Implementing a performance management system in a public organisation: barriers and outcomes

The complexities associated with the public sector have long attracted the attention of academics (Lapsley and Skaerbaek, 2012). One issue that has been the subject of much debate regarding public sector organisations is the implementation of performance measurement and management (PMM) systems, as considerable advancements have been made in this area in recent decades (Steccolini, 2019) and there is still the potential for further exploration (Steccolini et al., 2020). Public sector organisations have extremely diverse functions, large-scale operations (Lapsley, 1988), a wide variety of scopes (Arnaboldi et al., 2015), and a long-standing reputation for being inefficient (Modell, 2005). All of these characteristics, alongside the rise of New Public Management (NPM) principles (see Hood, 1991), make these organisations a fruitful realm for exploring the design and implementation of PMM systems, as well as any resistance to their introduction. Additionally, as stated by Steccolini (2019, p. 256), “public sector (reforms) and NPM were increasingly seen as ideal settings where scholars could explore the power of accounting to lead change (or how accounting change can be resisted …)”. The NPM paradigm is a reform process characterised by the mimicry of private practices and rationality on the design of public services (Lapsley, 2017), and despite being just one of the three most prevalent models in public management reform movements (Krogh and Triantafillou, 2024), it has opened the door for a golden age of accounting research in the public sector (Steccolini, 2019).

The NPM paradigm has led to the introduction of performance models, which are more goal-oriented and multidimensional (Modell, 2004; Arnaboldi et al., 2015), and as a result, the challenges associated with the management of change have attracted the attention of researchers (Modell, 2007). One of the performance models is the balanced scorecard (BSC), which was quickly translated and considered attractive because it focuses beyond the financials of an organisation (Arnaboldi et al., 2015). Its implementation comes with a number of unique challenges (Sharma and Gadenne, 2011), and numerous studies have focused on these (Sutheewasinnon et al., 2016; Bracci et al., 2017; Røge and Lennon, 2018; Zawawi and Hoque, 2020, 2022; Bertz and Quinn, 2022). Notwithstanding these challenges, management accounting change and the implementation of tools like the BSC remain one of the most challenging endeavours facing both researchers and practitioners. As Zawawi and Hoque (2020, p. 68) recognised, “there is still little research into issues of how and why the BSC has been diffused within public sector organisations generally”. Thus, the implementation of BSC remains an area that requires more research, and the aim of this study is to contribute to it.

The present paper reports on the adaptation and implementation of a BSC-inspired model in a public organisation in Portugal. It examines how the new model has been integrated into existing practices (Research Question 1), as well as how managers have coped with its implementation (Research Question 2) and what factors were behind their response to the new system (Research Question 3). In light of these research aims, the theoretical background of institutional theory and its understanding of the responses to management accounting change, resistance and the perceived consequences of its implementation were embraced. This research adopted a single case study approach (Yin, 2018), in an interpretative research tradition (Ryan et al., 2002), at an independent administrative organisation in the Portuguese public sector, which we have called organisation “Alpha” to preserve its identity.

The results of the Alpha case study presented here contribute to both theory and practice. Our findings add new material to the literature on the BSC and PMM that shows how the BSC was integrated and responded to the requirements of the previously existing mandatory practices. The introduction of the model was never intended to replace these practices, and in that sense, the management control division was able to adequately integrate the BSC with the other practices. At the same time, among most of the managers, there was a lack of integration of the newer model into the decision-making processes. This seems to have stemmed from a number of barriers and difficulties that led to resistance to the use of this newer model. This set of barriers was connected to organisational rigidity, education, training, a lack of involvement on the part of the board of directors, a lack of connection to a reward system, time consumption and data gathering. As this is a less successful example of using the BSC in public organisations, this study mainly adds to the literature by reflecting on the discouraging results.

The situation reported for the different groups of actors seems to have led to both stability and change, thus allowing us to contribute to the management accounting change literature. The interplay of change with the integration of the BSC into other practices, as well as the stability of older practices enacted by managers due to the aforementioned resistance, revealed a form of decoupling – a notion that is neither clear-cut nor monolithic (Leca and Laguecir, 2023), where the conditions in which it occurs, and the means by which it happens, need to be understood (Scott, 2014). The results presented show that decoupling occurred as a result of the emergence of some barriers to the implementation of the BSC model.

The study is also helpful to practitioners dealing with the implementation of BSC in their own organisations. Knowing the success and failure factors regarding the implementation of the BSC is essential for increasing its uptake in public organisations (Northcott and Ma'amora Taulapapa, 2012). As comes from the situation encountered, in the face of rigid structures, or insufficient support from top management, practitioners should emphasise the importance of the new implementation, their innerworkings through training, search superior sponsorship and work to design a model that is both efficient and less time-consuming. This is also a way for the managers not to go back to their old practices and embrace change.

The remainder of the paper is organised as follows. Section 2 presents a review of the main literature on the balanced scorecard for public organisations and the principal work on the theoretical background of the study. Section 3 describes the research methods and the procedures used for collecting data, as well as briefly introducing the case study organisation. Section 4 starts by explaining the fundamentals of the implementation of the BSC and its integration within other practices, before revealing the barriers and difficulties involved. Section 5 discusses the results and Section 6 provides concluding remarks.

