Understanding political accountability in a strong structuration framework

Abstract

Purpose

Research was conducted to investigate whether, and how, political accountability might stabilise when agents are faced with profound changes in external structures such as competition laws and austerity policies.

Design/methodology/approach

We performed a field study from 2007 to 2015 in a regional hub in Finland and worked with data from document analysis, interviews and meeting observations. We have used embedded research design, where we apply methodological bracketing as well as composite sequence analysis for field research.

Findings

Accountability declined when irresistible external structures were the dominant influence on the unreflective actions of agents-in-focus. With time, however, the agents started acting critically by drawing on structures that could facilitate strategic actions to stabilise political accountability.

Research limitations/implications

The field research and interpretation of the data were limited to the organisation analysed; however, the theoretical arguments allow for analytical generalisations.

Practical implications

The research demonstrates how public officials and political decision-makers can eventually adopt a strategic approach when faced with irresistible change in external structures.

Social implications

The research demonstrates how public officials and political decision-makers can eventually adopt a strategic approach when faced with irresistible changes in external structures.

Originality/value

The study locates political accountability in the context of strong structuration theory and discusses how it is redefined by external structures.

Citation

Mutiganda, J.C. and Järvinen, J.T. (2021), "Understanding political accountability in a strong structuration framework", Accounting, Auditing & Accountability Journal, Vol. 34 No. 9, pp. 80-103. https://doi.org/10.1108/AAAJ-12-2017-3264

1. Introduction

Political accountability – that is, accountability among public officials and political decision-makers – is attracting increasing attention in public sector accounting research (Fowler and Cordery, 2015; Hyndman et al., 2008; Nyland and Pettersen, 2015; Sinclair, 1995). An important contemporary theme in such research is the influence of widespread austerity policies on public sector accountabilities (Bracci et al., 2015). Previous studies have illustrated how accountabilities shift as budget cuts must be legitimised (Ahrens and Ferry, 2015; Pellinen et al., 2018). An important question therefore remains: how can individual agents reformulate accountability relationships when the structures that enable or limit action rapidly change and what tensions arise from redefining budgetary accountabilities (Mutiganda, 2013, 2014; Ferry et al., 2019)?

This study employs a revised version of structuration theory (ST), referred to as strong structuration theory (SST), in the manner suggested by Stones (2005) to study the redefining of accountabilities under austerity policies. According to Englund et al. (2011, p. 506), although traditional Giddensian ST has led to some lasting contributions to accounting research, it displays a major limitation in its lack of “overall standalone empirical achievement”, whereby its empirical use is restricted to an “ontological point of departure” and therefore manifests a constrained ability to address contextual accounting practices. Offering an alternative ontological point of departure, Stones (2005) argues for moving from the flat and local ontology – or “ontology in general” – on which ST is based (Giddens, 1979, 1984) to ontology-in-situ when analysing relationships between structure and agency.

This study argues that ontology-in-situ is particularly suitable for the study of accountability. Hence, using SST, our approach, rather than taking structuration concepts as mere guiding principles or “sensitising devices”, allows for the analysis of how particular agents draw from and act on their specific knowledge of both the structures and other agents involved in the network during structuration (Daff and Jack, 2018; Greenhalgh and Stones, 2010). It also directs attention towards empirical issues such as which agents the focus lies on, what their position–practices are and how the agents draw upon knowledge about various structures. Thus, SST represents a disciplined methodological approach to conducting empirical field studies (Jack and Kholeif, 2007; Stones, 2005). SST shows similarity to the older version of ST in terms of reaffirming that no primacy exists between agents and structure as well as concentrates on how and why structuration takes place (Daff and Jack, 2018; Giddens, 1979, 1984; Stones, 2005).

Our study locates political accountability in a public sector setting, wherein agents were faced with radical changes in the competition laws and austerity-driven fiscal policies that define their external structure. In particular, we contribute to SST studies with our investigation of the location of political accountability in a strong structuration setting, where we use “accountability” to refer to the ways in which a person or an organisation with an obligation to account for conduct delivers the account to other persons or organisations to whom it is due (Messner, 2009; Roberts, 2009; Roberts and Scapens, 1985). Many structuration studies in the domain of accountability have placed more emphasis on the structure of signification, domination and legitimation than on agency (Coad and Glyptis, 2014; Englund and Gerdin, 2014; Roberts, 2014; Roberts and Scapens, 1985); therefore, we argue that agency itself remains under-researched in accountability scholarship. In light of this, we asked the following research question: how do changes in structures influence the agency of those involved in political accountability?

This project involved a longitudinal field study in a “regional city”, referred to below as Viking City, in Finland during 2007–2015. We obtained our source data through document analysis, interviews and meeting observations. The field study used an embedded research design (Yin, 2018) that includes methodological bracketing (Stones, 2005) and composite sequence analysis (Miles et al., 2014; Greenhalgh and Stones, 2010) to identify and analyse the structures that influenced what active agents did in designing and implementing competitive tendering and explaining their actions in relation to political accountability. The case study covers two competitive tendering processes that took place in 2007–2008 and 2013. At first glance, political accountability seemed to decline over the course of the first competitive tendering process. During the course of the second process, however, public officials adopted strong positions to monitor macro-organisational practices so as to minimise procurement-related risks. In 2013, the restructuration of the competitive tendering process seemed to be leading to the stabilisation of accountability relationships – that is, the relationships become internalised and are not subject to critical reflection by the agents.

Our work makes several contributions to scholarship. First, we extend SST-based research by examining the location of accountability in terms of the quadripartite framework (Daff and Jack, 2018; Greenhalgh and Stones, 2010; Makrygiannakis and Jack, 2016). Next, we add to accountability studies by offering an investigation of abstract (structuration), meso (political and managerial) and ontic (practice) level concepts that shape accountability (Roberts, 2014; Stones, 2005) in a continuous process of structuration. Our field study reveals how and why accountability stabilises (or does not) when changes occur in the context and conduct of the agents-in-focus. Finally, the practical implications of our findings tie in with the implications of austerity policies and their localisation (Ahrens et al., 2020; Ferry et al., 2019; Ahrens and Ferry, 2015; Bracci et al., 2015): our work illustrates how public officials and political decision-makers encountering irresistible change in external structures have been able to strategically act to stabilise accountability (Järvinen, 2016; Pellinen et al., 2018).

In the next section, we describe the theoretical framework used in the study and its operationalisation, followed by the explanation of the research methods and context-related data. This background is necessary before the presentation of the findings, which first describe the case study narrative. Next, we illustrate how accountability relationships evolve under changing external structures. Finally, we discuss the emergence of a strategic setting for political accountability and present our concluding remarks.

2. The theoretical framework

At this juncture, it is important to explain the concepts that Stones (2005) used to refine and extend the conceptual framework of the ST of Giddens (1979, 1984). This explanation is followed by a brief review of accountability in relation to the topic of our study.

