costs to products. This point was recognized by Kaplan in 1992: When Robin Cooper and I first encountered ABC systems in the mid-1980s, in sites such as Schrader-Bellows, John Deere, and Union Pacific, we described as an allocation the procedure by which operating expenses were supplied, via activities, to products and services (p.59, emphasis added). By 1992 two things have happened to the presentation of ABC. Firstly, and perhaps of a cosmetic nature, references to “allocation” have been dropped. [Originally] Our use of the term “allocation” alarmed people who have been taught or are currently teaching that allocations are arbitrary and therefore provide no information relevant for management decisions ... [Now we realize] The process we had described initially as an allocation is an estimation. That is an attempt to estimate the cost of performing a setup or of processing a customer order ... we have found that estimates, based on interviews, employee judgements, and available operating data, are usually sufficiently accurate for the managerial use of the information from an ABC model (p.59, emphasis added). Second, there has been the identification of more non-product-related costs. This significantly changed ABC so that: Once Robin developed and articulated the hierarchical structure of activity-based costing, we understood that ABC was really a contribution margin approach, not an attempt to get “more accurate fully-allocated unit costs” (p.59) In the 1991-92 version of ABC it was no longer the more accurate allocation procedure it had been claimed to be (Cooper, 1989). Instead it was concerned with estimates of the costs of activities which are sufficiently accurate (according to Cooper & Kaplan’s experience and judgement) for managerial purposes. It turns out that ABC was really a contribution margin approach.

17

Neither according to Johnson, the third important figure in the disembedding of ABC, did it hold the promise he had earlier claimed for when he concluded: Activity-based management accounting information is the key to continuous improvement of profitability, a journey without end (1988, p.30, emphasis added). By the early 1990s he had changed his views considerably: As someone who helped put the activity-based concept in motion, I feel compelled to warn people that I believe it has gone too far. It should be redirected and slowed down, if not stopped altogether” (1992, p.26). Indeed he went further to argue that what had emerged was: a new competitive environment - call it the global economy - in which accounting information is not capable of guiding companies toward competitiveness and long term profitability (p.31). The man who had been the most enthusiastic cheerleader on the ABC bandwagon had jumped off, and the two who remained aboard were trying to steer it in new directions. How had these changes come about? Kaplan’s Retrospectives Kaplan (1990, 1994, 1998) gives us various explanations of the processes producing secondwave ABC in his reflections on his, and ABC’s, journey from the mid-1980s to the mid-1990s. The 1990 Retrospective The first reflection on the development of ABC comes from Kaplan’s contributions as a panel member at the 1989 Annual Meeting of the American Accounting Association (Kaplan, 1992). Kaplan begins by noting that the view of contribution analysis that ‘has been pervasive for fifty or sixty years’ (p.2) splits operating expenses into two cost pools - fixed and variable. The latest incarnation of this received wisdom is Eli Goldratt’s theory of constraints. However: many companies have resisted for the most part the attempts by academic accountants to convince them to ignore their fixed costs. Most companies persist in performing full cost allocations. One could be forgiven for thinking that ABC is therefore going to have something to say about cost allocations. However, by 1989, Kaplan was already changing the nature of ABC: The new approach to contribution margin analysis derives from activity-based costing (ABC). We did not invent this name: it was used by the John Deere company’ (p.5, emphasis added). Kaplan acknowledges that his readers might have misinterpreted ABC in their his and Cooper’s earlier writings: Our statement that the costs of “all of a company’s activities ... should be considered product costs6” has led many people to believe we were saying that all expenses, even expenses caused by activities above the product level, should be allocated down to a unitproduct cost calculation...[However] Intuitively, we knew that obtaining such a unit-cost number was impossible (p.6,7, emphasis added). Given the confidence with which the original statement was written, it is perhaps not surprising that some readers might have thought it meant what it said. 6

The full original statement is presented earlier in our paper.