Since NPM put the emphasis on performance (Krogh and Triantafillou, 2024), a vast body of research has explored the effects of PMM in public organisations (Steccolini, 2019). Indeed, in the midst of the NPM phenomenon, public managers have been pressured to drive public entities toward a more results-oriented operationalisation of their activities (Arnaboldi et al., 2015), probably due to the reputation of public organisations as being inefficient and wasteful (Modell, 2005) and dominated by bureaucracy (Lapsley and Skaerbaek, 2012). Moving to a more managerial focus, public organisations have also been asked to demonstrate how they create value for their stakeholders (Kaplan and Norton, 2004) in their range of functions (Lapsley, 1988). Emphasis has been placed on performance measurement (Osborne, 2010; Krogh and Triantafillou, 2024), and models and different practices that were initially developed for private and profit-oriented organisations started to be introduced and proclaimed (Modell, 2005; Arnaboldi et al., 2015). One of the models that was implemented in public organisations and explored in the literature was the BSC (Greatbanks and Tapp, 2007; Kasperskaya, 2008; Sharma and Gadenne, 2011; Northcott and Ma'amora Taulapapa, 2012; Dreveton, 2013; Zawawi and Hoque, 2020, 2022).

The BSC offers managers “a comprehensive framework that translates a company's strategic objectives into a coherent set of performance measures” (Kaplan and Norton, 1993, pp. 134–135). Therefore, the BSC represents a multidimensional model with different objectives (Northcott and Ma'amora Taulapapa, 2012) that enable the execution and management of a strategy (Kaplan and Norton, 2001, 2004) through the lens of financial and non-financial performance measures. Derived from strategy, the measures are organised into four perspectives: financial, client, internal processes and learning and growth (Kaplan and Norton, 1992, 1993, 1996, 2001, 2004). Within these perspectives, there are causal relations among the diverse objectives (Kaplan and Norton, 1996; Hoque, 2014; see also the results in Kober and Northcott (2021) for an empirical analysis of the causal relationships). In fact, as portrayed by Hoque (2014), the BSC was developed through a series of publications by Kaplan and Norton, whereby features were added and ideas were matured. Today, the BSC encompasses a richer model, with a holistic view that can be customised as needed (Hoque, 2014), tailoring the process to different strategies (Modell, 2005). Evolving from the four perspectives, Kaplan and Norton also proposed the development of a strategy map as a way to add another layer of detail focusing on the strategy components and their links (Kaplan and Norton, 2004).

Transposing the BSC to public organisations is challenging because these organisations are accountable to citizens rather than shareholders (Sharma and Gadenne, 2011). By promising to deal with both excessive performance indicators and the reliance on financial control (Modell, 2005), the model has evolved from a shareholder focus to more of a stakeholder focus, thus reducing the importance given to the financial perspective (Modell, 2004, 2009a). Financial success is not the main objective of public organisations, so a rearrangement of the model was proposed (Kaplan and Norton, 2001; Niven, 2008), and this adaptation, while keeping the strategy at the core of the model, raises the customer perspective and the mission to a higher position in the hierarchy (Kaplan and Norton, 2001, 2004; Niven, 2008). The financial features, therefore, are downplayed. Examples of this rearrangement and adopters of the BSC are also reported in the seminal literature on the BSC. Kaplan and Norton (2001) refer to implementations in the US city of Charlotte, North Carolina and the Australian cities of Brisbane, Cockburn and Melville. Kaplan and Norton (2004) further added examples of its application in the Royal Canadian Mounted Police and the UK Ministry of Defence.

Naturally, research has reported a wide range of differing situations regarding the implementation of the BSC. From studying the results of integrating the BSC with other tools (Modell, 2009a; Pimentel and Major, 2014) to analysing the implementation process and/or the interactions that such implementations trigger (Umashev and Willet, 2008; Modell, 2009a; Sutheewasinnon et al., 2016; Bracci et al., 2017; Røge and Lennon, 2018), the conclusions are diverse. Research has found that success factors include previous participative decision processes, management support, adequate resourcing of initiatives, continuous learning and training, and customisation of the model (Northcott and Ma'amora Taulapapa, 2012). Although the factors that underline either the success or failure of the implementation of the BSC are interdependent (Umashev and Willet, 2008), it is worthwhile pinpointing difficulties in the customisation of the model (Northcott and Ma'amora Taulapapa, 2012) or in the information systems to support their ongoing use (McAdam and Walker, 2003; Sharma and Gadenne, 2011; Maran et al., 2018). Table 1 presents a non-exhaustive summary of the factors that are deemed to be important.

Some studies in the public sector, for example, have reported on the use of the BSC in helping with changes to situated rationality (Bertz and Quinn, 2022), or in assisting with the alignment of purchaser–provider interest (Zawawi and Hoque, 2022), highlighting the need to review strategy (Dreveton, 2013) or even the perception of its implementation as being imperative to the internalisation of greater attention to long-term goals (Modell, 2012). Additionally, from situations that seem to be fully integrated in the practices of organisations (Pimentel and Major, 2014; Bracci et al., 2017; Zawawi and Hoque, 2020; Bertz and Quinn, 2022), other cases have been reported in which the model has become simply ceremonial (Kasperskaya, 2008), adapted or abandoned completely (Modell, 2009a; Sutheewasinnon et al., 2016; Bracci et al., 2017), or that helped in promoting a collective involvement of managers in a divergent change process (Pimentel et al., 2023).