The ontological foundation of ST (Giddens, 1979, 1984) is a combination of objectivist and subjectivist social realities. Whereas an objectivist ontology assumes that social reality exists independently of individual agents' knowledge of it, a subjectivist ontology presumes that social reality depends on the respective agents' knowledge and interpretation thereof (Giddens, 1979). Accordingly, ST is based on an assumption of duality of social structure – that is, an external reality that exists independently of individual agents and an internalised reality that is an outcome of a hermeneutic understanding of what the agent knows and does (Greenhalgh and Stones, 2010). Critics contend, however, that Giddens failed to clarify when a structure can be seen as an analytical memory trace vs a resource. It is also unclear how one can examine agents as entities who are separate from the structures upon which they draw to act and whose actions can change the structures. Critics often describe the issue as conflation or elision of structures and agents (Archer, 1995; Cohen, 1989).

By contrast, Stones (2005) argued structuration to possess a quadripartite nature, where the components of the accordant quadripartite framework are (1) external structure, (2) internal structure, (3) active agency and (4) outcomes. Each component is conceptually distinct from the others and can be independently analysed. Stones (2005) described external structures as those affecting internal ones and stated that both external and internal structures affect the actions of active agents, which lead to specific outcomes. Outcomes affect external and internal structures later on, which may explain why and how agents vary their actions in a continuous process of structuration (Chan et al., 2010; Greenhalgh and Stones, 2010; Jack and Kholeif, 2008).

One of the merits of SST is its inclusion of the notion of the time and space of positioned agents in the analysis of structuration (Makrygiannakis and Jack, 2016); this is connected with the intermediate zone of position–practices (Stones, 2005, pp. 81–84). From this perspective, external structures of an agent-in-focus can be analysed even though these are not external to agency (Makrygiannakis and Jack, 2016).

External structures, internal structures, agents' practices and outcomes are situated at the abstract ontological level; they are manifested in different spaces and at different times (Stones, 2005). Similar to Stones (2005), Makrygiannakis and Jack (2016) argued that investigating these abstract ontological concepts' operationalisation necessitates translating them “to ontic/in situ categories” – that is, to the empirical level. Stones suggested a further theoretical articulation of the factors that influence the abstract-to-ontic translation process at the meso-ontological level, which lies between the abstract level and the ontic level. Figure 1 illustrates this.

At the meso-ontological level, factors that constrain or afford external structures' influence need to be analysed. In fact, the major divergence of SST from Giddensian ST pertains to the existence of external structures that are independent of agents' cognition.

From this perspective, it is the agents' capabilities and knowledge about the social world that matters (Greenhalgh and Stones, 2010), in addition to each agent's position–practices (Stones, 2005). Accordingly, our study proceeds from an understanding that although changes in environment (in our case, competitive tendering regulation, austerity policies and the emergence of new market forces) do influence agents, irrespective of whether we can observe their actions, the agents' knowledge of the structures is situated and conjuncturally specific. Because the agent's conduct guides their practices, it is appropriate to analyse how and why that agent's practices are routine or strategic and whether the outcome of agency was intended or unintended (Makrygiannakis and Jack, 2016; Stones, 2005).

Our study exploits these meso-level ontological concepts to investigate the place and relevance of political accountability at the abstract and meso-ontological levels using the ontic-level data from Viking City. Previous ST literature has cited issues related to managerial accountability that are often present in decision-making that involves allocating financial resources and monitoring their use (Coad and Glyptis, 2014; Conrad, 2005; Roberts and Scapens, 1985); nevertheless, studies on political accountability are still scarce, although exceptions exist (see Ahrens et al., 2020; Ahrens and Ferry, 2018; 2015; Roberts, 2014).

From this study's viewpoint, the two forms of scholarship – managerial and political accountability – differ in a decisive way. In general, research on managerial accountability tends to focus on how persons assigned with specific performance targets demonstrate their achievements and explain any deviations from the targets to superiors in the hierarchy. However, research on political accountability goes a step further, focusing on how persons to whom authority has been delegated – such as political decision-makers and public officials – are answerable, directly or indirectly, to society or to other people who are the sources of that authority (Day and Klein, 1987). Accordingly, political accountability involves managerial accountability and the additional obligation of accountability to a larger audience, such as the society that has delegated authority to political decision-makers via democratic elections. Both political decision-makers and society expect managerial accountability from the public officials who have been given specific performance targets (Scarparo, 2008).

The web of external structures for public officials is extensive because public officials must be able to understand the political praxis and social positioning that unfold when politicians make decisions to allocate resources to them and monitor their use (Lapsley et al., 2011). At the same time, public officials may find it difficult to gain the trust of politicians and of the society electing them (Coad and Glyptis, 2014; Roberts, 2002; Scarparo, 2008). Therefore, public officials and political decision-makers must take on the role of active agents (Coad and Herbert, 2009) by constantly monitoring the position–practices and changes in their external structures and by updating their internal structures both in general and in particular (Mutiganda, 2016) such that their actions conform to political accountability expectations and the continuous obligation of fulfilling their duty (Adserà et al., 2003; Scarparo, 2008).

Active agents have to expend power and resources, communicate with other active agents, and demonstrate the ability to make or propose the decisions necessary for their actions to bring about the expected outcomes (Daff and Jack, 2018; Stones, 2005). Based on this line of reasoning, we made three theoretical arguments and operationalised them through sub-research questions.

The first argument is that when external structures change but internal ones are free to remain the same, the strong structuration outcome entails no change or merely surface/ceremonial changes in political accountability. This argument is informed by an assumption that if the law on competitive tendering provides the possibility for public officials and local politicians to continue utilising direct negotiation with local operators in the outsourcing of public elderly care services, their political accountability will remain the same as before or will only formally/ceremonially change.

The second argument is that when external structures change and internal structures are forced to adapt to the changes of external structures, the strong structuration outcome leads to unstable political accountability. This argument is rooted in the assumption that the internal structures of agents-in-focus (public officials and elected politicians) have potential weaknesses that cannot be ameliorated overnight. In this setting, they cannot avoid or limit the procurement risks that stem from dealing with highly experienced and sophisticated business operators interested in competing for a greater market share in Viking City. The two theoretical arguments play a central role in the investigation of the political accountability found in the process of designing and implementing Viking City's first competitive tendering process for elderly care services. The study operationalised these two arguments by asking this sub-research question: when external structures change, why and how does political accountability either change or not change?

The third argument is that when no further changes are made to external structures but internal structures do change, the strong structuration outcome leads to stabilised political accountability relationships. This argument was articulated to investigate the political accountability in the process of designing and implementing a competitive tendering process for elderly care services for the second time in Viking City. We operationalised the third theoretical argument by means of the following sub-research question: how does the stabilisation of internal structures stabilise or destabilise political accountability among agents-in-focus when external structures have not changed?

To discuss the above-mentioned arguments, we focused our data collection on agents' context and conduct (Makrygiannakis and Jack, 2016). For this setting, Stones (2005) advised applying methodological bracketing, originally formulated in part by Giddens (1979), to collect data of relevance for the SST's position–practice approach. Methodological bracketing is aimed at sharpening the focus of the research and improving the robustness of the findings by either combining the analysis of agents' conduct and their context or focusing on either one of the two (Coad and Herbert, 2009; Jack and Kholeif, 2007).

