Abstract

Activity-Based Costing (ABC) research often measures adoption as a binary organisational status, while activity-management research captures the extent to which firms use related practices. This descriptive survey study assesses whether these alternative measurement approaches produce consistent pictures of ABC use in UK non-manufacturing firms. Of 219 completed questionnaires, 204 provided usable observations for the reported ABC-status and activity-management analyses. Formal ABC adoption was reported by 21.08% of the usable sample, whereas 59.31% reported using ABC-level activity-management practices to at least a small extent. Activity Analysis (AA), Activity-Cost Analysis (ACA), and ABC-level use are treated as related, potentially overlapping practices rather than mutually exclusive stages. Cross-group patterns show that formal adopters report the strongest ABC-level use, while firms at earlier, rejected, or non-consideration statuses may still report selected activity-oriented practices. The findings demonstrate that reported ABC prevalence is sensitive to measurement choice and caution against comparing adoption rates that operationalise ABC differently. The study therefore contributes an empirical measurement perspective rather than a causal theory of ABC implementation.

Keywords

Activity-Based Costing, Activity Management, Activity Analysis, Activity-Cost Analysis, Full ABC Implementation

Share and Cite:

1. Introduction

Traditional absorption-costing systems assign direct costs directly to cost objects and allocate manufacturing overhead using predetermined allocation bases, often volume-related measures such as labour hours or machine hours. Non-manufacturing costs, including many selling and administrative costs, are commonly treated as period costs for external reporting, although they may be assigned to services, customers, or other cost objects for internal decision-making purposes. Heavy reliance on broad volume-based allocation bases can distort cost information in complex environments characterised by diverse products and services (Drury, 2018; Al-Omiri & Drury, 2007; Drury & Tayles, 2005).

In response to these limitations, activity-based costing (ABC) emerged in the late 1980s to provide a more accurate and nuanced understanding of costs. ABC addresses the shortcomings of TCS by focusing on the activities that generate costs, rather than merely grouping overhead costs into departmental pools. This system assigns costs based on specific activities and their drivers, distinguishing between volume and non-volume cost drivers (Kaplan & Cooper, 1998). ABC categorises activities into a hierarchy, including unit-level, batch-level, product-sustaining, and facility-level activities, thereby facilitating more precise cost allocation (Cooper & Kaplan, 1991).

Despite its advantages, the implementation of ABC has revealed measurement challenges. Differences in measurement across the various levels of ABC—such as activity analysis (AA), activity-cost analysis (ACA), and the ABC method itself—have resulted in varied outcomes regarding its application (Jusoh & Miryazdi, 2015). Furthermore, there is a notable lack of studies exploring the relationship between ABC and activity management (AM), which pertains to the practical application of ABC within organisations (Innes & Mitchell, 1995).

This research addresses the measurement gap by comparing formal ABC-adoption status (an organisational experience/status measure) with reported use of three activity-management practices: Activity Analysis (AA), Activity-Cost Analysis (ACA), and ABC-level use. Rather than assuming that these practices form mutually exclusive stages, the study treats them as related and potentially overlapping practices that may coexist within the same organisation. The objective is to assess whether alternative operationalisations of ABC generate different pictures of its prevalence and use in UK non-manufacturing firms (Alshamlan & Zverovich, 2018; Baird et al., 2004).

Accordingly, the contribution is empirical and measurement-focused rather than the development of a new causal theory. The study evaluates whether formal adoption categories align with reported intensity of activity-oriented practices and whether measurement choice can help explain inconsistent ABC adoption rates in prior research. The following discriminating propositions guide the descriptive comparisons:

P1. Firms reporting formal ABC adoption are expected to report stronger ABC-level use than firms in non-adoption categories.

P2. Firms closer to an adoption decision (intending to adopt or currently investigating adoption) are expected, on average, to report greater engagement with AA, ACA, and ABC-level practices than firms that rejected or never considered ABC.

P3. Firms that previously adopted and subsequently abandoned ABC may retain AA or ACA practices, but are expected to report lower current ABC-level use than current formal adopters. These propositions are directional and can be contradicted by the observed response distributions; no response category is automatically classified as supportive.

By comparing status-based and practice-intensity measures, the study provides evidence on how alternative operationalisations affect reported ABC prevalence. The analysis is descriptive and does not claim causal effects of AM on ABC adoption or implementation.

The subsequent sections will include a literature review and research propositions (Section 2), followed by a discussion of the research methodology and methods (Section 3). This will be succeeded by the research results (Section 4) and a conclusion (Section 5).

