Converging and diverging pressures in PBB development: the experiences of Finland and Sweden

ABSTRACT

This paper aims to provide institutional explanations for the development of performance-based budgeting (PBB). Particularly, we explore the tensions between converging and diverging pressures. By analysing PBB in two countries with significant experience with it, Finland and Sweden, the research explains why we observe only evolutionary changes in actual budgetary practices. Indeed, the introduction of this practice is part of a global reform process undertaken over decades by public sector organizations. Nevertheless, the countries still struggle with PBB and, despite global converging pressures, diverging pressures shape its development, challenging its implementation and use, and paving the way for evolutionary changes.

KEYWORDS:

• Performance budgeting
• isomorphism
• institutional logics
• ambiguities
• central government

Introduction

The modernization policies of public sector organizations over the last few decades have been influenced by the massive growth of studies in New Public Management (NPM). Although its origins go back to the late-1980s, the NPM movement is still alive and reflected in several tools currently adopted by public sector organizations (De Vries and Nemec 2013; Hyndman and Lapsley 2016). Among the various performance-oriented reforms introduced in the spirit of NPM, the so-called performance-based budgeting (hereafter PBB) has been revived. This reform is clearly connected to the managing for results philosophy, part of NPM, because of its focus on measuring and managing results. At the same time, it has the unique feature of being concerned with the budgetary use of performance information (Robinson 2007). This issue represents the main potential source of benefits of the reform but, at the same time, it paves the way to lively discussions on its feasibility. Indeed, PBB is still a controversial issue because of the difficult linkage between its two main components, performance management and public budgeting (Ho 2011; Hou et al. 2011). In particular, although the relevance of public budgeting with its different potential roles, including its managerial function, has been recognized over the years, this topic is surprisingly under investigated (Anessi-Pessina et al. 2016; Sicilia and Steccolini 2017).

Therefore, the current research will contribute to the analysis of public budgeting by investigating its potential innovations with a public management perspective through the study of PBB. Specifically, this research investigates the development of the practice of PBB by building an eclectic theoretical framework, which draws upon three concepts of neo-institutional theory: isomorphism (DiMaggio and Powell 1983), institutional logics (Thornton, Ocasio, and Lounsbury 2012), and ambiguity theory (March 1978). Complementing each other, these three concepts allow us to build institutional explanations for the development of PBB, investigating the pressures influencing it. The key research question is whether and how PBB development can be explained as a tension between pressures for convergence and divergence, both exogenous and endogenous (Kuhlmann 2006; Pollitt 2002).

On the one hand, public sector administrations worldwide have included the practice of PBB in their reform agenda generating a global convergence towards this idea (e.g. OECD 2007; Pollitt 2002). Isomorphic pressures, conveying exogenous pressures, can be considered the converging pressures leading to PBB adoption across countries (DiMaggio and Powell 1983).

On the other hand, such convergence is contrasted by a wide range of approaches to this change. Indeed, the budgetary use of performance information represents the heart of PBB but, at the same time, the source of its variation. This is first visible in the multiplicity of definitions of PBB (e.g. Joyce 2011). Some authors refer to it as ‘procedures or mechanisms intended to strengthen links between the funds provided to public sector entities and their outcomes and/or outputs through the use of formal performance information in resource allocation decision-making’ (Robinson and Brumby 2005, 5). However, a range of possible links between performance information and budgeting decisions can exist, thus paving the way for different forms of PBB, which have been the object of semantic splitting (Joyce 2011) and analytical classifications (Bleyen et al. 2017; OECD 2007). Consequently, the practice is characterized by a high variation of labels, interpretations and approaches (Curristine 2005), which seem to challenge the existence of a global convergence towards PBB.

This research investigates the pressures leading to divergence other than convergence to provide a comprehensive explanation for the institutional development of PBB. In light of the recognized need for combining the analysis of exogenous and endogenous forces to explain institutional development (Kuhlmann 2006; Liguori and Steccolini 2011), this study focuses on several endogenous pressures, here represented firstly by the specific country’s logics influencing the modernization of public administration (macro-level phenomenon) as a response to the exogenous pressures for convergence. Second, since in the case of PBB significant challenges to its implementation and use are posed by the ambiguity of public budget and performance, the research focuses on the ambiguities of PBB (micro-level phenomenon), which have been significantly under explored (Moynihan 2005).

By adopting this original approach, the research aims to provide enriched and comprehensive explanations of the development of PBB, contributing to this growing body of knowledge. A significant part of the literature has investigated the status of the practice emphasizing its current challenges and problems (e.g. Melkers and Willoughby 2001; Sterck and Scheers 2006), but only more recently has paid attention to its development (Clark, Menifield, and Stewart 2017; Mussari et al. 2016). The current study will contribute to the public management literature by explaining PBB development, analysing the contrasting pressures influencing such development.

The authors intend to achieve this purpose by studying the development of PBB at the central level of government in two Nordic countries, Finland and Sweden. By reviewing previous studies on PBB (Mauro, Cinquini, and Grossi 2017), it emerges that the majority of the comparative papers investigated the Anglo-Saxon world. This paper purposefully selects Finland and Sweden as a research setting because these countries can provide original insights. In addition, they have a significant and long experience with PBB, thus providing a suitable ground for studying the development of the practice.

In the following section, we discuss the exogenous pressures for PBB convergence. After which we provide insights into PBB, underlying the existence of both convergence and divergence in its development. We then present the theoretical framework used for analysing these pressures, which is followed by the introduction of the methodological design of the paper. Then we present and discuss PBB development in the two countries, and finally we make concluding arguments.