3. Research context, methods and data

3.1 Methodology

One of the merits of SST (Stones, 2005) is that it explains how a researcher can combine the bracketing of an agent's conduct and context analyses in designing, conducting and analysing research findings (Greenhalgh and Stones, 2010; Makrygiannakis and Jack, 2016). Conduct analysis focuses on the internal structure of an agent by examining that agent's ontological knowledge ability seen in their hierarchical sources of concerns, desires, ways of thinking and doing, reflective monitoring and interactions with other agents (Stones, 2005). Context analysis, in contrast, is outward-focused. The analysis starts with the examination of how specific internal structures are connected to the position–practice relations in external structures, such as rights and obligations, power relationships among or influencing active agents, conditions of action in a given societal setting and consequences of actions (Coad and Herbert, 2009; Stones, 2005). The analysis of an agent's context is performed to obtain information that helps a field researcher understand how the conduct of the agent shapes and is shaped by the position–practices situated in the field of the agent's actions (Stones, 2005) by focusing on the possibilities and limitations that may affect the agent's actions in a specific way (Chan et al., 2010; Greenhalgh and Stones, 2010).

To design and collect research data, we combined SST (Stones, 2005) with the embedded research design strategy (Yin, 2018) for the following reasons: (1) Agents-in-focus and their conducts are embedded in the process of implementing an open competitive tendering process and managing its outcomes. (2) Political accountability lies between two different categories of agents-in-focus, namely public officials and elected political decision-makers. Hence, political accountability is embedded in the context and conduct of agents-in-focus (Kholeif and Jack, 2019). (3) Combining SST with embedded research design allows us with appropriate steps to collect in-depth data on the subjects of change, objects of change and forms of change in relation to each agent-in-focus and among agents-in-focus belonging to the same category at a specific time and in a specific space (Greenhalgh and Stones, 2010; Makrygiannakis and Jack, 2016).

The embedded case study covers different sub-units (Yin, 2018), that is, different agents-in-focus – and some local operators – of a single organisation, Viking City. Having analysed the conduct of each agent-in-focus in their units and category, we moved on to analysing how the actions of each agent-in-focus influenced (or did not influence) the conduct and action of other agents-in-focus in specific contexts and the reverse (Adhikari and Jayasinghe, 2017). Greenhalgh and Stones (2010) and Stones (2005) recommended that conduct analysis should involve at least two steps: (1) identification and analysis of the general dispositional frames of meaning of each agent-in-focus and (2) analysis of conjuncturally specific internal structures set within the agent's general dispositional frames of meaning. Context analysis, too, should feature two key steps: (1) identification of position–practice relations that constitute clusters of external structures, such as power, authority, rules and resources available for the agent's use at various levels in the hierarchy and (2) analysis of the possibilities and constraints that influence each agent-in-focus during structuration (Stones, 2005).

In line with the approach adopted by Greenhalgh et al. (2013), the data of each agent-in-focus contributed to building an emerging “network”, a visualisation of all the agents-in-focus capable of exerting influence on political accountability among themselves and, similarly, the in situ variability among them. In this setting, agents and structures belonging to a similar category formed a sub-network, namely the political, managerial and outside observers' sub-networks. The outside observers' sub-network is composed of journalists, local operators in elderly care and active family members of the elderly.

To interpret the data from each sub-network and to analyse the relationships among all sub-networks, our study used composite sequence analysis (Miles et al., 2014). Composite sequence analysis is appropriate for identifying and explaining the sequential processes associated with active agency in each sub-network (Miles et al., 2014). In particular, composite sequence analysis helped us to identify, map, and examine the internal and external structures of each agent, their position–practices within the political, managerial and outside observers' sub-networks, and the actions taking place between different sub-networks-in-focus.

In line with the suggestions of Miles et al. (2014), we employed two analytic strategies to collect and analyse the data by relying on each theoretical argument in focus and examining the possible rival explanations based on the three arguments. Table 1 presents a concise overview of the data analysis, especially of how our theoretical arguments are linked to the sequences in our data. The table also illustrates how we divided one of our sequences (sequence 3) into two sub-sequences, enabling external structure change and irresistible change, to facilitate further analysis.

By analysing the relationships between different sequences, networks and sub-networks, we aggregated from an agent-in-focus to agents-in-focus and investigated the evolution of political accountability in the continuous process of structuration and change (Adhikari and Jayasinghe, 2017; Miles et al., 2014). The aim of this analysis was to enable analytical generalisation (Yin, 2018), where the first step is to illustrate how the sequence of events in our case study bears upon the concepts of SST. Table 1 summarises our data analysis.

The second step of analytical generalisation is to apply SST to implicate rival explanations (embedded in our sub-research questions) to find the pattern matches (Yin, 2018) between the “why” and the “how” – that is, the relationships between sub-research questions 1, 2 and 3 and their relevance for explaining the place of political accountability as a construct in SST at the ontic, meso and abstract ontological levels (Stones, 2005).

3.2 The context of the study

Viking City was determined to be an appropriate site for our field study when the decision was taken to implement a competitive tendering process for elderly care services for the first time in 2007. At the time, the parliament passed legislation that mandates all public sector procurements to be subject to competitive tendering. Thus, the decision to competitively operate was a direct result of changes in legislation. Direct access to public officials, political decision-makers, local media, the public and local providers of elderly care services as well as the transparency and openness implied by opportunities to observe political and management meetings in the city provided further motivation.

As a municipality in Finland, Viking City has the freedom of local governance. Political decision-makers are elected in municipal elections for a four-year term to represent their political party at several levels in the hierarchy (e.g. in the municipal council, the audit committee and the executive committee). The divide between politics and administration in the municipality means that the city director is usually a civil servant, making him/her a public official rather than a politician.

The municipality finances a substantial proportion of the elderly care costs, with the elderly being charged on an afford-to-pay basis (Tynkkynen et al., 2012). Municipal finances significantly depend on municipal tax revenues and the subsidies the central government allocates on a per capita basis once a year (Hyvönen and Järvinen, 2006; Kurunmäki, 1999). The Municipality Act (1995) requires each municipality to have a balanced budget and to avoid accumulating budget deficits over the four-year term corresponding to the calendar for municipal elections. Unfortunately, prior to 2007, the social and healthcare committee had not managed to avoid exceeding its budget. In 2007, the pressure to reduce elderly care costs increased because the city's tax revenues were decreasing and both social and healthcare costs and unemployment rates were steadily rising. The entry into force of the 2007 Act gave the public officials and politicians the opportunity to switch from direct negotiations with local entrepreneurs to an international competitive tendering-based system.