18

Kaplan then goes on to note that batch or product sustaining expenses can be divided by production volumes but this: is clearly an allocation process, with all the disadvantages that have been discussed for the past sixty years ... [It would be better to take] the opportunity of not allocating expenses down from north to south, but of aggregating them up from south to north ... No allocations are required to obtain this product-level margin (p7). Kaplan’s summary leaves us in no doubt that, in 1989, he sees the “correct” use of ABC as part of a marginal costing analysis: The ABC analysis strives to assign operating expenses to factors that cause them. Contribution margins can be calculated at the unit level; or after covering batch-related expenses, and even after covering product line, customer, and distribution channel expenses. Here the momentum for change to ABC is presented as initiated by their own intuition. Cooper & Kaplan had reported that, in their cases, managers were dissatisfied with unit costs derived from traditional costing systems, and had invented a new costing approach which they (the managers) believed provided more accurate unit costs. In retrospect Kaplan now claimed that he and Cooper intuitively knew that accurate product costs it wasn’t really realistic. Here they distance themselves from the case studies. They were merely reporters of, and not advocates for, the use of ABC. Our interpretation differs. We believe that Cooper & Kaplan did advocate ABC as a superior allocation system for unit costs: that they were “learning” from the cases, not merely “reporting” them. Their articles in practitioner-oriented journals were not neutral academic reports: they were committed exhortations to change. To put this another way, the 1989 revisionism was not driven by “intuition” but by “hindsight”. The 1994 retrospective. The next review is a more comprehensive history of events. Kaplan describes the origins of ABC, mentioning the following cases: Schrader Bellows, Schrader Bellows and Winchell Lighting, John Deere Component Works, Hewlett-Packard: Roseville Networks Division, Siemens Electric Motor Works and Kanthal. Kaplan acknowledges the “early” version of ABC: For the first few years of articulation and implementation of ABC systems, they were described as being “more accurate” than traditional cost allocation systems and of supplying “long-run variable costs” ... The breakthrough for ABC systems came from two theoretical developments that elevated the approach from deductive assertions to scientifically testable hypotheses: (1) discovery of the cost hierarchy of indirect and support expenses; and (2) the distinction between costs of resources supplied and the costs of resources used - the role for unused capacity costs (p.249). What are the origins of this change in the 1994 history? The conceptual advances occurred well after the initial discovery of ABC systems in practice. These advances also came after the first wave of writings that described ABC systems and how they differed from the traditional cost systems they were replacing. These advances came from studying and teaching cases describing early ABC implementations and by working closely with companies that were implementing ABC systems (p. 254, emphasis added). These companies come into two categories: those who had developed ABC systems on their own - many contacted through CAM-I; and those who approached Cooper & Kaplan for

19

assistance in building ABC systems, and with which they often worked over a period of several years. Here we are offered three drivers of change: reflecting on the original studies through studying and teaching; discovering new (independent) developments; and assisting in the implementation of (some form of) ABC in novice (dependent) companies. The creation of the cost hierarchy is explained in terms of an initial discovery of a limited aspect of it (the marketing, distribution and channel-sustaining hierarchy at Winchell Lighting) and as a ‘theoretical development’ reported by Cooper in 19907. At the same time, academics are distanced from practice through their reflexive reconsideration of past events, and moved closer to it through a second wave of disembedding in the present (e.g. the CAM-I companies), and reembedding, where academics actively engage in the construction of future ABC systems. The 1998 retrospective. This weaving together of the past, present and future is elegantly presented in the third retrospective (Kaplan, 1998) - just as though it had, self-consciously, been there all along. It is now identified as ‘innovation action research’. This mode of research engages both academics and practitioners in attempting to implement theoretical ideas and then modifying and extending the theory in the light of the knowledge gained. The research moves through four stages - which are recursive8. The first stage is ‘to observe and document innovative practice’ and Kaplan’s story begins (in the early 1980s) when: I became persuaded that a major gap existed between the needs of managers in leadingedge companies and the practices then being taught and studied by academics (pp.92-3). Recalling his first excursion into the field (see above) he decided that conventional ‘descriptive’field research was unhelpful: what if some scholars believe that the current practices occurring in companies are not desirable or optimal? ... Extensive studies of existing practice would merely document obsolete and ineffective practice (pp. 89, 94). The solution was to articulate the problem to those who might know of possible solutions. In the ABC case this was done through the Harvard MBA programme in 1984. Kaplan states that executives at Scovill (the Schrader Bellows) and John Deere cases alerted him to the innovative approaches he was seeking. Meanwhile Cooper, through CAM-I, gained contact with Siemens and Hewlett-Packard. The second stage is to teach and speak about the innovation: The preparation and delivery of these talks ... forced us to go “meta” - to generalize the phenomena occurring in these diverse organizations, not just describe the details of each companies experience (p.101). The third stage is to write books and articles: At this point, each concept had evolved far enough that it could be named [as ABC] and illustrated with experiences from the initial innovating companies(p101). 7