Exogenous pressures for PBB convergence

Since the 1980s, global trends emphasizing performance-oriented reforms have been supported by international organizations (such as OECD, World Bank, International Monetary Fund), and studied by public sector researchers creating institutional pressures towards the adoption of a package of reforms built on the measurement and management of performance. Accordingly, the global reform movement inspired by NPM principles (Hood 1991) has stimulated political decision-makers to adopt similar institutional arrangements across nations (Pollitt and Bouckaert 2011), creating a global convergence towards similar ideas and reform agenda (Pollitt 2002). Among others, the public budget, as the main economic policy document for the government and a key device for managers, has been called for adjustment even in response to recent economic crisis (Peters 2011), and thus budgeting innovations represent a relevant component of the broader reform movement linked to the performance movement.

In a globalized world, which has been strongly affected by economic and financial crisis and has shared the need for reform, these exogenous pressures have motivated the organizations to adopt a similar reform package through isomorphic behaviours, that is each organization tends to resemble the others for the sake of increasing their legitimacy and conforming to socially accepted rules and practices (DiMaggio and Powell 1983). Hence, the agenda of many countries worldwide converged towards PBB (OECD 2013; Pollitt 2002). The idea of convergence comes from the institutional theory that has emphasized the process of isomorphism, whereby organizations adopt similar procedures to gain legitimacy as result of coercive, mimetic and normative pressures (DiMaggio and Powell 1983).

Specifically, in the context of crisis and uncertainty that has shaped public sector organizations over the last few decades, governments have tried to model themselves on the others operating in similar contexts through mimetic processes. In particular, the example of the Anglo-Saxon world and its adoption of PBB has influenced the development of the practice in other OECD countries (Van Dooren, Bouckaert, and Halligan 2015). Further, coercive and normative pressures for meeting the expectations of international organizations, forums and professional networks, and to respond to economic demands and citizen pressures (Van Gestel and Teelken 2006; OECD 2007), have motivated governments to try to adopt a textbook model of PBB. Organizations such as the OECD (e.g. PUMA project) and the World Bank are committed to a view of international convergence towards some ideas and models, such as PBB, that they try to foster (Curristine 2005; Hood 1995; Robinson 2007). However, despite the global convergence towards the idea of PBB and its recognition in the literature, we need to go below this level of superficial convergence and find out what happens in specific contexts.

PBB development between convergence and divergence

The review of the literature on the practice shows that PBB represents only one of the labels most often attached to the attempts to link performance information and budgeting process. This practice is also known, for instance, under the labels ‘performance budgeting’, ‘output-/outcome-based budgeting’, or ‘results-oriented budgeting’. These labels have been used interchangeably in some cases (Klase and Dougherty 2008; Melkers and Willoughby 2001), while in other cases scholars have underlined the relevant differences existing among them (Martin 1997). In light of the type of use and the extent to which performance information can be used during the budgeting process, several classifications have been developed (Bleyen et al. 2017; OECD 2007). The OECD (2007), which defines PBB as a form of budgeting that relates funds allocated to measurable results, distinguishes three different categories of PBB: presentational performance budgeting, performance-oriented budgeting and direct formula performance budgeting. According to an increasing degree of sophistication, performance information is first produced and reported in the budget documents, second linked to proposed future performance or past realized performance in an indirect way, and third it is directly and explicitly linked to resource allocation.

In addition to this variety of labels and classifications, the trend among industrial countries to bring a stronger performance orientation into public financial management has resulted in a variety of approaches (Curristine 2005). This apparently high variety points out different decisions about this reform and consequently its different forms and techniques. In this regard, taking into consideration Pollitt’s (2002) classification of the different stages of convergence in public sector reforms, PBB reforms may include: (1) discursive convergence, in other words the conceptual agenda of talking about PBB converges across countries; (2) decisional convergence, that is countries make similar decisions about PBB systems; (3) practice convergence, as countries and institutions adopt similar forms and techniques of PBB; and (4) results convergence, wherein PBB systems in different countries start to produce similar intended impacts.

Empirical evidence suggests that beyond the convergence towards the idea of PBB, the development of PBB varies across countries, policy sectors and over time (e.g. Hou et al. 2011; Lee and Burns 2000; Melkers and Willoughby 1998). A significant body of literature has analysed the challenges connected with PBB implementation focusing on the current common problems and issues (e.g. Melkers and Willoughby 2001; Sterck and Scheers 2006), and only more recently the historical development of PBB has been analysed, investigating its adoption and re-adoption over time and its spread across countries (Clark, Menifield, and Stewart 2017; Schick 2014). Yet, there has been little effort in analysing the pressures that explain such development. In this context, a comparative analysis of PBB in multiple contexts can play a pivotal role to shed new light on PBB.

The majority of the comparative papers on this reform investigated mainly the Anglo-Saxon world, as shown by a recent review of the literature on the topic (Mauro, Cinquini, and Grossi 2017). For instance, Arellano-Gault and Gil-Garcia (2004) compared the implementation of PBB in three Latin American countries, demonstrating that, without a clear picture of the political feasibility, it may lose its direction and mission. Lee and Wang (2009) used data from China, Taiwan, and the United States to show how different purposes and implementation strategies may have different impacts on government entities. Sterck and Scheers (2006) analysed the PBB reforms in seven OECD countries (Australia, Canada, the Netherlands, New Zealand, Sweden, the United Kingdom, and the United States) and demonstrated the very limited use of performance information in the political budgetary decision-making process or in the legislative oversight function. The major impact of PBB reform appears to be situated in the internal management of departments and agencies. Similarly, through a comparison of PBB between Australia, Canada, the Netherlands, and Sweden, Sterck (2007) demonstrated that PBB changes budget formats rather than functions, thus pointing out a limited use of performance information. Bleyen et al. (2017) analysed the incorporation of performance information in local government budgets in 10 European cases (Belgium, Czech Republic, Germany, Ireland, Italy, Lithuania, the Netherlands, Norway, Poland, and Slovakia), demonstrating the existence of similar learning trajectories but, at the same time, differences in performance structure and span of performance. Their focus has been mainly on the current detailed features of PBB. As these studies show it is lacking an analysis of the pressures explaining PBB development. Mussari et al. (2016) analysed design, trajectories of reform and implementation of performance budgeting in Germany, Italy, Lithuania, and Norway by developing a framework for understanding the evolution of the reform and its key elements in several contexts. The current study aims at contributing to this approach by analysing the pressures explaining the development and the ambiguities of the practice, investigating the existing tension between contrasting converging and diverging pressures.

