The paper accounts for an Activity-Based Costing (ABC) analysis supporting decision-making concerning product modularity. The ABC analysis carried out is communicated to decision-makers by telling how much higher the variable cost of the multi-purpose module can be compared to the average variable cost for the product-unique modules that it substitutes to break even in total cost. The analysis provides the platform for stating three general rules of cost efficiency of modularization, which in combination identify the highest profit potential of product modularization. Finally the analysis points to problems of using ABC in costing modularity, i.e. handling of R&D costs and identification of product profitability upon an enhanced modularization.
In order to maintain competitiveness manufacturing companies in general aim to offer a wide selection of products to meet customers’ increased demands for variety. However, even though empirical results are not consistent (Anderson, 1995, p. 364), it is generally accepted that increased variety, or more correctly increased heterogeneity in the product mix, impacts negatively on costs and operational performance (e.g. Miller and Vollman, 1985; Banker et al., 1995; Kaplan and Cooper, 1998). The company will have to source, produce and sell in smaller batches and support functions will have to be expanded to accommodate increased internal demand for activities such as planning, set-ups, documentation, etc. To mitigate the negative effects from increased variety, manufacturing firms may pursue process-based and/or product-based strategies (Fisher et al., 1999). Product-based strategies, which are the topic of this paper, focus on product designs that allow for high product variety at reasonable cost. One such strategy is that of modularization (e.g. Heikkilä et al., 2002). When individual modules can be used in different end products, the manufacturing firm can offer variety at lower levels of component heterogeneity by combining modules and at the same time preserve some of the benefits of mass production (Perera et al., 1999).
A review of the literature on the concept and multiple effects of modularity, and paradigmatic approaches to manage modularity (Jørgensen, 2004) reveals that the concept of modularity has many faces (Hansen et al., 2003) and that a number of the economic benefits of modularization are taken for granted although the methods applied in identifying and assessing these consequences have something to be wished for.
The task of the paper is twofold. The first task is to investigate the merits of the Activity-Based Costing (ABC) as a method for assessing the cost consequences of modularization. This is done through a case study followed by reflections on how ABC (might) need to be developed to be able to serve as the relevant costing tool. The second task is to infer some general rules on the cost efficiency of modularization from the case study. In this way our contribution is of a more pragmatic character than, for example, Nepal et al. (2005), who develop a fuzzy logic model to handle cost information at the early stages of the product development process.
The paper proceeds as follows: Section 2 searches the literature on management accounting and costing to identify those parts of the (internal) value chain where cost effects of modularization are likely to occur. Section 3 provides a brief introduction to ABC, and Section 4 accounts for the ABC case study and points out some general characteristics of situations where modularization is cost effective. Section 5 reflects on problems of the ABC method in analysing the consequences of modularization beyond the specific case context. Section 6 concludes the article.
The paper accounts for an Activity-Based Cost (ABC) experiment in a case company—Martin Group A/S—to support decision-making concerning product modularity. The cost analysis pursued makes use of ABC's activity and cost object hierarchies, but the outcome of ABC analysis is communicated as “unit-level” information in terms of the maximum allowable increase in cost of materials for the over-specified potential common module compared to the average materials cost for the substituted product-unique modules. This information is instrumental in providing quick and easy-to-understand insights to designers, and can easily be expanded to encompass all unit-level costs. However, providers of this information should be aware of the prerequisites for these data to be relevant, i.e. that freed-up resources can be either taken out of the organization or redeployed in other profitable activities. The fulfilment of these prerequisites should be weighed by top management, before use of the calculations procedure is released for decentralized use in the organization to avoid distorted calculations.
If the prerequisites are satisfied, and if in addition it is assumed that an over-specified common module is at least as costly as the costliest of modules that it will be able to substitute, the paper identifies that the most profitable modularization efforts can be put where commonality between otherwise product-unique modules are high, and where volume and difference between unit-level cost of otherwise unique modules are low.
The paper also points to two areas where caution should be exercised in using ABC in assessing the economics of modularization. R&D cost of developing the initial common module is a common cost to all units and all periods in which the module is put to use. Thus, these costs are of an investment character and will in a “calendar-based” system as ABC be difficult to incorporate without arbitrary allocations to periods and/or products. In addition, it is argued that the product-profitability hierarchies resulting from extended modular structures are more complex than described in literature. This is mainly because sustaining costs of more common modules will only appear at very aggregate levels, i.e. above the level of the individual products, in non-arbitrary cost assignments; and this placement is essential to avoid distorted information.
In the specific case the materials costs of the common module were only allowed to increase by 3% of present materials cost. In the specific company this was deemed infeasible. This result provides tentative support to the existence of a modularity paradox suggested by Jørgensen (2004) as a parallel to Skinner's (1986)productivity paradox, which relates to the process-based strategy to mitigate the negative effect from increased variety (Fisher et al., 1999). We suspect that the same type of phenomenon is apparent in the product-based strategy of modularization. More research is needed in this area.