3.3 The data in context

Our longitudinal case study ties in with two major events, one in 2007–2008 and the other in 2012–2013. To leverage the full potential of SST, an embedded research design approach (Yin, 2018) is used in combination with concepts from the quadripartite framework (Stones, 2005). We adopted the following theoretical starting points: First, political accountability relations between elected politicians and public officials represent the ontic level. Second, structures through which political accountability arises represent the meso-level; this level includes board meetings and agendas, putting forth motions, having to vote and asking public officials to offer oral explanations in addition to written reports. Finally, community expectations and the factors that afford or impair public officials' meeting of those expectations are at the abstract level.

Among our empirical data sources were Viking City's internal documents as well as our own observations. The document analysis revealed various aspects of the political decision-making process of designing and implementing the calls for competitive bids and approving certain bids. Meeting observations yielded further information on how the process of accountability involving politicians and public officials was actualised and the reasons the politicians cited in support of their decisions – for example, when commenting on their decisions to local media outlets and officially explaining them to the municipal audit committee.

Our field study began in 2007 with agent context analysis. The data for this analysis was obtained from the analysis of documents such as the 2007 Act on Public Tendering, the EU's directive on public procurement, the Municipality Law (1995), Finland's social and healthcare regulations, material on municipal obligations to provide elderly care services, official annual financial reports of Viking City from 2006 onwards, budgets for elderly care in Viking City from 2006 onwards and details of their implementation, reports of the social and healthcare committee, local media articles and public debates on national television. Next, we performed content analysis and collected further information by talking with two key politicians who were members of the city council and with key public officials, such as the city's attorney, chief financial officer and manager of social and healthcare service finances.

To analyse the agents' conduct and internal structures, we conducted formal interviews between 2009 and 2013. Among the interviewees were the social and healthcare director, the manager of elderly care services, the head and two other members of the social and healthcare committee, representatives of the internal and external auditor and a member of the audit committee (which monitors the decisions of the social and healthcare committee). To increase the validity of the study, further interviews were conducted with people who had extensive experience in the local elderly care market. The interviewees included a politician who was a member of the national parliament, two key managers and chairpersons of local elderly care organisations and four members of the executive committee of one of the organisations.

To aid in the analysis of the relations among external structures, internal structures, what active agents did, outcomes, and political accountability, the field researcher attended four meetings of the social and healthcare committee between 2009 and 2011. Each meeting dealt with issues related to the procurement of social services, with particular focus on elderly care. The meetings demonstrated how public officials used accounting information in addition to healthcare criteria to explain to politicians the problems in the procurement of elderly care services. Further meeting observations were performed with the audit committee from 2009 to 2013. The committee invited the mayor and all key public officials involved in providing social and healthcare services to explain the city's budgetary and social policies, the outsourcing strategies and their implementation. In this way, we were able to follow the sequential processes associated with active agency as a part of composite sequence analysis. We ensured that sufficient data was collected to separately analyse each sequence and to separate the most important sub-sequences in the chain of events (Miles et al., 2014).

A final round of interviews was conducted in 2015 to examine the situation of political accountability among public officials and political decision-makers particularly during the implementation of the service contracts that were the outcome of the second competitive tendering process. Interviewees included the manager of elderly care services, the financial manager for elderly care services, two members of the social and healthcare committee (its head and one other) and the vice-chair of the audit committee. All politicians interviewed in 2015 also sat on the city council. Both rounds of interviews, with a total duration of 28.5 h, were recorded and transcribed. The time devoted to meeting observations was about 60 h. Tables 1 and 2 provide a summary of the data collection statistics. The sets of data collected via the various methods supported each other, thus establishing the validity of the study (Miles et al., 2014; Stones, 2005; Yin, 2018).

4. Findings

In this section, we first present the background and a case narrative of Viking City's competitive tendering processes. We then map the relevant networks within the sequences in focus. Our analysis is guided by Miles et al.'s (2014) guidelines on how to report the results of composite sequence analysis and by Greenhalgh and Stones' (2010) insights into conducting an empirical investigation from the perspective of SST.

4.1 Background and an overview of the case narrative

In late 2007, Viking City launched an international call for bids for its elderly care services. Local, regional and international companies positively responded, and in the beginning of 2008, the social and healthcare committee approved a proposal by the director of social and healthcare services to approve the bids of two particular companies, one international and the other local, because these had offered the lowest prices. The local company immediately started to implement its new contract with the city. The international company delayed implementation because it had no facilities or personnel in the city. One of its affiliated companies required two years to finish building a new healthcare facility.

When work under the international company's new elderly care contract with the city was finally about to start, the city became obliged to conclude two additional exclusive contracts for renting and maintaining public spaces within the new healthcare facility to make the main contract workable at a lower cost to the elderly. The city had neither intended to conclude these two contracts nor budgeted for them when approving the international company's bid in 2008. Therefore, the total cost of implementing all the contracts was much greater than what the public officials and political decision-makers had expected. Political accountability became a subject of lively debate in the local media and at political meetings. As a consequence, the head of the social and healthcare committee made a public announcement in 2011 that the city would no longer use international competitive tendering to outsource services for the elderly. Instead, a decision to recruit an expert in public procurement methods and strategies was made.

Two years later, the city attempted to arrange local competitive tendering for elderly care services but was unable to keep it local because the total amount of financial resources to be allocated to the relevant contracts was above the threshold set by the 2007 Act. At this time, austerity policies were being implemented across the board, and their effects were experienced in the form of an increasing pressure for cost savings. Accordingly, the city turned its local call for bids into an international one. The ensuing competitive tendering process proceeded more smoothly than the first one had. The bids of three companies – one local and two international – were approved, as they offered the lowest prices. Two of these were the companies that had won the previous competitive tendering, and both started working under their new contract in a timely and smooth manner. The second international company delayed the implementation of its contract, as it first had to build a new elderly care centre in the city. It has since begun fulfilment of its contract.

4.2 The embedded political accountability before a new competitive tendering law

Our study revealed that the praxis of outsourcing elderly care services before the implementation of a competitive tendering process was dependent on direct negotiations separately conducted between Viking City representatives and the representatives of each local elderly care organisation every year. Here, the agents-in-focus were the top local public officials and the head politicians (those sitting on the social and healthcare committee). Each negotiating party had a specific social position, such as representing the interests of the city and society in general or representing the private elderly care market and healthcare expertise.

The sequence of political accountability during this era was shaped by the professional expertise and power of the local producers of elderly care services (Mutiganda et al., 2013) to negotiate elderly care service prices (that covered all costs) with public officials responsible for elderly care (i.e. the head of the social and healthcare committee, director of social and healthcare services and manager of elderly care services of the city being politicians). The sub-category of local providers of elderly care services did not have a forum to set prices for these services because they were competing against each other. However, all of them wanted to cover their costs. The sub-category of elected politicians wanted to conclude the contracts that would be acceptable to the public to avoid criticism, even if those contracts were concluded at a relatively high price. In contrast, the sub-category of public officials insisted on lower prices of contracts and high-quality service standards. This context would explain why local providers and some politicians would have preferred to continue such budgetary negotiations rather than starting a risky international competitive tendering process. However, meeting observations showed lively debates between the three sub-categories during contract negotiations. A member of the social and healthcare committee commented:

Local providers of services for the elderly kept raising their prices […]. Negotiations with local organisations were difficult.