Institutional theory has long been a common paradigm to which research in management accounting has turned (e.g. Siti-Nabiha and Scapens, 2005; Nor-Aziah and Scapens, 2007; Kasperskaya, 2008; van Hengel et al., 2014; Bertz and Quinn, 2022; see also Barros (2025), who shows that institutional theory has been the most widely used theoretical perspective for informing research on the BSC and government entities). Indeed, institutional theory is well equipped to tackle aspects related to the implementation of, and organisational responses to, change processes. Therefore, the literature on the BSC and public organisations, and the theoretical body on management accounting change driven by institutional theory, seems to have overlapped (Kasperskaya, 2008; James, 2009; Modell, 2012; Sutheewasinnon et al., 2016; Rautiainen et al., 2017).

With the emergence of new techniques in the field of management accounting, researchers have become interested in studying the challenges around processes of change (Modell, 2007); the results of change have been addressed by many authors (e.g. Liguori and Steccolini, 2012). Indeed, at the very beginning of their paper, Burns and Scapens (2000) mentioned that management accounting change was already, at the time of writing, an area that attracted much debate. In order to understand the change processes in the context of management accounting, the field has often drawn on both old institutional economics and new institutionalisms (Lukka, 2007; Nor-Aziah and Scapens, 2007; Liguori and Steccolini, 2012).

Institutional theory adopts the view that practices (management accounting included) are shaped by the variety of unique factors related to each organisation, which means that in order to understand them it is necessary to study the “mish-mash of interrelated influences” (Scapens, 2006, p. 11) within the complexity of organisational life (Burns et al., 2003). Changes in organisations undeniably involve important behavioural and cultural issues that cannot be set aside and should be fully understood (Burns et al., 2003). Institutional theory helps in understanding this (Scapens, 2006), allowing researchers to gain insights into matters related to resistance to change, stability, or even power struggles (Modell, 2007). In other words, institutional theory provides a helpful theoretical background to assist researchers with the processes of institutionalisation, recognising that the evolution of practices should be understood by considering broader meaning systems and complex socio-political processes (Model, 2022). Using an institutional approach, it is evident that a complex relationship exists between institutionalised structures and different actors in the continuous transformation process of PMM (Modell, 2005).

In the study of management accounting change, Burns and Scapens (2000) developed an influential framework rooted in institutional theory with a view to interpreting and understanding it as an ongoing process that might encounter resistance (Burns and Scapens, 2000; Burns et al., 2003). This framework recognises the existence of both stability and change, attributing the roles of rules and routines to management accounting and control practices (Burns and Scapens, 2000). Burns and Scapens (2000) explained that, through encoding, institutional principles become rules and routines, which are then enacted by individuals and, in a third phase, the routines are reproduced. Finally, in a fourth phase, these rules and routines are institutionalised. Burns and Scapens (2000, p. 10) further explained that the “enactment of rules and routines may be subject to resistance, especially if the rules and routines challenge existing meanings and values …”

This initial model, developed by Burns and Scapens, was improved to accommodate criticism of the original framework (ter Bogt and Scapens, 2019) that started to appear in later works (e.g. Bertz and Quinn, 2022; Barros and Ferreira, 2023). Ter Bogt and Scapens (2019) expressed the opinion that their framework allows management accounting to be studied in a situated manner. The authors introduced new ideas but recognised that practices are shaped by both broader and internal institutions (ter Bogt and Scapens, 2019). Moreover, by introducing the notion of situated rationalities, the extended framework makes it possible to study how these institutions affect internal practices; however, the framework still recognises that change occurs and that resistance may arise. In fact, Burns and Scapens (2000) had already recognised that formal change processes could give rise to resistance that may lead to a failed implementation. Reporting on a previous case study, they observed that such resistance may appear due to competing interests, insufficient knowledge and experience to deal with the change, and the fact that other ways of thinking may already be embodied in the organisation's institutions. The factors that drive or hinder the change processes may vary, and an approach to these factors, which represent forces in the change process, was proposed by Kasurinen (2002; see also Pimentel and Major (2010), who based their analysis on this framework). Kasurinen (2002) recognised some of the barriers to change and classified them into three groups: confusers, frustrators and delayers. The first group occurs when different viewpoints and uncertainty about the role of the changes emerge (Kasurinen, 2002). The second group refers to organisational aspects, such as existing cultural barriers (Kasurinen, 2002; Modell, 2007). The third group is related to more technical and temporary issues (Kasurinen, 2002).

In more general terms, a wide range of effects may appear in any process of change. For instance, Liguori and Steccolini (2012) reported on one case in which, although two public organisations were subject to the same environmental pressures, two different patterns of change were observed.

Additionally, change processes can become largely ceremonial (Burns and Scapens, 2000), such as the situation that Kasperskaya (2008) encountered. And Siti-Nabiha and Scapens (2005) found that stability and change occurred in a process of change and, in some ways, the implementation ended up being rather ceremonial. Nor-Aziah and Scapens (2007) further explained that, in the ceremonial use of accounting, the intention is to show apparent compliance, but the change has no impact on day-to-day activities. In light of such cases, the literature introduced the concepts of loose coupling and decoupling (Siti-Nabiha and Scapens, 2005; Lukka, 2007; Nor-Aziah and Scapens, 2007; Kasperskaya, 2008; Bromley and Powell, 2012; van Hengel et al., 2014; Vakulenko et al., 2020). These concepts are taken from the organisational literature, particularly the work of Orton and Weick (1990), who described loose coupling as interdependent elements that are able to encompass some determinacy while at the same time possessing independence and indeterminacy.