Endogenous pressures for PBB convergence and divergence

In light of these considerations, exogenous pressures leading to convergence are not sufficient to explain the implementation and use of PBB. The outcomes of the change process stimulated by those pressures may vary across governments (Liguori and Steccolini 2011), implying that there might be differences in how various countries that implement and use PBB. For exogenous pressures to bring fundamental transformation, an endogenous evolution should prepare the groundwork (Greenwood and Hinings 1996). A certain degree of receptivity to change (Lapsley and Pettigrew 1994) is necessary for governments to successfully modify their budgeting systems. Such receptivity is influenced by the tension between exogenous forces and endogenous forces active within a specific country (see also Greenwood and Hinings 1996; Ter Bogt and Van Helden 2000; Argento, Peda, and Grossi 2018).

The country-specific context can explain the extent of change (McLeod and Harun 2014). The situations at the macro and micro levels of a specific country influence how and why PBB is implemented and used in a certain way. At the macro level, new logics can emerge and this contextual discontinuity may become a trigger for changing established practices (Mutch 2007). A budgeting system, as an accounting system, serves a given institutional logic (Anessi-Pessina and Cantù 2016), which is intended as an historical pattern of practices, assumptions, values, beliefs, and rules (Thornton 2004) that guides actors’ behaviour through its assumptions and rules.

Until the 1980s, the public budget mainly served the logic commonly known under the label ‘traditional Public Administration’ (PA), which has dominated the public sector from a cross-national perspective. Accordingly, cash- or commitment-based budgets have been emphasized as the preferred way to allocate resources in line with a bureaucratic and highly centralized structure, with the focus on the authorization function (Anessi-Pessina and Cantù 2016). Afterwards, since the 1980s, the public sector moved from the traditional logic of PA towards the logic of NPM (Hood 1991), an umbrella term used to indicate a wide set of principles and tools inspired by the private sector and strongly built on the buzzword of performance. In this context, public budgeting started to reflect this logic by including in the process the measurement of performance. Changes in the predominant logic motivate and explain institutional changes (Hyndman et al. 2014; Thornton 2004). Institutional logics affect the decisions of an organization by focusing the attention towards issues and solutions in line with the dominant logic (Thornton 2004). For this reason, it is crucial to analyse the predominant logic at work in the context under investigation to explain the development of PBB, since a logic shift can pave the way for a new path of changes (Torfing 2001). Hence, in this macro-level analysis, we analyse the modernization trends of public administration, reflecting the predominant logic at work and in turn reflected in the reforms introduced, where similar trends can stimulate convergence towards similar ideas and decisions.

The adoption of PBB would require the shift from a PA logic to a NPM logic, but in parallel, over the last few years, a new logic labelled New Public Governance (NPG) (Osborne 2010) emerged as potential response to NPM-oriented developments (Almqvist et al. 2013), although it is debated whether it is possible to shift completely from one logic to another or whether multiple logics can coexist (Lounsbury 2008). This research investigates which is the predominant logic at work shaping the development of PBB in Finland and Sweden. This approach can contribute to explain why traditional models of public administration can persist in making slow and incremental changes, as often shown by the analysis of public sector reforms (Van Gestel and Teelken 2006). Evolutionary changes are mostly expected to take place in this context. Greenwood and Hinings (1996) distinguish between revolutionary and evolutionary changes. In terms of process, revolutionary changes happen quickly and involve all parts of the organization simultaneously. Evolutionary changes occur slowly and gradually, consistently with an historical perspective according to which current decisions and actions are strongly path dependent.

The implementation and use of the new budgeting practice of PBB also involve issues that are subjective. This means that the role of key actors who enable changes at the micro level is of utmost importance. In the context of PBB, many ambiguities are connected to how performance information can be used by actors to allocate resources to different public activities and policies (Moynihan 2005; Noordegraaf and Abma 2003). The development of PBB is thus shaped by the inherent ambiguities of PBB that public administrations need to cope with when implementing the reform. Sometimes, solutions, such as a reform is expected to present, may solve existing policy problems, while at other times they may create new, previously unanticipated problems (Vakkuri 2010). In the context of PBB, many of these problems stem from the practice itself, because of the ambiguity connected to public budgeting and to the performance information used in public sector activities (Moynihan 2005; Noordegraaf and Abma 2003).

Ambiguity can emerge in several cases as an effect of different factors, like unclear goals, unclear and uncertain links between decisions and actions, and information about the past being difficult to understand and interpret univocally (March and Olsen 1976). Thus, different types of ambiguities can be detected such as: ambiguity of preferences, due to vague, contradictory and changing preferences; relevance, due to the lack of coherence between problems, solutions and actions; history, due to imprecise and limited calculation of possible consequences of actions; and interpretation, due to the difficulties in attaching unique meaning to information (March 1987). In the specific context of PBB, two main kinds of ambiguities can be considered influential in its development: ambiguity of relevance and interpretation.