The source of this theoretical development in unclear here. The previous version - ‘Robin developed and articulated the hierarchical structure’ (Kaplan, 1992) - suggests there was a more active academic role than reporting. 8

Kaplan’s third retrospective was published some time after we had become interested in Giddens’ disembedding-reembedding framework. The correspondence between the two we find eerie.

20

And the fourth stage is to implement the concept in new organisations. The first loop around this circuit takes us from 1984 to 1989 - the first-wave of ABC. The second loop produces second-wave ABC. Using this framework, Kaplan retells the story of the development of the resource use/supply distinction and the cost hierarchy. In the case of resource use/supply it is the ‘teach-speak’ and ‘write’ parts of the cycle which appear the more important. ABC was confronted by a hostile alternative - TOC (Theory of Constraints) - produced by Eli Goldratt9. This adopted an extreme version of marginal costing where all costs except materials are considered fixed (with specific exceptions, such as royalties, for particular cases) and where the concept of “excess capacity” is replaced with that of “protective capacity” - unused resources which buffer bottleneck machines or locations. Cooper and I participated in a series of three conferences, spanning 1988 to 1991, sponsored by the Institute of Management Accountants. These three conferences primarily featured two emerging management innovations; activity-based costing and Eli Goldratt’s Theory of Constraints (Goldratt & Cox, 1986). Recall the familiar expression that “nothing focuses the mind so much as an imminent hanging.” Well, running a close second place is engaging in public debate with Goldratt who is an articulate and persuasive advocate of his deductive theory of constraints. As we struggled with reconciling our emerging theory of activity-based costing with Goldratt’s well articulated theory of constraints, we came to understand the central role for measuring activity cost drivers based on resource capacity. This led us to articulate the difference between Goldratt’s cost definition, which related to the costs of resources supplied, vs. the activity-based costing definition, which related to measuring the costs of resources used (p.107). The amendment to ABC did nothing to shake Goldratt’s conviction that ‘cost accounting is enemy number 1 of production’ (the charge is repeated more strongly in the second edition of Goldratt & Cox, 1994), and the two camps have never been reconciled. However, the deductive reasoning of Goldratt certainly seems to have influenced the inductive (as it was originally presented) approach of ABC. The construction of the cost hierarchy, on the other hand, is attributed more to the implementation stage of loop one and the observational stage of loop two. Hewlett-Packard, Chrysler, General Electric, Kanthal, Maplehurst Bakeries, and British Columbia Telecommunications are cited as implementations: Hewlett-Packard, Maxwell Appliances, and Pillsbury as new observations. It was these new cases that replaced Schrader Bellows and John Deere as the sources of ABC information. As we worked with other manufacturing and service organizations, we learned that many indirect expenses helped to sustain individual products and customers; they were not supplied to support transactions. Cooper, after systematically studying the variety of cost drivers in our population of organizations ... developed a powerful taxonomy (p.105). The third retrospective, it appears, restores the practice-theory relationship of 1988. Business has supplied examples of existing innovative practices, the academics have generalized it, and

9

For a discussion of the history of the Theory of Constraints see Jones & Dugdale, 1998, and for its application as Throughput Accounting, Dugdale & Jones, 1998.