Adding to this, Nor-Aziah and Scapens (2007) showed, with their case study, that loose coupling did not occur as an automatic response, and the resistance that led to the loose coupling of budgets was not simply due to powerful ways of thinking. In explaining the treatment that researchers give to loose coupling, Modell (2009b) pointed to three situations: a display that is more symbolic and exists for legitimacy reasons; the separation of specific performance indicators or practices from the goals; reconciliation of clashing demands or satisfying different instrumental functions. Van Hengel et al. (2014) further distinguished between vertical and horizontal coupling, the former being between hierarchical levels, and the latter referring to the “loose coupling between intentions and actions” (51). Thereafter, decoupling continued to be investigated and various notions emerged, including reverse coupling (in which a mismatch between the ideas of the reformers and practices appears) (Vakulenko et al., 2020) and means-ends decoupling (in which the practice is implemented but the initial goals are not achieved) (Bromley and Powell, 2012).

The present study followed a single case study approach at Alpha, an independent administrative organisation in the Portuguese public sector, to explore the implementation of the BSC in that organisation. More specifically, we sought to understand how the process of implementing the BSC unfolded in terms of its integration with existing practices, particularly those related to management control processes (Research Question 1), how the managers coped with the introduction of the BSC (Research Question 2) and what factors were behind the managers' response (Research Question 3). These represent the three research questions that guided our analysis.

When studying the use of the BSC in public organisations, a case study is widely employed as the research approach (Barros, 2025). This is probably due to the opportunity it provides for the researcher to take the context into consideration and understand how an accounting practice is being used (Ryan et al., 2002). Case studies allow phenomena to be considered in depth, holistically, and in their real-life settings, and they are particularly useful when the phenomenon being analysed is related to the setting in which it happens (Adams et al., 2006; Yin, 2018). Our single case study approach has allowed us to undertake the desired empirical analysis of the implementation process and the results of adopting the BSC in the organisation studied. It represented a very particular case that provided a good illustration of some of the implementation issues that may arise. Therefore, according to the typology of case studies presented by Scapens (1990) and Ryan et al. (2002), it primarily exhibits the characteristics of an explanatory case study.

Data collection took place in December 2019 from different sources, including formal interviews, direct observations and documentation, thus ensuring both the triangulation of data and an in-depth analysis (Yin, 2018). At the time this research was conducted, it was already possible to see the impacts of implementation and the way that it was being used. After the aforementioned period, and in order to write this case report, additional informal contacts were made with an employee to complement certain aspects of the implementation process that were not clear to the researchers. This employee was a valuable informant, in the sense that they had spent some time in the management control division, where they were able to analyse the overall practices within the organisation and collaborate in the development of some of them.

The main sources of evidence were 12 formal interviews (Table 2), which were conducted in person at the headquarters of the case organisation by one of the researchers involved in this work. The interview strategy encompassed having a dialogue with the manager of each department located at the headquarters (in this case, 10 of the 14 departments), and two of the controllers who worked more closely with the BSC were also interviewed (making a total of 12 interviews). By adopting this approach, it was possible to capture the views and specificities of the managers who use the BSC and the controllers who have a broader view. A script for the interviews was prepared, but the interviewees were encouraged to talk about aspects that were not covered in that script. The interviews, during which the researcher took notes and recorded impressions and observations, lasted between 12 and 102 min, with an average duration of 36 min. Of the 12 interviews conducted, eight were tape-recorded and transcribed by the interviewer on the same (or following) day; however, the first four interviews were not recorded since they were integrated into the meetings that took place to establish the structure and the KPIs for the following year (Table 3). Given the sensitivity of the information discussed in these interviews, the choice was made not to record them; however, notes of the observations and impressions of the researcher were taken, organised and registered, in accordance with the proposed questions in the interview guide, as soon as possible. This approach enabled further analysis of the data and presented the need for context when describing the actions and characteristics of the internal practices. It was also beneficial that the same researcher maintained close contact with the organisation, enabling more direct engagement and an in-depth understanding of the functioning of these practices.

To protect the identity of the managers and controllers, the interviewees are only identified by a number.

In addition to this, to complement the previously collected data and observations, formal reports were consulted, including documentation of the BSC from earlier years, the strategic plan, the annual activity plans, activity reports from earlier years, human resources records and other documentation relevant to performance evaluation. Most of these documents were publicly available on the organisation's website.

The case organisation, Alpha (name anonymised to preserve the identity of the organisation), is a public institute integrated within a Portuguese ministry and involved in the indirect administration of the Portuguese government. It has a long history going back to the Kingdom of Portugal, but it only began operating with its current structure in 2007. Its restructuring process was aimed at adapting the work of the organisation in response to alterations in the environment, due to which the institute had to develop its activity while at the same time eliminating unnecessary formalities by simplifying acts and processes (as stated in the activity report of 2017). The organisation has administrative autonomy, its own assets and the jurisdiction to develop its activities throughout the entire national Portuguese territory.

At the time that this research was conducted, Alpha had a board of directors (BD) comprising three members, a consultative council and a series of departments/areas ^[1]. Overall, in 2018, the organisation had a total of 4,878 employees, 73% of whom were aged between 45 and 59, and 62% had worked for the organisation for 20 years or more (information gathered from a human resources report from 2018). With regard to the higher levels of management, according to information extracted from the social balance report of 2018, about 58% were over the age of 45, while only one was aged between 35 and 39. Additionally, the same report showed that about 79% of them had at least 15 years of service in the organisation and approximately 53% had been in the organisation for 20 years or more. These documents also reported that all of these employees had at least a bachelor's degree.