First, PBB requires the identification of goals and objectives in order to define what to achieve and how to influence the budgeting process accordingly. One of the main related challenges is to focus on all the relevant areas of public sector performance due to the bounded rationality of attention. Above all, it is difficult to determine what is relevant. Public budgeting is closely associated with individual, political and managerial preferences for making choices about societal change since it implies a process of allocation of relative values (Wildawsky 1986). As a consequence of this ambiguity of preferences, performance measurement information cannot determine such choices (Vakkuri 2010). The information may only assist in understanding the alternatives and limits for budget allocations. Consequently, with specific reference to the development of PBB, it becomes challenging to establish coherent linkages between relevant problems, solutions and actions – ambiguity of relevance (March 1987). Relevance is far from unequivocal in public sector budgeting processes (March 1987), thus representing the key challenge in the allocation of resources based on performance information.

Second, PBB requires the interpretation and use of the performance information produced and collected in order to influence the budgetary decision-making, but interpreting data and understanding cause–effect relationships between public interventions, funds invested, and outputs/outcomes produced or to be produced are highly complicated – ambiguity of interpretation (March 1987). This applies to most public organizations where valuation of activities cannot be conducted through market prices. In principle, it may be easy to accept the link between performance and budgeting, but in public agencies the actual ways of applying this principle are limited by complicated mechanisms of social modelling, where defining objectives, measures and resource needs is the result of a complex process of negotiation.

These ambiguities are expected to influence PBB reform. On the one hand, public sector reforms with a significant level of complexity, such as PBB, are subject to constant mechanisms of interpretation and re-interpretation of what is ‘useful’, ‘feasible’ or ‘politically desirable’ (March 1978). It could be argued that coping with complicated public sector reforms would incentivise countries to adopt similar reforms or apply reforms in a similar way. As Pollitt argues (Pollitt 2002), the stage of decisional convergence is fairly easy to identify, which may lead to a biased understanding of how isomorphic processes actually work in the context of government reforms. However, it is also one important source of evidence to create theoretical understanding on the development of PBB systems. On the other hand, empirical evidence often shows various results in terms of decisional and practice convergence. In light of these considerations, this analysis investigates the impact of practice ambiguities on the development of PBB.

The framework for analysis

The adoption of PBB can be considered the effect of several exogenous pressures for convergence represented by the isomorphic pressures towards performance-oriented reforms at work over the last few decades. As such, its implementation and use are shaped by endogenous pressures at work in each country, at the macro and micro levels. At the macro level, the institutional logic(s) influencing the modernization policies can reflect how each country has reacted to the isomorphic pressures towards the need for changing the public administration. This is made visible by the reforms adopted and the changes introduced over the years. At the micro level, PBB introduced new challenges to cope with because of the intrinsic ambiguities of its key components, public performance and budgeting. While the isomorphic pressures would support the development of similar reforms, the specific country logics and the practice ambiguities can determine convergence, but also a divergence in terms of decisions, implementation and use of the practice. Thus, the current research aims to investigate the development of PBB as the tension between exogenous and endogenous pressures for convergence and divergence (Figure 1).

Figure 1. PBB development as tension between convergence and divergence.

Research context and design

The Nordic countries have been at the forefront in developing public sector systems and adopting public sector reforms, and their similarities make them different from the rest of Europe (Rexed 2008, 111).

The selection of the cases has been guided by the aim to investigate less explored but well-developed contexts in order to fulfil the research purpose explaining the development of PBB. This research focuses on two Nordic countries, Finland and Sweden, both with a long and above the OECD average experience of PBB, which also distinguishes them from other Nordic countries such as Denmark and Norway that have adopted the reform later on and with results below the OECD average (Clark, Menifield, and Stewart 2017; OECD 2013). In both contexts, the research investigated the development of PBB at the central level of government, because of the need to address several challenges deriving from financial and economic crises, political shifts, and governance changes.

The Finnish system of public administration has historically been a dual two-tier system including a central government and highly autonomous local governments (Strategic Programme of the Finnish Government 2015). The central government, comprised of the Prime Minister and currently 12 ministries, has an important role in regulating service provision through legislative and grant systems, but a more limited role in the provision of basic public services. It is responsibility of the agencies, organizational bodies working under the control of ministries, to deliver public services. This excludes the police force, army, nationwide policy sectors, and universities, where the central government is the main provider, organizer and/or financier of services.

The Swedish system of public administration has three levels of government: national, regional and local. At the national level, which is the focus of this research, the Swedish model is characterized by a high degree of decentralized responsibilities from the central government to the agencies (Pollitt and Bouckaert 2011). The central government, made of the Prime Minister and currently 10 ministries and 22 ministers, provides agencies with the tools and conditions to implement decisions made by the government and the parliament. The government agencies thus carry out the activities and deliver public services (Swedish Government Offices 2013).

The empirical basis of this article consists of data from documentary analysis and semi-structured interviews. The documentary data include research articles and official reports concerning Finnish and Swedish central governments and their reforms. The documents were retrieved from the websites of the central government administrations (ministries and agencies) in both countries and integrated with the documents shared by interviewees. These archival material cover the period of development of the reform in both the countries, since it includes the different publications introducing the reforms and discussing them (e.g. handbooks on performance management published by the Ministry of Finance), thus providing key insights into the reasons and contents of the PBB reform.

Interviews were then conducted at the Ministries of Finance in Finland and Sweden and at the Swedish National Financial Management Authority (ESV) in order to gather information on PBB development. The interviewees have been selected because they are well informed on the development of PBB in light of their long and direct experience with the practice. These interviews were integrated with those of sector ministries and agencies to deepen the understanding of the practice ambiguities connected to PBB. In particular, these interviews were conducted at Finland’s Ministry of the Interior and Ministry of Economic Affairs and Employment, and at Sweden’s Ministry of Justice and the Statistics Sweden agency (Annex 1). The three ministries provide examples of PBB application (or lack of) in contexts directly influenced by high societal pressures such as migration (which is under the responsibility of the Ministry of the Interior in Finland and the Ministry of Justice in Sweden) and unemployment, while the Swedish agency’s activities are indirectly affected by these pressures. Appendix 2 reports the interview outline (see the lists of interviews in the Appendix 1). In total, 485 min of interviews were tape-recorded and interpretatively analysed according to the adopted theoretical approach. The in-depth study of documents and interviews aimed to identify the main pressures influencing the development of PBB, and group and categorize them according to the framework.