21

they will go on to promote it. The distancing of researchers as mere reporters (Kaplan, 1990) has gone - and Cooper & Kaplan are striding out on their third loop (Kaplan, 1998). The reconciliation between theory and practice might be seen (as implied in Kaplan, 1998) as ABC practice being reflexively reconsidered following the same path as that taken by Cooper & Kaplan - and so the two converge after the 1989 parting of the ways. In essence there is a “basic” and an “advanced” ABC and researchers and companies move towards it - sometimes independently; sometimes together. This would fit with an evolutionary view of the “progress” entailed in accounting history. We offer another possible10 interpretation. That in the early 1980s the researchers discovered new forms of full-absorption costing for producing unit costs, and in the late 1980s/early 1990s they discovered new forms of marginal costing for understanding cost behaviour. Of course, both were concerned with “activities” in some way, but in other respects they were very different. What they had in common was that they were both packaged and promoted as new accounting solutions under the label of ABC. Perhaps it was simply the power of “ABC” as a brand name, plus the personal and corporate investment (and reputation) involved in its construction, that led researchers to regard their observations as two forms of the same phenomenon, rather than as two different phenomena. DISCUSSION In the development of first- and second-wave ABC we see a complex inter-relationship between theory and practice: between academia and business. In the early in 1980s we have a group of academics with theoretical perspectives on the general nature of the ‘new manufacturing environment’and the ‘failure of traditional accounting’. At first they can find no innovatory management accounting practices to study. Then, around 1985, there are three independent discoveries made by Cooper, Kaplan, and Johnson in US manufacturing companies. Soon there are more ‘essentially identical’ cases - including three in Europe although the first case (Schrader Bellows) continues to be the most important. These disembedded practices are then packaged and disseminated as the ABC theory - a ‘world-class management accounting’ which provided more accurate product costs through a two-stage allocation procedure. Then this first-wave ABC is confronted by academic critique, by issues arising in implementation (under advice from the academics), and new independent developments in practice. Thus second-wave ABC appears which is a contribution margin approach which helps managers understand their costs better. While second-wave ABC is emerging, first-wave ABC continues to be reembedded in local contexts - reinforced by discussion of ABC in textbooks. What then is ABC by the mid-1990s? Is it an absorption costing system for product costs, or a marginal costing system to inform managers on cost structures? It seems to us that ABC is in a state of confusion in the period. There are both first- and second-wave practices, and the theory shows elements of both. For example, where do Cooper & Kaplan stand on the question of fixed versus variable costs? By the mid-1990s the confident assertion that ‘virtually all costs are variable’ (Johnson & Kaplan, 1987) has gone. Instead we have the more careful statement: 10

We raise this as a possibility in order to problematize the “official” history set out by Kaplan.

22

Costs are not intrinsically fixed or variable. ABC analysis permits managers to understand the sources of cost variability (Cooper & Kaplan, 1991, p.135). This, however, is not an acceptance that there need be fixed costs. Second-wave ABC analysis promises to enable managers to reduce demands made on the organization’s resources. However: If managers fail to follow up any reductions in the demands on organizational resources, improvements will create excess capacity, not increased profits. Managers might then conclude erroneously that operating expenses were indeed fixed and not variable ... Having reduced the demands, managers can then increase throughput or to reduce spending to convert the savings into increased profits (p.135). In short, costs are variable if and when people manage to vary them - which seems something of a truism. Even the question of whether or not second-wave ABC allocates costs to products at unitlevel is not clear-cut. Cooper & Kaplan (1991) find that unit costs are not useful. However, in Kaplan and Atkinson’s (1998) textbook, students are told: When pricing policies are derived from a traditional (unit-level) standard costing system ... managers can make poor pricing decisions ... [However] Companies, after an initial ABC analysis, have frequently been able to sustain price increases of 50% or more for their speciality, customized, and - as they now see - much more costly products (p.151). This seems to imply that meaningful unit-level ABC costs can still be derived to inform pricing decisions - and we are returned to the 1988 claim that ABC produces more accurate product costs. In Giddens’(1990) discussion of disembedding and reembedding in modernity, new forms of knowledge are constantly emerging and being bound into expert systems. This knowledge is created through the reflexive monitoring of actions and contexts. Knowledge is unstable because the self-doubting nature of modern rationality which means that it is constantly being undermined, and expert systems must be reformulated and re-legitimated if they are to be trusted to inform decisions in the modern world. We suggest that the development of ABC is classic example of the speed and complexity of these processes. REFERENCES Abbeglen, J.C & Stalk, G.Jnr. (1985), Kaisha: The Japanese Corporation, New York, Basic Books. Armstrong, P. (1987), “The rise of accounting controls in British capitalist enterprises”, Accounting, Organisation & Society, Vol. 12, pp.415-436. Berliner, C. & Brimson, J.A. (1988), Cost Management for Today’s Advanced Manufacturing: The CAM-I Conceptual Design, Boston, Harvard Business School Press. Boland, R.J.Jnr. (1993), “Accounting and the interpretive act”, Organizations and Society, Vol. 18, No. 2/3, pp.125-146.