Being non-profit organisations, the local providers of elderly care services used corresponding business logic to justify their positions (see Järvinen, 2016). A chief executive officer of one local organisation explained:

This is a not-for-profit organisation […]. I have a responsibility to pay salaries and all allowances to our personnel as dictated by national law […]. I also have a responsibility to design elderly care services that put the elderly at the centre of our activities […]. All of these are costs […]. We cannot offer under-priced services to the city.

The final conclusion, however, usually involved asking local providers to slightly lower their prices, after which the politicians would approve the newly amended prices and defend the approvals as political decisions made at higher hierarchy levels (i.e. in the social and healthcare committee and later in the executive committee of the city). In this way, elected politicians and public officials collaborated to achieve further political legitimacy for the outsourcing. For example, a member of the social and healthcare committee explained:

The problem was that the budget of our board was not enough to cover the costs of outsourced elderly care services […]. We had to negotiate lower prices.

Document analysis confirmed that the council and executive committee of Viking City insisted on reducing the costs for its healthcare services, including elderly care. The executive committee used austerity-driven budgetary policies, such as the last-minute budget cut, to reduce the amount allocated to the social and healthcare committee every year. For instance, the analysis of the minutes of board meetings indicated that the social and health care committee had never taken a vote on approving or rejecting budget increases. In practice, contracts were drafted by the public officials in concert with the board chairman, who then received the board's approval before actually signing the contract.

Consequently, we argue that although local entrepreneurs tried to dominate the market as price-setters, Viking City representatives held both financial and political monopolies and could exploit their political power to set lower prices for outsourced services. In particular, the external structures of public officials and elected politicians as agents-in-focus were independent (Stones, 2005). Their social positioning in the elderly care market enabled them to do all they could to avoid criticism from local media and the family members of the elderly, including concluding elderly care contracts at higher costs. Because the prices for elderly care were negotiated every year, the internal structures of elected politicians and public officials were based on established routines that had somehow become taken for granted at Viking City. In this context, the agency of their political accountability was ceremonial – that is, politicians and public officials had all the tools to maintain the status quo, despite the continuously rising costs of public elderly care. This finding provides a good background for our first theoretical argument, which is elaborated on in the next section.

4.3 Weakness of political accountability owing to new and constraining external structures

The entry into force of the 2007 Act (the competitive tendering act) marked a major shake-up in every sub-network that was part of the elderly care market in Viking City. First, the sub-network of elected politicians knew almost nothing about how to implement the 2007 Act in political decision-making. The internal controller commented:

[O]ur public officials and political decision-makers were not aware of what they were doing and what they needed […]. They have no skills in dealing with competitive tendering contracts […] that involve multinational business organisations.

Second, the sub-network of public officials had extensive knowledge about how to manage public elderly care services but did not understand how to operationalise international competitive services in the city to limit procurement risks. Interviews with elected politicians and local entrepreneurs showed this weakness. The city's elderly care service manager explained:

When the social and healthcare board meeting made a decision to start […] competitive tendering […], technical aspects were not an issue […]. However, I had no toolbox ready to use […]. I had no prior background […] in this specific matter […]. We were simply not ready.

Third, the sub-network of local providers of elderly care services saw a possibility of being overtaken by outside competitors almost overnight and did not know how to successfully compete with them. A local organisation executive committee member commented:

In competition, the risk of a local company being overshadowed by an international business competitor was clear.

We conclude, therefore, that the internal structure of each agent-in-focus, regardless of the network that agent belonged to, was weak and uncertain.

The sequence of political accountability in this setting became marked with new positioning of public officials and an unreflective approval of the majority of political decision-makers. Local media and private elderly care providers became very critical about the way competitive tendering was being handled. For instance, the director of social and healthcare services used a legal argument to justify why competitive tendering was unavoidable:

The new public procurement law is compulsory […]. It addresses competition in public-sector organisations […]. The city has decided to implement it in elderly care […]. National and international entrepreneurs are welcome to submit their bids.

The previous statement, however, was based on weak technical certainty – that is, specific internal structures or habitus (Stones, 2005) for how to operationalise international competitive tendering in healthcare. For public officials, having to consider what was documented was, in such circumstances, a justification of why they should not be blamed for making wrong decisions as long as what is written is in line with what public officials promised to leading political decision-makers. A member of the audit committee, however, remained critical:

When I read all the offers for elderly care services submitted to the city, I made my own calculations on the soundness of their price levels […]. On that basis, I warned the city's director of social and healthcare services that some of the apparently appealing offers were seriously under-priced.

Critics also addressed how the new chosen companies could guarantee the quality of elderly care services. The director of social and healthcare services explained:

[T]here are so many laws and recommendations […] about the quality of healthcare services […]. We have to consider all of them […]. But what is quality? […] And how can we measure it…

In addition to the apparent weakness of the internal structures of agents-in-focus, particularly public officials of elderly care, there seemed to be a lack of trust in the reliability of the bid evaluation process. For example, when facing criticisms about bid evaluation in choosing an international company for the Viking City market, the manager of elderly care services responded:

That company has workshops elsewhere in Finland […] and abroad […] and quality certificates […]. There was no reason to give it lower scores for quality.

The underlying criticism here was not that the company obtained full points on theoretical quality standards during the bid evaluation but that local companies with high-quality healthcare standard records did not obtain full points.

Here, it seems that the embeddedness of political accountability arises from the relations between the external structures that shape and challenge the internal structures of political decision-makers and public officials in focus when dealing with and justifying (or giving account of) specific events for which they have taken specific positions (or decisions). In the case of Viking City, political accountability became ceremonial during bid evaluation and during the implementation of contracts based on selected bids.

When dealing with rectifying the mutual mistakes of each agent-in-focus by approving additional exclusive contracts with the international company that won the competitive tendering, a local politician who was member of the social and healthcare committee commented:

Politicians […] and public officials […] do not like to be told that they have acted wrongfully.

An influential politician who was on that committee at the time provided further explanation:

We first approved the outcome of the competitive tendering […]. It could have been difficult to request further accounts from public officials later with regard to something that we had already approved.

A member of the social and healthcare committee confirmed:

At the board meeting, during which we approved these additional contracts […], board members did not really know what else to do […]. It was like […] 'we cannot do anything other than approve them […] so that the elderly can afford to live there […] at costs that are affordable for them.

In this setting, political accountability at Viking City was at the intersection of many networks (the politicians, public officials, and outside observers). However, its form and content depend on how and why members of the political and managerial networks could decide to collaborate with each other, ceremonially or not. The findings of Viking City that were ceremonial in nature are explained by a willingness of both influential politicians and public officials to cover up each other’s mistakes. Note that although these mistakes were not intentional, they could have been avoided, for instance, by listening to warnings from some members of the audit committee and critics of local providers of elderly care services who did not win the competitive tendering.

The director of social and healthcare services echoed this sentiment, later offering further thoughts:

In a way, I have now realised that competitive tendering for elderly care services is not the most appropriate strategy […]. It can even lead to a monopoly in the long run.