Within its structure at the time of this research, Alpha included a division specifically related to the management control function. In accordance with legal requirements, in line with a publicly available deliberation that had been emitted in 2012, most of the functions of the management control division related to supporting the top management team. Specifically, within its competences, we found that they had support functions related to the formulation and communication of the strategy used to execute the organisation's mission, support the definition of the mission charter, update the “framework of evaluation and accountability” in line with the ministry, define the annual objectives, elaborate the activity plan, accomplish objectives and implement business intelligence tools, among others. With this wide range of tasks, the management control department is structurally positioned on the organisation's charter to respond to the board of directors – interestingly, above all other departments.

These descriptions of activities confirm the perception that diverse practices exist and the management control division plays an important part in them. Being a public organisation, Alpha is subject to a system of performance evaluation imposed by law, namely the Performance Appraisal System for Public Administration (or the Sistema Integrado de Gestão e Avaliação do Desempenho na Administração Pública – SIADAP 1, in Portuguese). The aim of this system is to improve the performance and quality of services and provide the basis for the main instrument to be presented by a public organisation: the “framework of evaluation and accountability” (hereinafter known as “QUAR” – Quadro de Avaliação e Responsabilização). This document presents the main objectives for the year and indicates how close they are to being accomplished, along with an explanation of the variations that may have occurred (information on the case organisation's activity report for 2018); it is an important framework for evaluating and managing the performance of Alpha. The QUAR includes the strategic objectives that were defined in the mission charter, breaking them down into operational objectives. As described in the annual activity report for the year 2018, there were nine operational objectives measured by 17 indicators, which were monitored and reported to the ministry.

Another legal tool that is required of Alpha is the mission charter, which also serves as a basis for the QUAR. The mission charter is a declaration of purpose and provides guidance for decision-making and for the strategy, mission and main objectives. It is a framework for public administration that allows the board of directors to be evaluated and should be signed by them at the beginning of their tenure. At Alpha, the last mission charter was issued for a period of five years (2019–2023).

Alpha also prepares an annual plan of activities. This plan, as explained on the organisation's website (consulted in April 2024), is a document in which the objectives of the organisation and its projects are set out, including information on the resources that are assigned to accomplish them. This plan of activities serves as a basis for the definition of the QUAR and is also supported by the balanced scorecard model, which allows the organisation to have a more detailed tool than the mission charter and the QUAR required by law.

Apart from the previously described practices, which were mostly defined long ago (the SIADAP legislation was first published in 2004 and obtained its current characteristics in 2007), another model has started to be implemented. SIADAP legislation does not define specific internal systems to be implemented, so each public organisation may apply what its managers consider to be adequate. After collecting the data, it was also determined that the management control division would be given the competences to collaborate in defining a management control model. This also appears to be defined among the competences of the department referred to at the beginning of the previous section.

Therefore, the implementation of a BSC-inspired model was also triggered, around 2016, by the management control division. This was seen as a way to more visibly clarify the mission among its collaborators, recognising its strategic management features, monitoring performance and communication (as reported in the 2016 activity plan). Indeed, all the implemented practices, imposed either by law (activity plan, QUAR, or mission charter) or the BSC, attempt to undeniably create a move towards a more results-oriented organisation. An awareness of this exists within the organisation, as is shown in the annual activity plan for 2016:

The implementation started with the management control division conducting a training programme in all departments. During this training, the BSC was contextualised and the perspectives and goals of the management control system were explained. The management control division retained the responsibility for preparing, implementing and supporting the BSC construction. This internal division also prepares the annual plan of activities and the annual report.

Since the beginning of the implementation, the BSC has been formally integrated with the remaining practices and seems to have contributed to their elaboration. This can be seen in this passage from the annual activity plan of 2017:

Moreover, the BSC was included in the Annual Activities Plan after it had been adopted. The BSC was also updated annually, which involved a process that took at least one month to be concluded. For this reason, it usually started in October and was finished by the end of November. The first step was the definition of the overall strategic objectives. After these had been approved by the board of directors, they were disclosed by the management control division to all of the other departments. Subsequently, an overall objectives diagram, aligning all the departments and the measures leading to each of the objectives, was defined. The management control division met with the manager of each department individually, and two significant moments were observed (the researcher was able to attend some of these meetings): first, a segment dedicated to collecting data about the KPIs to consolidate the final report; and second, a section focused on validating or defining the KPIs for the following year.

In more recent years, the number of KPIs in each department has ranged from 18 to 50, with an average of 25, and the identification and selection of these measures was undertaken by the department, considering its activity. Moreover, each KPI has a sheet on which are listed: the KPI designation, a description, the calculation formula, the person responsible, the person responsible for data collection and monitoring, the data source, the interlocutor to obtain data, the frequency of measuring, the unit for measuring and a disclaimer. These KPI sheets were made public in the annual plan of activities (e.g. in 2016, 2017 and 2018).

At Alpha, the BSC was applied to the different units of the institute (2017 annual plan of activities) in a segmented way. For each of these units, a strategy map was created and internally designated as a process activity diagram (PAD) with a similar structure (this was seen in the various maps we had access to and reported on in the last passage of the 2017 activity plan). In the activity plans for 2016 and 2017, these maps are said to work as a guide and a way to articulate the diverse management documents. Thirteen performance maps were created for 2016, and in 2017, the number evolved to 14 with the addition of another department. Initially, the BSC model was constructed using a common structure with four perspectives, namely: the stakeholder perspective; the financial perspective; the process perspective; and the learning and knowledge perspective (as explained in the plan of activities for 2017). And although not all perspectives were used in every map, as is usual in public organisations (including Alpha), the stakeholder perspective is considered the top perspective and achieves the highest position in the hierarchy (Niven, 2008), while most of the 14 maps created for 2017 exclude the financial perspective. Comparing the annual reports of 2017 and 2018, the financial perspective only appeared for the financial department in 2017, and by 2018 it had been eliminated entirely.