Accounting,

Boland, R.J.Jnr. (1996), “Why shared meanings have no place in structuration theory: a reply to Scapens and Macintosh”, Accounting, Organizations and Society, Vol. 21, No. 7/8, pp. 691-97. 23

Brent, M. (1995), “Activity-based costing at Wavin”, Management Accounting (UK), Vol. 73, No. 5, May, pp. 28-30. Brent, N. (1992), “ABC in the UK: a status report”, Management Accounting (UK), Vol 70, No. 5, May, pp. 22-23, 28. Bright, J., Davies, R.E., Downes, C.A. & Sweeting, R.C. (1992). “The deployment of costing techniques and practices: a UK study”, Management Accounting Research, Vol . 3, pp. 201-211. Brimson, J.A. (l988), “CAM-I cost management systems project', in Capettini R. and Clancy D.K. (eds) Cost Accounting, Robotics and the New Manufacturing Environment, American Accounting Association. Bussey, B. A. (1993), “ABC within a service organisation”, Accounting(UK), Vol. 71, No. 11, December, pp 40-41, 65.

Management

Cooper, R. (1989), “ABC: key to future costs”, Management Consultancy, October. Cooper, R. & Kaplan, R. S. (1988a), “How cost accounting distorts product costs”, Management Accounting (US), Vol. 69 No. 10, April, pp. 20-27. Cooper, R. & Kaplan, R. S. (1988b), “Measure costs right: make the right decision” Harvard Business Review, September-October, pp. 96-103. Cooper, R. & Kaplan, R. S. (1991), “Profit priorities from activity-based costing.” Harvard Business Review, May-June, pp 130-135. Cooper, R. & Kaplan, R. S. (1992), “Activity-based systems: measuring the costs or resource usage”, Accounting Horizons, September, pp. 1-19. Cooper, R., Weiss, L. & Montgomery, J. (1985), “Schrader Bellows”, Harvard Business School case 186-050/4, reprinted in Cooper, R. and Kaplan, R. S. (eds) (1991) The Design of Cost Management Systems, New Jersey, Prentice Hall Inc, pp. 321 - 345. Dirsmith, M.W., Heian, J.B. & Covaleski, M.A. (1997), “Structure and agency in an institutional setting: the application and social transformation of control in the Big Six”, Accounting, Organizations and Society, Vol. 22, No. 1, pp.1-27. Dugdale, D. & Shrimpton, S. (1990), “Product costing in JIT environment”, Management Accounting (UK), March, pp. 40-42. Dugdale, D. & Jones, T.C (1997), “How many companies use ABC for stock valuation? A comment on Innes and Mitchell's questionnaire findings”, Management Accounting Research, Vol. 8, No. 2, pp. 233-240. Dugdale, D. & Jones, T.C. (1998), “Accounting for throughput: transforming practices?”, British Accounting Review, Vol. 30, pp.203-220. Edwards, K. A. & Emmanuel, C. R., “Diverging views on the boundaries of management accounting”, Management Accounting Research, Vol. 1, 1990, pp. 51-63.