In this context, a political decision to recruit an expert in public procurement methods and strategies was unanimously approved. Local entrepreneurs and politicians were satisfied with the city's announcement that unknown, new companies would not be brought into the local market. A member of the Finnish parliament had a similar approach and made proposals to change the competition law in elderly care:

Competition in elderly care undermines our social values and welfare principles […]. In some cities and municipalities, the elderly have had [to move] from their elderly care institution to new ones, owned by the winners of the competition […]. I have, therefore, asked the parliament to consider amending this law.

We argue, therefore, that when agents-in-focus and their networks have difficulties in controlling the risks arising from irresistible and constraining external structures, they tend to reach a consensus at an organisational level on how to amend the external structures so that these structures are more comfortable for them. The rationale behind this at Viking City was an attempt to stabilise political accountability by reducing further criticisms from members of the outside network, which is in line with our second theoretical argument and provides an answer to its underlying sub-research question.

4.4 Towards strategic settings for instrumental political accountability?

The attempts to revise the 2007 Act to make it more comfortable for local politicians and public officials of elderly care did not lead to quick results at the national level. Consequently, Viking City became obliged to organise a second competitive tendering process in 2013. In this phase, the agents-in-focus who were members of the political and managerial networks had changed: with the municipal elections in 2012, the composition of the social and healthcare committee had changed; the city had recruited a new manager of elderly care services and a new procurement manager. The procurement manager had documented experience of dealing with open competitive tendering processes and claims in Finland. The new manager of elderly care services was a well-regarded health professional but had no previous experience in international competitive tendering. Leading re-elected politicians had previous experience of dealing with the first competitive tendering process at Viking City. Obviously, one key way for an organisation such as a local government to resolve the paradox of embedded agency – how to change internalised structures (see Englund et al., 2011) and overcome restrictive internal structures – is to bring in new agents.

The city first invited all local providers of elderly care services to a general meeting in early 2013, where they were asked to express their opinions on what the new competitive tendering process among them should look like. We interpret this meeting as the first sub-sequence that shaped the process of planning the second competitive tendering process. Representatives of all the local providers attended this meeting. It was now evident that competition laws were more strictly being enforced and that irregularities in tendering processes could easily lead to court proceedings. We interpret this as external structures having become stable and solid.

After the presentation part of the meeting, each interested company provided feedback. The atmosphere was calm and focused. Each company brought up issues that required consideration and clarification in the invitation to submit bids. In a post-meeting follow-up, the manager of elderly care services and the executives of interested local companies continued the work.

Our results, which are organised according to the quadripartite framework (Stones, 2005), show that public officials and key politicians in Viking City started the process of designing the new call for tenders by analysing the positions of local companies on key praxes that should receive focus. We interpret the bringing of new agents together as a process of collecting information about the internal structures that had to be in focus to re-establish the trust embedded at the intersection of the political, managerial and outsider networks.

In this setting, we interpret trust as an internal structure, a deeply embedded belief that matters can be resolved by co-operating with long-time partners, with whom the agents have done business before. The meeting with all local companies also served to identify the social practices that all companies were willing to follow. The head of the audit committee commented:

I have learnt that the director of social and healthcare services now has good relationships with all local suppliers of elderly care to the city […]. I trust that the new competition process will work out well.

All members of the audit committee agreed to that the process was now working well and that the partners were trustworthy. A press release from the director of social and healthcare services confirmed that public officials were confident in the city's ability to organise a new competitive tendering process. Even if the set of services to be procured was to vary, the agents (public officials, politicians and local providers participating in the procurement process) were envisioned as remaining the same.

Political accountability in position–practice relationships between public officials and politicians could be restored by means of good communication, which, in turn, was made possible by the high level of trust between all the agents-in-focus (Coad et al., 2015. More specifically, although new external structures influenced how competitive tendering must take place, the high-level public officials were able to convey the impression to the political decision-makers that this tendering process would be the best course of action and easy to accomplish. Accordingly, they were able to persuade the politicians to unanimously approve their proposal to start the process.

The second sub-sequence started with the restructuration that took place in mid-2013, when public officials had to amend their initial call to submit bids from local companies into an international call. The amendment was necessary to comply with the financial thresholds set forth in the 2007 Act on Competitive Tendering and to cope with the austerity pressures that drove outsourcing. The head of the social and healthcare committee offered the following explanation:

At first, we wanted the competition to be local […]. Then we realised that the law was going to be against us […]. This obliged us to issue the official invitation to submit bids at EU level.

The chief procurement officer verified this account of the process and commented:

…the work of designing the invitation to submit bids was intensive […]. I participated as an expert in legal issues related to public procurement […]. I have done this many times in fields other than healthcare […] [and] the manager of elderly care did substantive market research to find out which model was the most suitable.

That manager mentioned in the above quote provided further explanation:

I reviewed public procurement models applied in other fields and designed a specific model for us for elderly care.

Previously, accountability relationships between politicians and public officials had been weakened owing to quality-related issues and increasing costs of elderly care. To preclude criticisms related to the evaluation of bids arising this time too, the procurement committee specified numerous metrics addressing the quality–price relationships in elderly care services and various types of service packages that each competitor had to include in the price that it offered to the city. The manager of elderly care services explained:

We included 200 criteria on quality of services and elderly care service packages that each company had to comply with before its bid could be considered […]. In the bid evaluation, we focused on the price levels that the competing companies offered.

A member of the social and healthcare committee explained how the decision to approve the winning bids took place:

Actually, the public officials did not give us full details of the evaluation process […]. But we knew that the key variable was the price level for each service package […]. We had to approve their proposal […] because once the competition process is launched, one has to go on with it.

The manager of elderly care services commented:

Because all the competitors had agreed to comply with our requirements specific to each service package […], the decision on the winning bids had to be based strictly on price levels […]. Actually, this was transparent.

The key agents – public officials – took control of designing and implementing the second competitive tendering by setting strict and specific criteria regarding their expectations of the process. In effect, this meant that external agents had become internal and that structures had been endogenised (Ahrens and Ferry, 2018). Note that the actions of public officials thereby influenced the structure of the competitive tendering process itself, which led to the expected outcomes of outsourcing of elderly care services. Importantly, the quantification of service quality seen here serves as evidence that creating and imposing accounting systems does not automatically lead to increased accountability (see Mutiganda et al., 2013), although risk management for contracts that involves formal performance indicators is articulated as a distinct outcome for the process. In combination with quantified reports, the interactions of agents influence accountability. In the case of Viking City, some of the external agents had become, in effect, internal, which made them no longer politically accountable.

All in all, three companies – one local and two international – won the second competitive tendering process. Many others unsuccessfully participated. The total budget allocated to the winning contracts varied within the range of around 20–40 million euros for each contract period, which varies in duration between 4 and 7 years, depending on the city's need to outsource elderly care services. Because the second competitive tendering process was successful, political accountability became stable and was appreciated by all parties. A vice-head of the audit committee explained:

The audit committee is satisfied with the outcome of the new competitive tendering process for elderly care […]. [For instance,] the chief procurement officer has done a good job in limiting the legal risks in procurement […]. The manager of elderly care services has saved on costs for elderly care […]. In fact, the elderly care budget was not overspent in 2014, which is very good […], and no warnings have come to our attention, from the media or elsewhere.