After the process activity diagrams had been defined, a monitoring phase and a reporting phase were meant to start, with the latter being mandatory, as requested by the management control division, to gather all the information available about performance.

Moreover, at the beginning of every year, Alpha presented the annual activities report, which is also connected to the rest of the documents, as the annual activity report of 2018 explains:

Therefore, this annual activity report summarises the performance of the previous year in both quantitative and qualitative terms, including the self-assessment report based on the QUAR, which is a legal obligation.

Certainly, by the time the research reported herein has taken place, the BSC has had time to mature at Alpha. In particular, there had been an initial training period before the implementation, and the management control division was responsible for keeping it running. Nonetheless, in 2019, some of the managers had reported difficulties in taking full advantage of the BSC, with some of them revealing that they did not use the model unless it was absolutely imperative. Variations in the use of the model have emerged, and during the interviews conducted, we noted that only three of the ten managers interviewed were integrating the BSC into their management practices – to a certain extent. Examples of the answers given when asked if they made use of the KPIs included:

Although others revealed some recognition of the merit in its implementation, they showed a reluctance and a lack of interest in the BSC, as can be seen in the next quote:

This lack of interest also appears in a few of the comments made by managers, who mentioned that some of their departments feed the system and monitor some of the results, but only when the management control division asks for the results of the indicators:

There was even some confusion about the periodicity of the monitoring, despite the fact that this, and those responsible for collecting the data, were clearly identified in the KPI sheet, which was publicly shared in the annual activity plans.

Other situations were detected in which the managers had more proactively embraced the use of the BSC and had taken advantage of the components of the model for their day-to-day monitoring of the evolution of the principal variables of their departments, or at least they had an interest in pursuing the necessary changes to be more involved with the practices than the system would permit.

In summary, the implementation of the BSC was somewhat uneven within the different departments inside Alpha, revealing different degrees of the integration of the tool, with some managers being “users” and some being “non-users”. Indeed, a closer look at the BSC non-users shows that the managers do not use it as a tool for strategising or for measuring performance, but rather as an information system to clarify and to communicate what they are doing during the year. This seems to stem from a different set of interrelated factors that might be understood as being more closely related to organisational and technical aspects. When we seek to understand the difficulties related to adopting the model, managers have described some of the obstacles and considerations regarding what may have possibly gone wrong, and these are explored below.

As is common in most public organisations, Alpha has employees who, given their ages and longevity in the organisation, may be expected to adopt a posture of resistance and be less predisposed to change. They have a lot of experience with the processes as they are and are more prone to maintain their ways. Looking at the age of the employees and their contract terms (as of 2018), we see that 73% are aged between 45 and 59, 99% have permanent contracts, and 62% have been with Alpha for 20 years or more. Additionally, the fact that most department managers come from areas other than management (generally from law), and lack knowledge about the inner workings of these kinds of practices, poses further difficulties, as depicted in the comments below:

This may be one of the explanations for the lack of adherence, but the absence of comprehensive and successful training has not helped in this matter. It is important to mention that, at the beginning of the BSC introduction process, there were indications that some of the training conducted was felt, by some managers, to be insufficient or “embryonic”:

Every year, departments were invited to define the KPIs for the following year, in collaboration with the management control division. At the time of this research, all of the departments met in person to discuss their strategy and goals for the coming year, with an initial brief explanation from the management control division's coordinator about the BSC (how it is divided, how it works, the Alpha strategy, and other important information), but this is not enough to deal with the characteristics of the model.

Another issue pertaining to the above-reported situations is the lack of connection between the model outcome and the individual performance evaluation of the managers and employees in the department, as described by one of the managers:

Explaining some of the points made in the previous passage, as with the organisational performance, the evaluation of employee performance in Portugal's public administration also has rules and principles determined by law (another level of the SIADAP), which, at Alpha, seem to be disconnected from the BSC practices, and employees seem to be more focused on the aspects dictated by that evaluation.

Another of the major difficulties that surfaced in the interviews was related to a lack of, or minimal, involvement on the part of the board of directors. As mentioned above, the implementation of the BSC was an initiative implemented by the management control division, and it is this area of responsibility that pushed to continue the project. One of the controllers interviewed (Controller 1) commented that, since the development of the updating process was done by the management control division, the BD felt they did not need to look at it:

All communication about strategy and goals is done through the management control division (after being validated by the BD). For this reason, the departments do not feel any kind of pressure from the BD, and probably not from the top managers either, which could represent a further aspect in explaining the lack of interest in the process.

There were also comments pointing to a lack of time as being a possible obstacle preventing managers from devoting more attention to the BSC. This is reflected in the following passages:

It is also important for the organisation to have more automatic ways of conveying the information, with less need for input and contact from the managers. First, the way in which the information technology (IT) support and the information management were designed is perceived as not being user-friendly. As mentioned above (Section 4.3), feeding the model was the task of the departments and was a very rudimentary process. Excel spreadsheets were the only support tool for the BSC, and it has been mentioned that they are not easy to use. For example, Manager 10 complained that the file was complex and did not permit agility when used. Collecting data was, indeed, a “manual process” (Manager 1) and Alpha does not have the necessary IT support systems to make the process more straightforward and less dependent on the inputs introduced by the managers. Furthermore, some information may even be required from other areas of the organisation, which leads to further delays and adds further complexity to the practices, thus increasing the amount of time consumed. Manager 3 also referred to concerns over the dependence on data from other Alpha sectors.