2. Literature Review

Various researchers have defined ABC adoption and non-adoption in numerous ways, leading to inconsistencies in measurement. For instance, Bjørnenak (1997) classified ABC adopters as those currently implementing, planning, or recently abandoned the system. However, this study did not consider other classifications, such as those merely considering ABC, which could lead to inaccurate assessments of adoption rates. In contrast, Krumwiede (1998) suggested that companies that had abandoned ABC should still be classified as adopters due to their prior experience, complicating the distinction between adoption and non-adoption.

Van Nguyen and Brooks (1997) provided a clearer classification by identifying three categories of ABC adoption, including those who planned to adopt it again after abandonment. However, their framework still lacks nuance regarding the varying degrees of commitment to ABC. Similarly, Innes et al. (2000) and Clarke and Mullins (2001) offered simplified classifications that failed to capture the complexity of adoption behaviours, merely distinguishing between users and non-users without considering those in transitional phases.

Brierley (2011) criticised the prevailing classifications, arguing that they lump diverse groups together, which can obscure meaningful differences in operational performance between adopting and non-adopting units. He proposed more precise definitions, yet the challenge of accurately categorising organisations persists, as highlighted by Cohen et al. (2005) and Jusoh and Miryazdi (2015), who also noted that definitions of non-adoption varied widely among studies.

Furthermore, the measurement of AM adoption has shown similar inconsistencies. While some studies treat ABC as a singular practice, others break it down into levels of activity management, such as Activity Analysis (AA), Activity-Cost Analysis (ACA), and ABC itself (Baird, 2007; Gosselin, 1997). This differentiation is crucial, as AM relates to the practical implementation of ABC for cost management rather than solely identifying costs.

The research by Baird et al. (2004) indicates a high adoption rate of all three levels of AM in the private sector, yet Baird (2007) found similar usage in the public sector, raising questions about the generalisability of these results across different contexts. The varying adoption rates and definitions across studies suggest that the diffusion process of ABC and AM is not homogeneous, leading to mixed findings.

Given inconsistencies in definitions and classifications of ABC and AM adoption, this study examines whether formal ABC status and activity-management intensity tell the same empirical story. The focus is therefore on measurement alignment: a firm may not identify itself as a formal ABC adopter while still using activity analysis, activity-cost analysis, or ABC-type cost information for selected purposes.

A clearer distinction between organisational adoption status and the intensity of related practices is important for interpreting and comparing survey evidence. Treating these measures as interchangeable can inflate or depress reported adoption rates and obscure partial or selective use of activity-oriented techniques.

The study therefore uses the two measures side by side and interprets discrepancies as evidence of measurement sensitivity rather than as proof of a sequential implementation theory.

3. Methodology

This study uses a descriptive quantitative survey design to compare formal ABC-adoption status with reported intensity of activity-management practices. The design is intended to identify distributional and cross-group patterns rather than estimate causal effects or test a structural theory.

A structured questionnaire was used to collect data from UK non-manufacturing companies. The survey format enabled consistent measurement of ABC status and AA, ACA, and ABC-level use across a broad organisational sample (Collis & Hussey, 2014; Bryman & Bell, 2015).

Designed to enhance response rates and data quality, the questionnaire adhered to best practices (Dillman et al., 2014; Saunders et al., 2016). Key elements included conciseness, with a limit of eight pages to facilitate higher engagement; a booklet format printed on A3 pages, providing clear definitions of terms; and primarily closed-ended questions utilising a five-point Likert scale to assess various aspects of ABC and AM. Additionally, pilot testing was conducted with practitioners and academics to refine question clarity and wording.

The intended respondents were senior accounting or finance personnel with sufficient knowledge of the business unit’s costing practices. Respondents were offered mail and online completion modes. Each questionnaire was accompanied by information explaining the research purpose, confidentiality, voluntary participation, and researcher contact details.

The target population comprised medium-to-large UK non-manufacturing companies with at least 50 employees and £20 million in sales revenue. Firms were identified from the FAME database using these size criteria and non-manufacturing industry classifications. The sampling frame contained 2000 eligible companies. The industry classification rule excluded manufacturing firms so that the study focused on service and other non-manufacturing settings, where costing-system design has received comparatively less attention (Abdel-Kader & Luther, 2008; Al-Omiri & Drury, 2007).

The questionnaire was distributed to the 2000 firms in three mailing stages, with follow-up reminders to non-respondents; 78 firms requested the online version. Fieldwork produced 219 completed questionnaires (10.95%). The analyses reported in Tables 1-10 are based on 204 observations. The available study records used for this manuscript do not establish the disposition of the remaining 15 completed questionnaires; accordingly, no unsupported reason for their exclusion is assigned here. This discrepancy is acknowledged as a limitation and should be resolved from the original item-level dataset if it becomes available. The available records also do not contain an early-versus-late respondent comparison or another non-response-bias test; therefore, no such result is claimed, and possible non-response bias remains a limitation when generalising to the full population.