Endogenous pressures for PBB convergence and divergence

Macro-level analysis in Finland and Sweden

Since the 1980s, in Finland and Sweden, performance management reforms have been initiated by the central government.

The Finnish institutional context at the time was that of public corporations. The basic objective of the legislation on state corporations was to implement a new kind of accountability, where more independent units were responsible for attaining social goals set by the parliament (‘social efficiency’) and business goals set by the Finnish government (business profitability). New and more independent business-like agencies were created to address the problem of the lack of financial accountability. The logic behind these changes was to make independent agencies accountable for their performance and subject to growing competitive pressures to be effective and efficient. A programme was launched to make the government ‘less bureaucratic’ and ’more efficient’ (Bill 349/01/1988). These attempts to overcome bureaucracy and pursue efficiency reflect the shift from a PA logic towards a NPM logic. This shift has had strong repercussions on the budgeting process (Anessi-Pessina and Cantù 2016) since it influenced the central government budget reform (1989), replacing historical line-item budgeting with lump-sum budgeting. Rather than focusing on detailed regulation and ex-ante financial control, part of a more traditional PA logic, the lump-sum budgeting is interpreted as an extension to improve institutional accountability. Hence, organizations and their managers were given more freedom to influence the management systems of their agencies. The greater autonomy, on the one hand, and the higher degree of accountability, on the other hand, were clear signals of a predominant NPM logic.

It was already a long time ago, it was 20 years back, when performance management was introduced in the Finnish public administration ― late 1980s and beginning of 1990s … the focus was to shift the focus from administrations to what the actual results are, to the end users, or to the citizens, and more so of course even to the outcomes and impacts, from inputs to outputs, outcomes and impacts. (Interview A)

Similarly, in Sweden, following the several attempts made during the 1960s and the 70s to improve cost-consciousness and introduce performance management (Wilks 1995), during the 1980s, after and because of the economic crisis, performance management reforms started (Küchen and Nordman 2008; Modell and Grönlund 2007). The NPM logic became predominant as the introduction of management–by-objectives, heavily focused on output, processes and efficiency (Sundström 2006), testifies.

The activities part of performance management-based logics, such as setting goals, assigning resources, appointing managers, and holding them accountable for results (Rexed 2008), were predominant:

The Swedish government system was not really efficient and effective in the past. In the 1980s, there was a strong need to introduce a wide performance management reform, not only PBB as such, for all ministries and government agencies. This change was really radical and revolutionary, and finally the focus was no more on inputs but on outputs and outcomes. (Interview F)

According to the perceptions of the actors interviewed in Sweden, the shift from a traditional PA logic to a NPM logic was thus revolutionary, and the consequent tendency to measure performance was so strong that nowadays it is considered part of the traditional way of governing. Flexibility and performance became keywords in Sweden, balancing the decentralization of tasks with higher result controls, as occurred in Finland. The intention was to improve efficiency and effectiveness, but at the same time overcome the Swedish ministries’ limited capacity in managing their organizations and control public agencies (Küchen and Nordman 2008; Sterck 2007).

At that time, there was strong political pressure from the ministries, which were interested in steering and controlling the government agencies in term of activities, resources and goals. The largest changes were not in budgeting but in the reporting systems of the agencies to the government. The innovations were related more to the agencies. (Interview F)

As emerges from our interviews, a strong focus has been on control and accountability, emphasizing the need to meet strict reporting requirements and thus recall values that are part of a more traditional PA model. In the late 1980s, budget reforms intended to encourage and enable long-term planning moved away from incrementalism techniques and replaced rule-based input control systems with an emphasis on results and outcomes (Wilks 1995), consistent with the shift from a PA logic to a NPM logic. The budget process was managed through expenditure ceilings and subsequently accrual accounting was introduced. In this context, budgeting reforms and decentralisation trends contributed to the development of PBB.

In Finland, according to the decentralized structure of the public sector, the budget proposals presented to the Ministry of Finance by the sector ministries were based on negotiations between the sector ministry and its subordinate government agencies. The higher flexibility granted to agencies had to be counterbalanced with higher accountability and control for results, especially during times of economic and financial crisis. Thus, sector ministries and their agencies achieve a mutual understanding of the services to be delivered and the level of performance to be achieved with the agreed level of resources, which is reported in the so-called performance agreements (Salminen and Viitala 2006). This latter experiment of the government concerning the introduction of performance agreements was designed to clarify the boundaries of accountable entities, representing the core element of performance management and thus of PBB. In 1988, when the Ministry of Finance in Finland introduced the three Es principle (economy, efficiency, effectiveness) to be used in measuring organizational performance, a shift from a focus on operations to one on results was perceived as a ’revolutionary moment’ (Interview D), confirming the shift towards a NPM logic. This was reinforced in 2005, when the Ministry of Finance established a new conceptual scheme for providing performance measurement information: the so-called performance prism.

PBB was then designed and introduced in Finland in order to strengthen ex-ante and ex-post accountability by supporting the budget proposal with performance targets and reporting the results in the final accounts. However, the achievement or the failure to achieve the stated results was not linked to any rewards or sanctions (Blöndal, Kristensen, and Ruffner 2003).The PBB system was considered a necessity and to be completely implemented by 1995, but in practice its development took only marginal steps. On the one hand, the state budget was seen as a holy book, which is why politicians and public managers were extremely hesitant to change the status quo in budget allocation systems and through political pressures tried to maintain the existing sectoral-based budgeting structure. Thus, the role of performance information in PBB remained marginal in practice. The system of performance agreements between ministries and agencies was considered appropriate and sufficient to support the idea of PBB. On the other hand, the perseverance of the traditional approach to budget and the pressures on controls make central government agencies develop ‘strategy processes’, influencing also the budgeting process by this development, including stages of ‘vision talk’, ‘strategic priorities’ and ‘monitoring indicators’ (Vakkuri 2010). However, the allocation mechanism was not modified.