The financial manager confirmed the social and healthcare committee's budgetary position. A member of that committee provided further comments:

As you may know […], I used to be quite critical […]. However, I have learnt to trust our public officials in healthcare […]. For example, the manager of elderly care services is doing a good job now […] and the financial manager regularly informs us of how well, or not, our budget is doing […]. I feel confident that my political responsibility is not in danger anymore.

This finding ties in with our third theoretical argument, related to how change in internal structures eventually stabilises accountability. The change is evident in the public official–politician accountability relationship. Here, the agents started acting strategically: establishing trust among all active network members, using quantification metrics to evaluate bids and follow up on them, and maintaining active communication with members of the outside network to avoid criticisms. This strategic stance provided an appropriate setting to structure a comprehensive bidding and contracting process that considers all possible risk factors. This stance would also influence the politician–constituency accountability relationship, as the use of quality criteria and performance metrics was seen as a way to avert negative publicity related to service quality. We interpret these actions as strategic actions with stability as an expected outcome for the agents-in-focus. The combination of risk awareness and extensive measures facilitated critical reflection, which is essential to evaluating one's internal structures (Englund et al., 2011). Because the intended stability was in place from the outset and the agency was strategic, internal structures were critically mobilised and the outcome in terms of political accountability was stable. By demonstrating this, our study shows that accountability mechanisms are situated at the meso-ontological level. Trust among agents-in-focus is established at the ontic level and facilitates the processes that stabilise accountability relationships at the meso-level. However, the ontology of political accountability – that is, the intrinsic need to provide transparent accounts of what has happened between politicians and public officials at Viking City – remained compelling during the first and second competitive tendering processes.

Since new methods such as using performance metrics for service quality and shifts in agent positions (from external to internal) have the potential to influence accountability relationships, a question arises about trust in creating and maintaining such structures (Roberts, 2014). We conclude that political accountability is necessary for rendering the use of accounts ethical, particularly in the relationship between politicians and their constituencies. That said, because not all winners of the competition have started fulfilling their new contracts with Viking City, it is still too early to take for granted that political accountability will continue to appear as positive as it currently does. Table 2 provides a summary of our research findings obtained using the composite sequence analysis methodology.

Table 2 presents the key findings related to our theoretical arguments and the respective sequences of data analysis. The position–practice of the agents-in-focus, with taken-for-granted structures and stable accountability relationships of the first sequence, shifts between sequences and destabilises as external structures rapidly and irresistibly change (sequence 2). An illustration of this sequence provides an answer to the “why” question embodied in our sub-research questions 1 and 2: when external structures change, why and how does political accountability either change or not change? An analysis of sequence 3 provides an answer to our sub-research question 3: how does improvement to the internal structure stabilise or destabilise political accountability among agents-in-focus when external structures have not changed? As the agents-in-focus strengthened their internal structures and gained control of the outsourcing process, political accountability gradually stabilized (sub-sequence (1). This required fostering trust between the key agents. Next, critical examination of the new model followed (sub-sequence 2) and position–practice shifted to an international level, with tenders invited from large multinational companies, signalling permanent change in how the outsourcing process was conceptualised. Eventually, politicians and public officials could fully trust that the process is handled in a competent and professional way, stabilising the accountability relationships.

All in all, as the agents-in-focus interacted with each other during decision-making, political accountability at the abstract ontological level remained embedded in (meso and ontic levels) the processes of designing and implementing competitive tendering processes. In addition, the new structures related to the opening up of the healthcare markets were able to influence political accountability, which seemed to weaken at first. In the second competitive tendering process, the agents-in-focus learnt how to apply knowledge of the new structures (and to trust one another). In this process, we found the role of accounting technologies and, in particular, performance measurement and cost accounting.

5. Discussion and conclusion

This study was motivated by the need to extend the concepts of SST to address accountability. To this end, we drew on the longitudinal data from the pseudonymous Viking City's competitive tendering process. Our aim was to investigate political accountability around public-sector competitive tendering processes under conditions wherein agents are faced with change (here, in competition laws) that determines their external structure in a very fundamental manner. Moreover, the impact of austerity policies on our case organisation significantly increased over time, constraining decision-makers' policy options in a significant way. The research question looked at how changes in structures influence the agency of those involved in political accountability. We made three theoretical arguments related to this question: (1) when external structures change but internal structures are free to remain the same as before, the strong structuration outcome entails no change or a purely ceremonial change of political accountability; (2) when external structures change and internal ones are forced to adapt to the change in external structures, the strong structuration outcome leads to unstable political accountability; and (3) when no new changes occur in external structures but internal structure changes, the strong structuration outcome leads to stabilised political accountability.

The three arguments were operationalised through the following sub-research questions: (1 and 2) when external structures change, why and how does political accountability either change or not change and (3) how does improvement to the internal structure stabilise or destabilise political accountability among agents-in-focus when external structures have not changed?

Empirically, we approached these questions by employing composite sequence analysis. With regard to the era prior to the change in competition laws, we found support for the first theoretical argument, namely that should new laws not have forced public officials to open the competitive tendering to outside parties, the changes in political accountability would have been ceremonial at best. For our sub-research question 2 – why accountability is altered in situations of major change in external structures – we found that as long as the agents-in-focus still possessed conjunctural knowledge of the operations, they would draw from knowledge about the new structure that would eventually influence their dispositions. For many of the agents-in-focus, skills in competitive tendering were lacking, however, and conjecturally specific knowledge was unreflectively used, with little awareness of the risks associated with competitive tendering.

Turning to sub-research question 3, we found that as competition laws entered full force, agents began to act strategically to take into account various possible risk factors. Now, stability was sought as an expected outcome. Risk awareness coupled with extensive quantitative measurements and the utilisation of performance indicators facilitated critical reflection, which, again, is essential to evaluating one's internal structures (Englund et al., 2011). Because internal structures were critically mobilised, the outcome in the realm of political accountability was stable, with accountability mechanisms situated at the meso-ontological level.

Addressing another factor, Coad et al. (2015), Conrad (2014), and Stones (2005) conceptualised trust as being part of the external structure in position–practice relations. For instance, work by Cunningham and Nickson (2011) illustrates how procurement regulations can negatively influence co-operation. In our case, whereas the first competitive tendering process had the same overtones, the second yielded different results. In fact, our longitudinal study provides evidence of growing co-operation and trust between public officials and political decision-makers over time. Although media outlets still criticised going over budget and tried to frame it as a failure of competitive tendering to save costs, the relations between the political and managerial decision-makers remained close and collaborative. Our document analysis did not uncover any evidence of managers receiving warnings related to this or of associated voting by the board, etc. This enables us to extend the discussion by showing that trust can also emerge from a positive relationship between internal and external structures among the active agents that operate in different organisational settings, such as political and managerial ones. This stands in contrast to the views expressed in earlier works (e.g. Pellinen et al., 2018) that point to whether and how the accountability “blame game” ushered in by structural changes can be overcome with trust.