The survey questionnaire included two primary questions addressing Activity-Based Costing (ABC) and Activity Management (AM). The first question assessed the business unit’s experience with ABC through a multiple-choice format, allowing respondents to select one option from a list of nine possibilities. This measurement was adapted from Brierley (2011) and included the following options: 1) have adopted ABC, 2) intending to adopt ABC, 3) investigating whether to adopt ABC, 4) intending to investigate adopting ABC in the near future, 5) adopted ABC and subsequently abandoned it, 6) investigated whether to adopt ABC and decided to reject it, 7) have considered adopting ABC but did not investigate and decided to reject it, 8) never considered adopting ABC, 9) other, please specify. Respondents selecting the first option indicate full implementation of ABC, while those choosing options two to nine may reflect varying levels of engagement with Activity Management, which will be further explored in the subsequent question. The options from two to nine were derived from various studies by Clarke and Mullins (2001), Cohen et al. (2005), Elhamma and Fei (2013), and Ittner et al. (2002). Focusing solely on the fourth construct of ABC adoption allows for consistent sample sizes across the research models, eliminating discrepancies in sample representation.

The second question of the survey aimed to measure AM usage, including AA, ACA and ABC. For AA, the survey defined it as the process by which a business unit identifies and analyses the various activities involved in providing services, without recording their associated costs. For ACA, the definition provided was that a business unit identifies, analyses, and records the costs of the various activities involved in providing services but does not use this cost information to calculate the cost of each service provided. Lastly, for ABC, the definition specified that a business unit identifies, analyses, and records the costs of the various activities involved in providing services, and subsequently uses this cost information to calculate the cost of each service provided. These definitions were adapted from Baird (2007), Baird et al. (2004), and Gosselin (1997) to ensure clarity and precision in measuring the variables related to AM usage.

The analysis is descriptive. Frequencies, percentages, and cross-group response distributions are used to compare formal ABC status with reported AA, ACA, and ABC-level use. SPSS was used to produce the descriptive statistics. The propositions are evaluated from directional patterns in the observed distributions. Because the earlier AA and ACA expectations encompassed every Likert response category, those categories are reported directly rather than converted into a match percentage.

4. The Research Results

Among the 204 usable cases, 43 firms (21.08%) reported formal ABC adoption. The remaining firms were distributed across intending to adopt (7.84%), currently investigating adoption (9.31%), intending to investigate in the near future (8.82%), previously adopted and abandoned (7.84%), investigated and rejected (5.39%), considered but did not investigate and rejected (8.33%), and never considered ABC (31.37%). Table 1 reports these eight status categories. The denominator is therefore 204, not the 219 completed questionnaires received during fieldwork.

Table 1. ABC adoption in UK non-manufacturing companies.

Using the second approach, this section discusses findings that relate to AM levels rather than ABC adoption and non-adoption. As shown in Table 2, of the business units that use each level of AM, Baird et al. (2004) combined response points “3 = medium extent”, “4 = large extent”, and “5 = very large extent” to present the use of three levels of AM. However, the current research includes a small extent under the use of each level of AM. Business units that use each level to either a small, medium, large, or very large extent are 59.80% for AA, 66.67% for ACA, and 59.31% for ABC. In addition, Table 2 presents the percentage of respondents who reported that they did not adopt any level of AM (40.20% for AA, 33.33% for ACA, and 40.69% for ABC).

Table 2. AM level in UK non-manufacturing companies.

The adoption rate for ABC (the third level) indicates a much higher result than in the first approach (ABC adoption). According to Table 2, 59.31% (22.06% + 8.82% + 10.29% + 18.14%) of UK non-manufacturing companies have adopted ABC (which includes small, medium, large, and very large extent adoption of ABC). This figure is more than double that of the companies that adopted ABC in the ABC adoption question. A comparison between the ABC adoption rates reported in the two approaches (“ABC adoption” and “AM levels”) for UK non-manufacturing companies can help explain some variations in ABC adoption rates in the literature. According to the findings, a substantial share of the variation in adoption rates of ABC in the UK non-manufacturing companies is due to the implementation of the two different approaches (“ABC adoption” and “AM levels”). These findings may provide some insights into how and why some literature found a high ABC adoption rate, especially when they used the AM levels. The reason for this variation may be the possibility that these companies adopted ABC for different purposes, such as budgeting and control rather than costing, and they have not adopted it permanently.