In Sweden, in 1997, the performance budgeting project (so-called VESTA) introduced a non-financial activity structure into the budget bill, although this activity structure was implemented only in 2001 (Tagesson and Grossi 2015). The aim was to allocate resources in accordance with political priorities and thus to relate planning and budgeting processes, although the budgetary instructions were written with a textbook logic and not with a focus on information needed in the budget process of the Ministry of Finance and other ministries (Czarniawska-Jorges and Jacobsson 1989). However, the new structure consists of policy areas and programmes for which objectives, outcome budgets and costs are defined (Sterck 2007). A performance dialogue takes place once a year between each state agency’s management and the responsible ministry with the basic purpose of following up each agency’s work in the previous year and discussing activities in a future-oriented perspective (Küchen and Nordman 2008).

PBB, like other performance management methods and techniques, was implemented in Sweden with gradual and incremental changes, and expected to be improved through pilot projects and evaluations (Sundström 2006). In practice, there were no key changes in the budgets:

The budget system was changed later, during the 1990s, and after this it remained quite stable with only small changes. The actual budgetary procedures were not really revised or reshaped after this reform. Consequently, the changes were gradual and adaptive. The changes were not radical and mainly related to the broader context of performance management reforms. (Interview F)

Therefore, the implementation and use of PBB was not as noticeable as expected.

Micro-level analysis in Finland and Sweden

The analysis of PBB documentation and interviews, both in Finland and Sweden, shows some important problems of using performance information in the budget allocations, as summarized in Table 1.

Table 1. Ambiguities shaping PBB development.

Finland	Sweden
Ambiguity of relevance
What is controllable is thus relevant
• Political logic vs. managerial logic	• Political logic vs. managerial logic
• Lack of coordination across processes and organizations	• Indicators decoupled from real performance
• Indicators decoupled from real performance
Crisis management
• Performance measurement and management logic challenged by unexpected events and societal problems	• Providing public services vs. measuring public performance
Ambiguity of interpretation
Does ‘A’ always lead to ‘B’?
• Abundance of information	• Call for integrating different sources of information
• Lack of learning process
• Call for networking and coordination	• Call for a balanced analysis of external and internal factors
‘Public’ matters
• ‘Papers because of papers’ and not of reality	• Not feasible and inopportune mechanical links between performance information and budgets

First, identifying what is relevant represents the first challenge – ambiguity of relevance. There are a number of factors explaining this difficulty. One, exemplified by the Finnish case, is that ‘Ministries are very strong leaders in their administrative fields’ (Interview B) and they focus on their respective administrative field, making it difficult to identify relevant areas of intervention from a comprehensive point of view and even more difficult to change money between different administrative fields. In both cases, the political interests, which represent the political logic continuously evolving because of changes in political systems, interfere with PBB managerial logic, which rely on a constant measurement of performance in a way suitable for comparing results over time and consequently affecting decision-making.

Further, in Finland and Sweden, the problem of controllability of performance leads administrations to set goals that they can completely control because of the fear of being judged and punished for objectives out of their control. This can result in identifying indicators that are not particularly relevant nor opportune or that can produce distorted effects, as clearly pointed out by the Swedish agency interviewed:

When you set performance indicators you always run the risk of doing well according to the performance indicators, and if you are not clever enough to choose indicators you run the risk of encouraging the wrong type of behaviour. There has been, for instance, a debate about police. There have been a lot of indicators in terms of how many alcohol tests they do on the roads, and it has sometimes led the police to do tests where they knew there were a lot of people and not very many drunk, because if they caught someone drunk they have to report it and it takes time. (Interview H)

Therefore, indicators are not always suitable for depicting real performance and they are used in a decoupled way from reality. As clearly outlined in the empirical evidence in both cases, it is not easy to conceptualize and measure different aspects of performance, and it is very difficult to evaluate the significance of government’s contribution to positive or negative development of societal changes. Indeed, many areas of government activities are difficult to measure such as social and health care, education, research and culture, as well as impacts on public security, unemployment and immigration. According to the interviews, it is hard to quantify public performance. An example is that of employment policies in Finland. In this field, unemployment rates have often been used as key indicators although considered unsuitable for evaluating employment policies. Similar problems are expressed by the Statistics Sweden agency, whose indicators measure, for instance, the number of published reports but not their quality or their achievable goals. Consequently, performance indicators appear decoupled from real performance.

Due to these ambiguities, there is a tendency to focus on the most easily measurable and comparable metrics between agencies and ministries, where the degree of accuracy can range from detailed and concrete indicators to abstract and vague declarations.

Performance budgeting ‘sounds like an easy system, but it is not’ (Interview D), and, more broadly, as recognized by the experts of the Finnish Ministry of Finance, ‘Performance management systems as such create problems’ (Interview B). An example is provided by the Finnish migration office. Despite its attempts to identify what to measure and how, the unexpected flows of immigrants make it highly difficult to respect initial performance agreements. The societal nature of the problem challenges the logic of measuring performance. In this context, complicated by the difficulties in analysing the environment – ’We do not know how many refugees will come during the next 6 months’ (Interview D) – the use of performance information may be perceived as not useful and the measurement of performance as a bureaucratic complication – ‘We are doing papers because of papers and reality is something different’ (Interview D).