In fact, prior research has shown the importance of trust-building activities when accountabilities shift owing to irresistible changes in external circumstances (Väisänen et al., 2020). In our case, the two competitive tendering processes illustrate that a high level of trust between agents is needed if changes in accountability are to occur (Coad and Glyptis, 2014). Furthermore, as Coad et al. (2015) and Conrad (2014) have conceptualised, trust is a key part of the external structure in position–practice relations. Because new ways of reporting (involving performance measures and cost accounting) and new agent positions (shifting from external to internal) represent potential to influence accountability relations, a question of ethics in creating and maintaining such structures arises (Roberts, 2014). Importantly, accounting reports alone were not sufficient to establish accountability relations (Mutiganda, 2014). Accordingly, we emphasise in our conclusion that political accountability is needed to make account use ethical (again, because not all the companies that won a contract with Viking City have started performing it, it remains too early to assume that political accountability will continue to manifest as positively as it currently does).

This study makes several contributions to the literature. First, we take SST research further by examining the location of accountability in the quadripartite framework (Daff and Jack, 2018; Greenhalgh and Stones, 2010; Jack and Kholeif, 2007, 2008). Our findings dovetail with the discussion offered by Coad et al. (2015) on active agency embedded in ongoing structural relations and changes. By focusing on the factors that influence the actions of active agents during the implementation of a market mechanism (specifically, competitive tendering and political accountability at organisational and societal levels), our study serves as a constructive response to recent calls to analyse active agents and accountability in terms of a longitudinal process of structuration (Coad et al., 2015; Conrad, 2014; Roberts, 2014).

Second, we contribute to studies of accountability (Ahrens and Chapman, 2002; Conrad, 2005; Roberts, 2014) by conceptualising accountability at the abstract and meso-ontological levels, alongside its ontic empirical specificities (see Makrygiannakis and Jack, 2016). Our case study provides an illustration of the stabilisation of accountability when changes take place in the context and conduct of the agents-in-focus. The results illustrate how accountability weakens when position–practice relations come to dominate internal structures and the actions of active agents and demonstrate how agents can later shape the structures to stabilise their accountability relationships. However, our longitudinal analysis also revealed a positive relationship between active agents operating in different organisational settings, such as political and managerial ones, which is at odds with views expressed by some scholars (e.g. Pellinen et al., 2018), showing why and how the above-mentioned accountability “blame game” attendant to structural changes can be overcome.

Finally, our case study provides practical implications for work on competitive tendering (Cunningham and Nickson, 2011; Jones, 2013; Pîrvu and Bâldan, 2013; Mutiganda, 2014) in conditions of austerity policies (Ahrens and Ferry, 2015; Ahrens et al., 2020; Bracci et al., 2015; Ferry et al., 2019). In connection with this, we illustrated how the agents were able to act strategically to stabilise accountability (Järvinen, 2016). These contributions reveal SST's potential to go beyond serving as a mere so-called sensitising device, as it so often seems to be in traditional ST work in the structuring of empirical data (Englund and Gerdin, 2018). This is especially true with regard to external structures such as competitive tendering laws and austerity policies, which we found useful to treat as a category separate from agents' dispositions.

As with any field research, the findings of our study cannot be directly extended to organisations other than those analysed. The study does show, however, how SST (Stones, 2005) provides a systematic framework for conceptualising policy on organisational change, and it exemplifies the empirical research illustrating the applicability of this framework (Chan et al., 2010; Coad et al., 2015; Jack and Kholeif, 2008). An avenue for further work is connected with the argument by Stones (2005) that trust is one of the constitutive elements of position–practice relations. Although our findings point to a more extensive role of trust during structuration and suggest that trust can also emerge from the general dispositions of each active agent when that agent is confident about his/her relations with another active agent in specific contexts and settings, we feel that this issue would merit further research. In addition, further studies are needed to analyse the ethics of trust in the inter-organisational management of business contracts within public-sector organisations.

Figures

Figure 1

The quadripartite nature of structuration and the three abstract-concrete levels of ontology

Table 1

A summary of our research methodology and data analysis

Year and sequence	Theoretical arguments	External structure and change	Internal structure and change	Agency (action)	Political accountability
Up–2007	Argument 1	Independent	Unreflective Habitus	Routine	Intended but ceremonial
Sequence 1		Enabling
2007–2011	Argument 2	Irresistible	Unreflective	Non-routine (somehow disorganised)	Unintended, ceremonial, and problematic
Sequence 2		Constraining	Habitus failure
2012–2013	Argument 3	Enabling (Subsequence 1)	Reflective	Strategic (but not yet routine)	Intended and instrumental
Sequence 3		Irresistible (Subsequence 2)	Critical	Strategic (but not yet routine)	Non-intended and instrumental
2015	Follow-up of	Independent	Reflective	Strategic	Intended, instrumental, and gradually stable
Sequence 4	argument 3	Enabling	Critical	(but not yet routine)

Table 2

Summary of the links between theoretical arguments and research findings obtained using the composite sequence analysis methodology

Theoretical argument	Sequence of analysis	Position–practice	Political accountability
Argument 1 If external structures change but internal structures are free to remain the same as before, the strong structuration outcome entails no change or a purely ceremonial change of political accountability	Sequence 1. Up to the 2007 Act	Taken for granted external and internal structures Trust between agents-in-focus and local entrepreneurs is questionable	Political accountability is stable
Argument 2 If external structures change and internal ones are forced to adapt to the change in external structure, the strong structuration outcome leads to unstable political accountability	Sequence 2. Designing the first competitive tendering	Agents-in-focus face changes of external structures and must internally implement them Internal structures are weak Positioning among agents-in-focus is problematic Trust between agents-in-focus and the general public is under challenge from social media	Political accountability becomes unstable: the general public is more attentive to what has happened
Argument 3 If no new changes occur in external structures but the internal structure changes, the strong structuration outcome leads to stabilised political accountability	Sequence 3Sub-sequence 1	Agents-in-focus invite other agents to cooperate Agents-in-focus stabilise and strengthen their internal structures (using accounting and managerial performance metrics) External structures are now more familiar Trust between agents-in-focus and the general public is improving but weak Irresistible regulation obliges to change competitive tendering from national to international position–practice Managers- and politicians-in-focus can now fully trust each other but still face challenges in minimising costs and controlling the performance of outsourced services	Political accountability gradually stabilises: politicians feel safe as public officials regain control of the outsourcing process
	Sub-sequence 2		Politicians feel safer as public officials gain control of the outsourcing process
Follow-up of argument 3 If internal structures change but external structures remain stable, structuration outcome lead to a stabilised political accountability	Sequence 4Learning from previous errors to work better		Both politicians and public officials have good understanding of accountability in relation to managing outsourced services at lower costs