24

Giddens, A. (1990), The Consequences of Modernity, Cambridge, Polity. Goldratt, E. M. & Cox, J. (1984, 1986, 1994), The Goal, London, Gower. Gwynne, R. & Ashworth, G. (1993), “Implementing activity-based management at Mercury Communications”, Management Accounting (UK), Vol. 71, No. 11, December, pp. 34-36. Hayes, R.H. & Abernathy, W.J. (1980), “Managing our way to economic decline”, Harvard Business Review, July-August, pp.67-77. Hayes, W.W. & Wheelwright, S.C. (1984), Restoring Our Competitive Edge, New York, Wiley. Innes, J. & Mitchell, F. (1991), “ABC a survey of CIMA members”, Management Accounting (UK), Vol 69, No. 9, October, pp. 28-30. Innes, J. & Mitchell, F. (1995), “ABC: a follow up survey of CIMA members.” Management Accounting (UK), Vol. 73, No. 7, July-August, pp. 50-51. Innes, J. & Mitchell, F. (1995), “A survey of activity based costing in the U.K.'s largest companies”, Management Accounting Research, Vol. 6, pp. 137-53. Jaikumar, R, (1986), “Postindustrial manufacturing”, Harvard Business Review, November-December, pp. 69-76. Jeans, M. & Morrow, M. (1989), “Management accounting in AMT environments: product costing”, Management Accounting (UK), November, pp. 42-44. Jones, T.C. & Dugdale, D. (1998), “Theory of Constraints: Transforming Ideas?”, British Accounting Review, Vol. 30, pp. 73-91. Johnson, H. T. (1988), “Activity-based information: a blueprint for world-class management accounting”, Management Accounting (US), June, pp. 23-30. Johnson, H.T. (1992a), Relevance Regained: From Top-down Control to Bottom-up Empowerment. New York, Free Press. Johnson, H. T. (1992b), “It's time to stop overselling activity-based concepts: start focusing on customer satisfaction instead”, Management Accounting (US), September, pp 26-35. Johnson H.T. & Kaplan, R.S. (l987), Relevance Lost: The Rise and Fall of Management Accounting Cambridge, Mass; Harvard Business School Press. Kaplan, R. S. (1984), “Yesterday’s accounting undermines production”, Harvard Business Review, July-August, pp. 95-101. Kaplan, R. S. (1985), “Accounting lag: the obsolescence of cost accounting systems”, in Clark K. B., Hayes R. H. and Lorenz C. (eds.) (1985), The Uneasy Alliance: Managing the Productivity-Technology Dilemma, Mass., Harvard Business School Press.