Set up to explore and accompany the implementation of the BSC in a public organisation, the aim of this study was to examine how the integration of the newly inspired BSC system with the organisation's existing practices unfolded (RQ1), how managers have coped with the implementation (RQ2) and what factors were the forces behind the managers' response (RQ3). In answering these questions, Figure 1 provides a visual synthesis of our results and discussion, interpreting the management accounting change process of the BSC and its integration in the newer practices at Alpha in light of institutional theory and its branches.

In looking at how the new model was integrated into the existing practices, and thus answering the first research question, we start to see that it was the management control division of Alpha that pushed for the implementation and that managed the whole process. Essentially, this is the inside division that represented the driving force behind the whole “BSC project”. With regard to the initial process of change, it is clear from our findings that the organisation, in trying to abide by the commitment to a more efficient management style, especially in terms of the management control area and its manager, decided to implement the BSC. In line with other practices that had already been implemented, the BSC was a way to guarantee clarification of the mission in a visible way. The implementation of the BSC was not imposed by force of law, unlike the contrasting situation in which the previously existing elements had to be prepared.

Notwithstanding this, the results allow us to present an example of the integration of the BSC with the other practices that had already been implemented in the organisation. Lapsley (2022) referred to layers of reforms when he spoke about NPM or NPG, expressing the view that these new practices are not expected to enter a blank field and the older practices will continue. As a corollary, in specific organisations, a change process, such as the introduction of the BSC-inspired model that occurred in the case study organisation, does not happen against a blank canvas. In this sense, Modell (2009a) revealed that previous experimentation with different management techniques influenced the practices that were implemented. In the case of Alpha, previous practices were demonstrated, imposed by force of law, and the introduction of the BSC led to another layer of practices that had to be constructed according to the necessities and requirements of these practices, i.e. they were introduced on top of the existing practices. The existing practices were implemented earlier (the time period and analysis are not covered in this research), and afterwards, it was intended that the change to using the BSC would respond to the needs of the existing practices. Thus, this represents different layers of change with successive implementations occurring, and these were formally integrated by the preparers of the different practices and documents. In fact, for Alpha's management control team, the BSC was a way to prepare and integrate the QUAR into the activity plan, along with the annual activities report.

Therefore, the Alpha case provides a further example of the integration of the BSC with pre-existing practices (extending the examples of Modell (2009a) and Pimentel and Major (2014)). This integration could be undertaken thanks to the commitment of the management control department, which was responsible for preparing all the control documents. It is through the practices and activities of this division that we can clearly see an integration between the previous practices and the newer BSC-inspired model.

On the other hand, successful integration appeared to have only been achieved by the management control division and did not reflect the reality for most of the managers. The following considerations will allow us to ponder the response to Research Question 2. In fact, this apparent integration does not represent one of the most successful examples of the use of the BSC in the public sector, in the sense that the implementation was not used as the theory suggests it should be. It is clear from our results that the majority of the various managers and employees were not keen on the BSC.

First, it is undeniable that the BSC was not used to support decision-making processes by most managers. In fact, most of the managers' activities were decoupled from the BSC and they did not take advantage of day-to-day use of the model (Meyer and Rowan, 1977). Decoupling situations are a complex phenomenon and were inclusively categorised by Leca and Laguacir (2023) as being non-clear-cut notions and non-monolithic. Complexity is involved in this case, and it is also clear that the older practices have continued at Alpha since the introduction of the BSC.

Although the implementation of the BSC does not seem to have evolved into chaotic situations across departments, it has emphasised the stability of the previous practices. Some of the managers still use their previous and specific approaches (their “organisational routines”, according to Burns and Scapens (2000)), and they only interacted with the BSC when it was first defined and when feeding it the information required by the management control department. This reveals some parallels with the study by Siti-Nabiha and Scapens (2005). As these authors found in their case, ours also shows signs of an effective process of change involving coupling of the new model with current practices. However, a clear separation exists regarding the decision-making processes of the managers, who have not fully endorsed the characteristics of the BSC. This is similar to the “passive resistance” described by Kasperskaya (2008), in which managers preferred to cling to the practices that were familiar to them, while recognising the virtues of the BSC. Therefore, a rather ceremonial stance may have been assumed, decoupling the decision-making processes from the BSC. Nevertheless, this was enacted by the management control division, in order to comply with the mandatory practices; in other words, there is a situation concerning both stability and management accounting change that is both successful and unsuccessful (Siti-Nabiha and Scapens, 2005). In contrast to Maran et al. (2018), who described an early disengagement with the BSC and favouring the stability of older practices, our data show that, at least in the realm of the management control division, change existed, but in fact, most of the managers relied heavily on their previous ways.