The problem is shared by the Swedish actors involved in the same field. The migration emergency ‘was such a big task that we did a sort of “crisis management’’ and the main intention was to “take care of people’’’ rather than measuring results (Interview G). This example clearly shows the tension existing between measuring public performance and providing public services, where the latter should prevail but is considered inconsistent with the former under emergency conditions.

Then, the information produced should be interpreted and evaluated for use. Among the numerous measures produced, useful information from different sources and departments should be selected to avoid the risk that the abundance of information can then impede an effective use of information. Then, such information should be interpreted. These issues point out another key challenge – ambiguity of interpretation. In this regard, a learning process whereby it is possible to fully understand what the information says to explain poor performance appears lacking. To overcome this issue, proposed solutions have been identified in the need for networking and collaborating across horizontal units in Finland, while in Sweden more attention has been paid to balanced analysis of external and internal factors by relying on different sources of information.

Following the difficulty in understanding poor and good performance, their budgetary consequences are not uniquely defined. Interviewees agree that a mechanical resource-result link could not work and, according to some of them, it is even better to not be implemented at all. As the analysis of the Swedish context has underlined, ‘An agency can have very bad results last year, but there can be a pressing issue, which means they need more money for the next year’ (Interview G). However, it could also happen that ‘An agency can say, “We do not have enough money; this is why we are not effective” and the government can say, “We think you can work more effectively within your means”’ (Interview G). In this case, a fight for resources can undermine the logic of PBB.

Generally speaking, the practice ambiguities contributed to move the attention of the government towards different changes to implement. In Finland, the government focused on the need for managing horizontality. Service delivery travels across ministries, departments and agencies’ boundaries, requiring flexibility, coordination and collaboration among the different administrative silos – a need strengthened by the organizational restructuring (ministry and agency mergers). The attention paid to these new issues signals limited current efforts in PBB development and reflects the development of new logics such as NPG. In Sweden, after the political elections, the new government, in light of the strongly perceived limitations of NPM, fostered the public agenda towards a different orientation, built, as pointed out by interviews, on governance by trust, as sponsored by the logic of NPG, where the slogan ‘governance by trust’ reflects the importance of autonomy and flexibility reducing controls instead of increasing the requirements.

Discussion: tension between convergence and divergence

Both Finland and Sweden have accepted and introduced the package of cultural reorientation towards performance management, following the global trends in this direction, as response to the overall imperative for economizing the public sector. The powerful NPM discourse of ‘economizing the public sector’ and ’increasing value for money for taxpayers’ (Hood 1991) has required fairly comprehensive decentralization of decision-making powers, and instigated problems of demonstrating performance in public agencies. The central governments in Finland and Sweden, moving from a traditional PA logic to an NPM logic, introduced managerial tools and techniques designed to strengthen accountability and improve efficiency and effectiveness by managing flexibility and performance. The governments have thus developed cost accounting, performance measurement, and accountability systems, pointing out a converging path towards these changes. Hence, PBB could be expected to be only a small marginal extension. This paper demonstrates that, in the context of performance management reforms, budgeting is different.

The decentralized government structure has called for reinforcing accountability and political control by reporting results rather than by deeply reforming the budget. After the initial changes to budget processes (e.g. lump-sum budgeting and expenditure ceilings), politicians and a number of public managers have shown to be less willing to disturb the status quo of the budget, which is why the radical development of PBB systems may be in most policy areas more talk than action (Brunsson 1989). Despite the existence of converging pressures towards NPM, in general, and PBB, more specifically, and despite the coherence between the managerial logic implied by PBB and the predominant NPM logic, the practice of PBB has not completely replaced the existing systems, as expressed by the interpretation of the public budget as a holy book. Further, the shift that occurred in the management style has not been completely reflected in the budgeting system, where the pressures to monitor and control appear to signal still elements that are part of a more traditional PA logic.

Therefore, the institutional development of PBB systems appears gradual and limited, where its development has been challenged by diverging pressures stemming from the practice ambiguities. As both a technical method and an element of cultural orientation, linking performance information to the budget process in difficult-to-measure environments is a highly complicated and ambiguous process. The interference of political interests with the PBB managerial logic (what is relevant to do) and the need for controlling performance (what is relevant to measure) result in the elaboration of ambiguous goals and indicators. Thus, as occurred in the case of immigration policies, the difficult-to-measure environments result in ambiguous metrics, which determine difficult-to-interpret information because of the lack of a univocal and clear way of establishing cause and effect relationships and taking funding decisions accordingly. As the budget allocation problem is extremely complex (Wildawsky 1986), decision-making needs to incorporate heuristics, such as performance indicators, to solve the problem (Browley and Powell 2012). These indicators can be used for symbolic purposes and only in a few cases for directing attention to relevant areas of resource allocation choices (Vakkuri and Meklin 2003). In this context, contrary to the attempts at solving problems with technical decisions, decoupled indicators may produce less distorted effects. In public administration, the close linkage between performance information and budget allocation may not always be a sensible strategy for policy making in difficult-to-measure environments, as demonstrated by the current analysis. Thus, decoupled performance indicators may actually make sense. This creates several limitations for using PBB in different contexts of the public sector, and points to the limited development of PBB even in countries with significant and apparently successful experience with the reform. Further, unexpected events and societal problems may generate difficult-to-control situations, where delivering public services and values become incompatible with measuring performance, reinforcing the perception of ambiguities.

These findings show that PBB ambiguities in Finland and Sweden exist, as intrinsically linked to public budgeting processes, and contribute to explaining the partial implementation and use of the practice. Performance-informed budgeting (OECD 2007) seems in place since the allocation of resources is indirectly influenced by performance information in the majority of the policy areas. This indicates that use of performance information in the budget process is boundedly rational and an ambiguous process where the textbook model of implementation does not always work properly (De Bruijn 2007; Vakkuri 2010). Divergence from a textbook model emerges in Finland and Sweden, where the perceived criticalities pave the way for different reactions of the central government. Signals of new shifts in the logics of modernization of the public administration have emerged, consistent with recent movements towards NPG. In Finland, the central government has paid increasing attention to managing coordination and horizontality, while in Sweden trust-based systems have attracted the attention of the newly elected government.