25

Kaplan, R.S. (1986), “The role for empirical research in management accounting”, Accounting, Organizations and Society, Vol. 11, No. 4/5, pp.429-452. Kaplan, R. S. (1988a), “One cost system isn’t enough.” Harvard Business Review, JanFeb, pp. 61-66. Kaplan, R. S. (1988b), “Regaining relevance”, in Capettini R. and Clancy D.K. (eds) (1988) Cost Accounting, Robotics and the New Manufacturing Environment, American Accounting Association. Kaplan, R. S. (1990), “Contribution margin analysis: no longer relevant - strategic cost management: the new paradigm”, Journal of Management Accounting Research, Vol. 2, pp. 1-32. Kaplan, R. S. (1992), “In defense of activity-based cost management”, Management Accounting (US), November, pp. 58-63. Kaplan, R. S. (1994), “Management accounting (1984-1994): the development of new practice and theory”, Management Accounting Research, Vol. 5, pp. 247-260. Kaplan, R. S. & Atkinson, (1998), Advanced Management Accounting, 3rd edition, Prentice-Hall International. Kaplan, R.S. (1998), “Innovation action research: creating new management theory and practice”, Journal of Management Accounting Research, Vol. 10, pp. 89-118. Kaplan R. S. & Norton D. P. (1992), “The balanced scorecard - measures that drive business performance”, Harvard Business Review, January-February, pp. 71-79. Laing, E. F. (1994), “Accounting tools really can help: product profitability analysis.” Management Accounting (UK), Vol 72, No. 9, October, pp. 50-51. Latour, B. (1987), Science in Action: How to Follow Scientists and Engineers Through Society, Milton Keynes, Open University Press, 1987. Lucas, M. (1997), “Absorption costing for decision-making”, Management Accounting (UK), Vol. 75, No. 9, October, pp. 42-44. Lyne, S. & Friedman, A. (1996), “Activity-based techniques and the ‘new management accountant’”, Management Accounting (UK), Vol. 74 No. 7, Jul/Aug, pp. 34-35. Macintosh, N.B. & Scapens, R.W. “Structuration theory on management accounting”, Accounting, Organizations & Society, Vol 15, pp.455-477. Macintosh, N.B. (1994), Management Accounting and Control Systems: An Organizational and Behavioural Approach, Chichester, John Wiley. March, A. & Kaplan, R. S. (1987), “John Deere Component Works”, Harvard Business School case 187-107/8, reprinted in Cooper, R. & Kaplan, R. S. (eds.) (1991), The Design of Cost Management Systems, New Jersey, Prentice Hall Inc.

26

Maskell, B. (1988), “Management accounting: relevance regained: an interview with Professor Robert S. Kaplan”, Management Accounting (UK), Vol. 66 No. 8, September, pp. 38-42. Mason, B. (1996), “Activity-based budgeting at Scottish Courage”, Management Accounting (UK), Vol. 74 No. 7, Jul/Aug, pp. 32. Miller, P. (1994), “Accounting as social and institutional practice: an introduction” in A.G.Hopwood & P.Miller (eds) (1994) Accounting as Social and Institutional Practice, Cambridge, Cambridge University Press. Miller, J.G. & Vollman, T.E. (1985), “The hidden factory”, Harvard Business Review, September-October, pp. 142-150. Mitchell, M. & Wycherley, I. (1994), “ABC from first principles”, Accounting (UK), Vol 72, No. 6, June, pp. 13-15.

Management

Ostrenga, M. R. (1990), “Activities: the focal point of total cost management”, Management Accounting (US), February. Scapens, R.W. & Macintosh, N.B. (1996), “Structure and agency in management accounting research: a response to Boland’s interpretive act”, Accounting, Organizations and Society, Vol. 21, No. 7/8, pp.675-690. Staubus, G.J. (1971), Activity Costing and Input-Output Accounting, Illinois, Irwin. Skinner, W. (1974), “The focused factory”, Harvard Business Review, May-June, pp. 113-121. Schonberger, R.J. (1982), Japanese Manufacturing Techniques: Nine Hidden Lessons in Simplicity, New York, Free Press. Weber, M. (1922), Wirtschaft und Gesellschaft, Tübingen, J. C. B. Mohr, translated by G.Roth & C.Wittich (1968) Economy and Society, New York, Bedminster Press.

27

Figure 1

Analytical Framework

global-abstract level

disembedding

reembedding

local-concrete level

28

Figure 2

CAM-I Sponsors and University Affiliates, 1988

-------------------------------------------------------------------------------------------------------Sponsors Corporations Automotive Sector, Allied Signal Alison Gas Turbine Division, General Motors Boeing Military Aircraft British Aerospace Eastman Kodak Eaton Corporation Garrett Turbine Engine General Dynamics/Fort Worth General Electric Grumman Corporation BMY Division, Harsco Corporation Hughes Aircraft International Computers ITT-AMT Center Lockheed Aeronautical Systems Lockheed Missiles & Space Martin Marietta Energy Systems McDonnell Douglas Motorola Plessey Pratt & Whitney Aircraft RHP Industrial Bearings Rockwell International TRW Westinghouse Electric Williams International Government Agencies GTE Government Systems AFWAL/MLTC, US Airforce US Department of Energy US Navy Accountancy Firms Arthur Andersen Arthur Young Coopers & Lybrand Ernst & Whinney Peat, Marwick, Main Price Waterhouse