As regards the factors understood as reasons for the non-use of, or disengagement from, all the BSC model facets (and thus answering the third research question), this seems to be related to a certain resistance stemming from organisational and technical barriers and factors. Factors and barriers resulting from different difficulties felt in the implementation of the BSC are not new (Kasurinen, 2002; Pimentel and Major, 2010; and see Table 1 in this regard). It is noteworthy that the organisation has a human resource structure that is less prone to change, offering some rigidity and cultural resistance, compared to the way in which work is organised. Additionally, managers tended to have educational backgrounds from outside the management area. These two points represent a strong challenge, in which the lack of efficient training in the characteristics and specificities of the BSC, to support its introduction, has not helped in the “buy-in” process, as a lack of knowledge and experience in dealing with systems has long been recognised as a handicap to the processes of change (Burns and Scapens, 2000). Adding to these difficulties/barriers, our case study organisation presented a lack of solid involvement from the board of directors, a time-consuming process of obtaining the information required and connecting it to individual performance, and difficulties in gathering the data. All of these factors played their part and are no more than the visible faces of resistance used by these actors to resist change and couple it to daily activities, ultimately guaranteeing the stability of pre-existing routines in each department. Maran et al. (2018) went further and also highlighted some of these factors as forces of stability.

Together, these aspects further reinforce the problems and recurrent limitations faced by the BSC and expressed in the literature (Holmes et al., 2006; Umashev and Willett, 2008; Sharma and Gadenne, 2011; Bracci et al., 2017; Røge and Lennon, 2018; Maran et al., 2018). Resistance has indeed long been recognised in the literature as the killer of processes of management accounting change (e.g. Burns and Scapens, 2000; Kasurinen, 2002), and it is fair to say that this is not one of the most successful implementations.

Supported by data collected from a single case of a public agency, this study has provided an account of the implementation of the BSC. Public organisations face challenges in implementing the BSC, or a PMM for that matter (Sharma and Gadenne, 2011), implying a very interesting situation for studying the processes of management accounting change. Our case study looked at the implementation of the tool, observing how it was integrated into pre-existing practices (which comprised our first research question), the hurdles encountered, especially regarding how managers have coped with this change situation (the second research question) and what factors underlined their response (the third research question).

Therefore, the intention was that this study would contribute to the challenge of implementing a performance model in a public sector organisation, an area that still requires more research and it is therefore pertinent to both the literature and practice in a number of ways. First, it adds to the literature that addresses performance management, measurement systems and the BSC. The case makes it possible to show that the BSC was implemented on top of the existing practices and was built to respond to their requirements. With this change project, the situation emerged in the division that was responsible for the control practices (and that pushed for the implementation), which was able to integrate them with the older legally mandatory practices in a public service in Portugal; however, in the rest of the departments, some managers resisted incorporating the new practice into their decision-making processes. This non-involvement with the model seems to be justified by a set of barriers, difficulties and resistances. Roughly speaking, these aspects included organisational rigidity, education, training, a lack of involvement on the part of the board of directors, a lack of connection to individual compensation, time consumption and data gathering.

Second, our results allow us to contribute to the literature on management accounting change. In particular, they contribute to exploring the dynamic underlying processes of change that result in both the stability of older practices and change. As described in the previous section, the results point to a situation of change, with adequate integration of the newer model by the department that pushed for the implementation. For the managers' part, stability of the older practices was maintained. Most managers only paid attention to the BSC when feeding it to support the management control division and help it keep the practices running, thereby revealing the interplay between stability and change in a situation where a kind of decoupling had appeared. As Scott (2014) stated, it should not be assumed that decoupling just happens, and the question of when it occurs, and under what conditions, should be studied. Thus, our results demonstrate that the set of barriers and difficulties is at the heart of this form of decoupling, and in doing so, the study contributes to the debate on decoupling (Siti-Nabiha and Scapens, 2005; Kasperskaya, 2008; Vakulenko et al., 2020; Leca and Laguecir, 2023).

Third, following the previous identification of barriers, difficulties, and resistances, this may be of special interest to practitioners, in the sense that it can provide valuable insights into the situations that may arise upon implementing the BSC. And, as evidenced by Northcott and Ma'amora Taulapapa (2012), learning about the successful/unsuccessful factors is key to the uptake of the model. Consequently, to avoid maintaining old and new ways simultaneously, implementers should emphasise the importance of the new system and support its adoption through training, particularly when facing rigid structures or limited prior experience with the systems being implemented. Additionally, special care should be taken to design a system that is easy to use.

This study, as is the case with all research, is not exempt from limitations. In implementing management accounting techniques, each case represents a separate situation, and the results presented here are not entirely generalisable. Nevertheless, they provide insight into similar situations. Future studies could, therefore, address how different public organisations have dealt with distinct barriers to the implementation of the BSC, highlighting their strategies and showing the impacts on the model developed. It would also be valuable to study how decision-making processes may be affected by using BSC information (Steccolini, 2019).

In addition, this study describes a less-successful instance of implementing the BSC, while the literature has many more cases of successful implementation of various PMM tools. The implementation processes can benefit from an in-depth understanding of the failures that may occur, which would allow us to respond to some of the calls of Zawawi and Hoque (2020), and look into organisations that have discarded the BSC.

The authors gratefully acknowledge the insightful comments and suggestions provided by the journal's editors, Professor Orthodoxia Kyriacou and Professor Dorothea Greiling, as well as by the two anonymous reviewers. We also wish to thank Ricardo Barradas, Maria Antónia de Jesus, and the participants of the following conferences and seminars for their valuable feedback: the XXIII Grudis Conference and Doctoral Colloquium, held in Aveiro; the MCA Online Workshop; the 46th European Accounting Association Annual Congress, held in Bucharest; and the 2024 Research Seminar of Universidade Europeia, held in Lisbon. We thank all interview participants for their time and valuable input.