The adoption rate of ABC (under the AM approach) was 47.82% in Canadian manufacturing companies (Gosselin, 1997), 78.1% in Australian manufacturing and non-manufacturing private companies (Baird et al., 2004), 66.3% in Australian manufacturing and non-manufacturing public companies (Baird, 2007), and 56% in Australian companies (Chenhall & Langfield-Smith, 1998). To interpret this finding related to the variation between the two approaches (the ABC adoption and AM levels), first, some companies (e.g., those intending to adopt ABC) may answer that they have used the third level of AM to a large or very large extent; however, they do not choose option one (i.e., adopted ABC) under the first approach (ABC adoption). This may be the case if some companies have started practising the first levels of AM (AA and ACA), which are considered the first and second levels of ABC adoption under the AM levels but would not classify them as ABC adopters under the experience of ABC. Second, in the case of reporting higher ABC adoption rates when using AM levels, it is also possible that the higher ABC adoption rates in the AM levels question may stem from a misunderstanding of the participants regarding ABC as a whole concept.

4.1. Companies Have Adopted ABC

The 43 formal ABC adopters provide the clearest benchmark for P1. As Table 3 shows, 41 of 43 adopters (95.35%) reported ABC-level use to a large or very large extent. AA and ACA responses remain dispersed, which is consistent with treating these practices as overlapping rather than requiring low AA/ACA scores once ABC is implemented. The result therefore supports the descriptive expectation that formal adopters exhibit strong ABC-level use without imposing a sequential-stage interpretation.

The substantive descriptive evidence is the concentration of formal adopters at the upper end of ABC-level use. The AA and ACA distributions are presented in full to show the variation in activity-oriented practices. These patterns are descriptive and should not be interpreted as causal relationships.

Table 3. The relationship between companies that have adopted ABC and AM levels.

4.2. Companies Intending to Adopt ABC

Among the 16 firms intending to adopt ABC, AA and ACA responses span the full scale, while eight firms (50.00%) reported ABC-level use as not at all or small extent. The remaining firms reported medium-to-very-large ABC-level use, showing that stated adoption status and practice intensity are not equivalent measures. This divergence is directly relevant to the paper’s measurement argument.

Table 4 reports the full AA, ACA, and ABC-level response distributions for firms intending to adopt ABC. The ABC-level distribution indicates partial activity-based practice among some firms that do not yet identify themselves as formal adopters.

Table 4. The relationship between companies that have intended to adopt ABC and AM levels.

4.3. Companies Currently Investigating Whether to Adopt ABC

Among the 19 firms currently investigating ABC, AA and ACA are widely used, with the modal response at “large extent” for both practices. Twelve firms (63.16%) reported ABC-level use as not at all or small extent, while seven reported medium-to-very-large use. This pattern is broadly consistent with an investigation-stage group showing less intensive ABC-level use than formal adopters, while also illustrating overlap between adoption status and practice.

The revised interpretation does not claim 100% confirmation for AA or ACA. Their distributions are presented as descriptive evidence of activity-oriented engagement, while the ABC-level responses provide the more discriminating comparison with formal adopters.

Table 5. The relationship between companies currently investigating whether to adopt ABC and AM levels.

4.4. Companies Intending to Investigate Whether to Adopt ABC in the Near Future

Eighteen firms (8.82%) intended to investigate ABC in the near future. Their AA and ACA responses span the five-point scale. For ABC-level use, 8 of 18 firms (44.44%) selected “not at all”, 6 (33.33%) selected “small extent”, and 4 (22.22%) selected medium or large extent.

For ABC-level use, 8 of the 18 firms (44.44%) reported “not at all”, while 6 (33.33%) reported a small extent and 4 (22.22%) reported a medium or large extent. The AA and ACA distributions are reported directly rather than being converted into a match statistic, allowing the observed variation across response categories to remain visible.

The distribution suggests that some firms may use selected activity-based practices before describing themselves as formal ABC adopters. Because the survey is cross-sectional and descriptive, the data cannot establish whether these practices represent a temporal implementation sequence.

Table 6. The relationship between companies intending to investigate whether to adopt ABC in the near future and AM levels.

4.5. Companies Adopted ABC and Decided Subsequently to Abandon It

Sixteen firms (7.84%) reported that they had adopted ABC and subsequently abandoned it. AA and ACA use remains substantial in this group, whereas ABC-level use is concentrated at the bottom of the scale: eight firms reported “not at all” and eight “small extent”. This residual use of AA/ACA alongside low current ABC-level use is consistent with P3 and suggests that activity-oriented practices may persist after formal ABC abandonment.