29

Universities & Professional Societies University Affiliates George Foster (Stanford) Robert S. Kaplan (Harvard) Jeffrey Miller (Boston) Wickham Skinner (Harvard) University Review Panel Michael C. Burstein (Industrial Technology Institute) Bernard Coda, (North Texas State) Robin Cooper (Harvard) Bela Gold (Claremont) Anthony Hopwood (LSE) Dan Shunk (Arizona State) William G. Sullivan (Tennesse) Gerry Susman (Pennsylvania State) Professional Societies National Association of Accountants American Association of Accountants American Society of Mechanical Engineers -------------------------------------------------------------------------------------------------------Source: Berliner & Brimson, 1988

30

Figure 3

Articles on activity-based accounting and management published in Management Accounting, US, 1988-98.

--------------------------------------------------------------------------------------------------Year Expo. Case Report Survey Crit. All Specific General --------------------------------------------------------------------------------------------------1988 1 1 1 1989

2

-

-

-

-

2

2

-

1990

9

3

-

-

-

12

9

3

1991

7

2

-

-

1

10

9

1

1992

7

4

-

1

2

13

11

2

1993

4

10

8

1

-

23

14

9

1994

8

6#

3

2

-

19

13

6

1995

6

4

2

1

-

13

8

5

1996

3

6#

1

2

-

12

11

1

1997

-

9#

1

-

-

10

10

-

1998*

4

7

4

1

-

16

6

10

--------------------------------------------------------------------------------------------------All 51 51 19 7 3 131 94 37 --------------------------------------------------------------------------------------------------NOTES: Expo.

= expositions of ABC/ABM theory, general reviews of developments, student examples. Case = single or multiple case studies (# = includes one case which is seen as “failure” of ABC or where an alternative approach is advocated). Report = reports of meetings (including IMA and CAM-I), announcements of SMAs (Statements of Management Accounting), book reviews. Survey = postal questionaire or survey interview research of ABC/ABM practice (or teaching). Crit. = critiques of ABC/ABM essentially opposed to theory or practice. Specific = item wholly or predominantly concerned with ABC/ABM General = item which discusses a range of issues with ABC/ABM being given significant weighting 1998* = Only January-September surveyed . WARNING: These are provisional findings: please do not quote the statistics from draft table.

31

Figure 4

Articles on activity-based accounting and management published in Management Accounting, UK, 1988-98.

--------------------------------------------------------------------------------------------------Year Expo. Case Report Survey Crit. All Specific General --------------------------------------------------------------------------------------------------1988 1 1 1 1989

2

-

1

-

1

4

3

1

1990

4

1

-

-

1

6

5

1

1991

3

-

-

1

1

5

4

1

1992

2

1#

1

1

1

6

4

2

1993

4

3

4

-

2

13

8

5

1994

6

2

3

-

-

11

8

3

1995

19

3

4

2

2

30

24

6

1996

7

3

2

1

-

13

9

4

1997

6

1

1

-

1

9

6

3

1998*

7

-

1

-

-

8

4

4

--------------------------------------------------------------------------------------------------All 61 14 17 5 9 106 76 30 --------------------------------------------------------------------------------------------------NOTES: Expo.

= expositions of ABC/ABM theory, general reviews of developments, interviews with experts, student examples. Case = single or multiple case studies (# = includes one case which is seen as “failure” of ABC or where an alternative approach is advocated). Report = reports of CIMA meetings, announcements of research grants or outcomes. Survey = postal questionaire or survey interview research of ABC/ABM practice (or teaching). Crit. = critiques of ABC/ABM essentially opposed to theory or practice. Specific = item wholly or predominantly concerned with ABC/ABM General = item which discusses a range of issues with ABC/ABM being given significant weighting 1998* = Only January-September surveyed . WARNING: These are provisional findings: please do not quote the statistics from draft table.

32

33