Therefore, a global discursive convergence towards the idea of PBB has inspired an initial convergence towards similar decisions about PBB, calling for introducing changes in the budgeting process in order to reflect the managerial logic of NPM. These findings point out a decisional convergence demonstrated by the similar evolutionary changes within the context of NPM. However, the practice ambiguities stemming from PBB implementation and use have challenged the convergence towards a unique model of PBB. Practice difficulties have inspired the emergence of divergences influencing the logics of modernization of the public administration. In both the countries, PBB has represented only a small component within the broader context of performance management, pointing out how the final outcomes of PBB appear not revolutionary, especially when the analysis goes beyond what it is on paper (Van Gestel and Teelken 2006).

Conclusions

Many PBB-related reforms date back more than 20 years, which indicates rich experience in understanding possibilities, but also extremely important limitations in implementing PBB. In this context, due to their characteristics as early adopters of NPM-oriented reforms, Finland and Sweden have been at the forefront in integrating performance information in the public budget process, and are considered as modernizers (Pollitt and Bouckaert 2011). The significant experience of the practice in the two countries has been a crucial factor allowing us to analyse and interpret the institutional development of PBB. In addition, the countries’ specificities have influenced the findings, emphasizing some aspects of the reforms such as the strong tendency to decentralize due to the strong democratic basis of the Scandinavian governments (Kisner and Vigoda-Gadot 2017). However, although the focus on two countries can be considered a limit to the generalizability of the findings, the research aims to build some room for thoughts that can be useful for explaining the development of budgeting practices in different contexts.

The analysis of PBB in Finland and Sweden has underlined the existence of both converging and diverging pressures, contributing to explain the PBB development. While the global trend of NPM has been reflected in the initial predominant logic of modernization, guiding the introduction of a performance-based basket of reforms, many problems in linking performance information to the budgetary process still seem to prevail. Our findings show that PBB ambiguities are recognized but not solved over time in both Finland and Sweden because of the tendency to emphasize the technocratic expectations of performance management, assuming a tight coupling between government budget objectives and uses of performance information in the process (Browley and Powell 2012), and thus reinforcing rather than addressing the ambiguities of the system. The practice ambiguities challenge the development of PBB and new reforms appear to receive growing attention, reflecting the existence of multiple logics of modernization (Wiesel and Modell 2014). As part of the broader management-by-results movement, PBB systems were only a means to an end in the reform package, not the end itself.

Our paper argues that much of the current doctrine of PBB system development is based on a narrow understanding of how budgetary decision-making actually takes place in government. Contributing to the study of PBB development (Clark, Menifield, and Stewart 2017; Mussari et al. 2016), this research emphasized the pressures leading to such development. The existence of converging and diverging pressures is suitable for shedding light on the features finally assumed by a practice and its coexistence with different reforms. Indeed, the development doctrine should benefit more from institutional and ambiguity theories, since the question is not only about the diffusion of PBB, but also about something more fundamental in the ways in which it works in context. Our findings contributed to this debate, pointing out the need to consider and integrate the ambiguity of the budgetary systems and the context where it takes place. If PBB is and should be only a small component of the reform package, it is of utmost importance that its study explores the integration with the other reforms and practices. We argue that there is thus a significant need for more detailed research and theorizations of public budgeting systems, using institutional development and ambiguity theories, in different countries and in different contexts of public administration.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Giuseppe Grossi

Giuseppe Grossi is Professor in Accounting at Nord University (Norway), Professor in Public Management and Accounting at Kristianstad University (Sweden) and Visiting Professor at Kozminski University (Poland). Grossi’s research focuses on hybrid forms of governance, smart cities, and public budgeting.

Sara Giovanna Mauro

Sara Giovanna Mauro is Research Fellow at the Institute of Management, Scuola Superiore Sant’Anna of Pisa, where she received her PhD in Management. Her primary research interests are in public sector accounting and management, with a focus on performance budgeting, performance management, and participatory management.

Jarmo Vakkuri

Jarmo Vakkuri is Professor at the Faculty of Management, University of Tampere, and the director of the research group on Public Financial Management. His research concentrates on public financial management, performance measurement in public administration, theories of decision-making, hybrid forms of governance, and urban policy-making.

Appendix 1

List of interviews

Table

Organization	Interviewee	Role	Length (minutes)
FINLAND
Ministry of Finance	A	Financial Counsellor – Personnel and Administrative Policy, Administration and management policy unit	60
		Negotiating official	50
	B	Personnel and Administrative Policy, Administration and management policy unit
Ministry of Economic Affairs and Employment	C	Development Manager – Employment and Entrepreneurship Department	67
Ministry of Interior	D	Director General – Migration Department	75
SWEDEN
Ministry of Finance	E	Desk officer – Department of Finance	64
The Swedish National Financial Management Authority	F	Senior Analyst	69
Ministry of Justice	G	Deputy Director – Division for Management Support and Administration	45
Statistics Sweden	H	Director General	55

Appendix 2

Interview Outline

• Interviewee background (e.g. positions held, current position and main responsibilities)

• Main reasons to introduce PBB

• Existence and content of official requirements/guidelines/laws introducing and disciplining PBB

• Definition of PBB

• Initial design of the reform and current use

• Main differences compared to pre-reform scenario

• Main problems to address during its development

• Solutions adopted (which kind and by who) or lack of

• Changes in the reform over time

• Main problems currently connected to PBB

• Ongoing changes and prospect for the future