The observed AA and ACA distributions are reported directly, while the ABC-level responses provide a discriminating basis for comparing this group with formal adopters.

Table 7. The relationship between companies that have adopted ABC and decided subsequently to abandon it and AM levels.

4.6. Companies Investigated Whether to Adopt ABC and Decided to Reject It

Eleven firms (5.39%) investigated ABC and decided to reject it. Five (45.45%) reported no ABC-level use, four reported a small extent, and two a medium extent. AA and ACA responses are presented descriptively and are not classified as 100% matches. The presence of some activity-oriented practices despite rejection again demonstrates that formal status and practice intensity capture different aspects of ABC-related behaviour.

Compared with formal adopters, this group is concentrated at much lower levels of ABC use, which is directionally consistent with P1 and P2.

Table 8. The relationship between companies that investigated whether to adopt ABC and decided to reject it and AM levels.

4.7. Companies Have Considered Whether to Adopt ABC, but Did Not Investigate It and Decided to Reject It

Seventeen firms (8.33%) reported that they had considered ABC, did not investigate it, and decided to reject it. For ACA and ABC-level use, all 17 cases provide responses. The AA row contains 13 valid responses; four cases have missing AA values, so the AA percentages in Table.9 are correctly based on valid n = 13 rather than 17. This distinction is now stated explicitly.

For ABC-level use, eight of 17 firms (47.06%) selected “not at all”, six selected “small extent”, and three selected medium or large extent. The lower ABC-level intensity relative to formal adopters is directionally consistent with the measurement-based propositions.

Table 9. The relationship between companies that have considered whether to adopt ABC, but did not investigate it and decided to reject it, and AM levels.

Note: AA has 13 valid responses in this subgroup; four of the 17 firms have missing AA responses. Percentages for AA are based on valid n = 13.

4.8. Companies Never Considered Whether to Adopt ABC

Among the 64 firms that had never considered ABC, 46 (71.88%) reported no ABC-level use, while 18 reported some degree of use. AA and ACA also show selected activity-oriented practices. Rather than attributing all discrepancies to respondent misunderstanding, the revised interpretation treats them as evidence that formal adoption status and practice-based measures are not interchangeable. This group shows the lowest concentration of ABC-level use and is directionally consistent with P2.

Table 10. The relationship between companies that have never considered whether to adopt ABC and AM levels.

5. Conclusion

This study provides a descriptive assessment of whether formal ABC-adoption status aligns with reported intensity of Activity Analysis (AA), Activity-Cost Analysis (ACA), and ABC-level use in UK non-manufacturing firms. The central finding is a substantial measurement gap: 21.08% of the 204 usable cases identify as formal ABC adopters, whereas 59.31% report ABC-level use to at least a small extent.

This difference should not be interpreted as evidence that one measure is correct and the other incorrect. Instead, the two approaches capture different dimensions of organisational practice. A binary status measure identifies whether a firm regards itself as an ABC adopter, whereas the activity-management measure captures the reported intensity of related practices. Firms may therefore use selected ABC-related practices without classifying themselves as formal adopters.

AA, ACA, and ABC are treated as related practices that may coexist rather than as mutually exclusive stages. Their coexistence within firms suggests that activity-oriented practices may overlap across departments, purposes, and periods of use. Formal adopters nevertheless show a strong concentration at high ABC-level use, while rejection and non-consideration groups generally show lower ABC-level intensity.

The study’s main contribution is therefore measurement-related. Reported ABC prevalence is sensitive to how adoption is operationalised, and comparisons across studies should distinguish formal adoption status from the extent of activity-based practices. This distinction offers a plausible empirical explanation for some of the variation in ABC adoption rates reported in prior survey research.

Several limitations qualify the findings. The design is cross-sectional and descriptive; the reported analyses use 204 observations from 219 completed questionnaires, and possible non-response bias cannot be ruled out because the available records do not contain a non-response-bias comparison. In addition, the available study records do not establish the disposition of the 15 completed questionnaires not represented in the reported analyses. Future research could replicate the measurement comparison with complete item-level datasets, test group differences inferentially, and use interviews to examine why organisations report practice intensity that differs from their formal adoption status.

Overall, the findings support greater precision in ABC survey measurement: researchers should specify whether they are measuring formal organisational adoption, prior experience, implementation status, or the intensity of activity-based practices, rather than treating these constructs as interchangeable.

Conflicts of Interest

The authors declare no conflicts of interest regarding the publication of this paper.

References