Is the ability of politicians to act as representatives between elections hampered in systems inspired by NPM?

Abstract

In this article, I explore and analyse how the application of systems inspired by New (Normal) Public Management affects the capacity of politicians to act on behalf of the people. The aim is to acquire new knowledge about how output-based financial and performance measurement systems affect the form and content of deliberative processes between elections. The analysis is based on an extensive empirical process, both backwards and forwards in time, tracing analysis of a critical event that occurred within education policy in a Swedish municipality. One main hypothesis generated is that NPM-inspired systems have a negative effect on deliberations between elections in ways that generate lock-in effects that hamper the ability of politicians to act as representatives between elections.

Keywords:

• new (Normal) public management
• performance management
• representative democracy
• deliberative processes
• municipal managers
• local politicians

Introduction

Performance management systems have the potential to provide politicians, the people and public officials with knowledge about costs and quality of public policies and thereby serve as a basis for accountability and organisational and operational development (Heinrich 2002; Christensen and Lægreid 2015). Performance management systems are consequently often perceived to be essential and necessary components in a well-functioning representative democracy. Performance management systems are implemented with the aim of altering values and roles within organisations in order to boost promotion of overall goals (Poulsen 2009; Nutley et al. 2012). Indeed, a huge amount of empirical research demonstrates that behaviour and values are affected, though often in unintended and undesired ways. There is a prodigious amount of research that includes, for example, studies focused on the use and misuse of performance information (Heinrich 1999; Pollitt 2006; Moynihan and Pandey 2010) and the effects of performance measurement on values and behaviour among professionals within organisations (e.g. Brignall and Modell 2000; Pollitt 2006; Verbeeten 2008; Evetts 2009; Green 2013; Dahler-Larsen 2013; Kallio and Kallio 2014). However, the effects of performance management systems on the quality of representative democracy have not been widely scrutinized. I therefore pose the question in this article of what effect output-based financial and performance measurement systems have on the roles and behaviour of politicians and public officials. My intention is to explore and discuss whether representatives’ ability to represent between elections might be affected by the design and application of management systems. The point of departure for the study is theories of representative democracy with the focus on deliberative processes between elections and theories of performance management. An extensive process, both backwards and forwards in time, tracing analysis of a critical event that occurred within education policy in a Swedish municipality, will serve as an empirical base to explore and identify potential causal mechanisms affecting the quality of representative democracy.

The paper is organised as follows. First, the theoretical concepts of representation, deliberation and public management systems are outlined and discussed. Second, the methodology, contextual setting and empirical case are presented. Third, the case is described and analysed, and finally the conclusions are presented and discussed.

Representative democracy and public management

Theories of political representation and empirical analysis of representation have engaged numerous scholars within the research community since the emergence of representative democracy in the 20th century. Central to researchers’ interests are questions of how representation can be understood and defined, along with questions of how political representation can best be studied and measured. Apart from the obvious philosophical issues of whether it is at all possible to represent something or someone, the question of different forms and content of representation has been constantly on the agenda (Eulau et al. 1959; Pitkin 1972; Eulau and Karps 1977; Mansbridge 2003; Karlsson 2013). The relational link between representatives and the demos and what it is or can comprise and what (whom) (can be) represented has attracted a great deal of theoretical attention and generated a considerable number of typologies and normative positions (Pitkin 1972; Mansbridge 2003; Andeweg and Thomassen 2005; Held 2006; etc.). Regardless of the researcher’s normative or methodological position, one common conclusion is that representation can take a large range of forms. Over the course of time, the early distinction between delegates (i.e. representatives who act in accordance with the preferences of the demos) and trustees (i.e. representatives who act in accordance with an individual assessment), has been extended.

Deliberative processes between elections

Mansbridge has put forward four forms of representation in relation to the concept of accountability (Mansbridge 2003). Three of them – promissory, anticipatory and gyroscopic representation – highlight the need for deliberative processes within the demos between elections. The fourth – surrogate representation – is of less relevance in this context as it relates to situations where representatives represent actors outside the demos, i.e. people and groups with whom the representatives have no electoral relationship. While promissory representation presumes that the demos appoints representatives based on campaign promises made, and that the demos punishes representatives who do not fulfil their election promises, anticipatory representation presumes that the demos acts retrospectively, i.e. that it punishes representatives that fail to deliver, regardless of their campaign promises. The third form, gyroscopic representation, presumes instead that the demos appoints representatives that act in accordance with certain values and norms rather than implementing specific programmes. Mansbridge states that promissory representation, but above all anticipatory and gyroscopic representation, requires effective deliberative processes between the demos and its representatives between elections in order to make representation even possible. In line with this reasoning, the quality of representation is assumed to be linked to and dependent on actual deliberative processes taking place between elections. As articulated by Mansbridge, deliberative processes should meet the criteria of being ‘free, equal, and rational or reasonable (2003, 525)’. She also points out that each criterion is somewhat unclear and needs further elucidation.

No matter which representation model is applied or presumed, whether a promissory, anticipatory or gyroscopic model, in practice, representatives will have to manage all policy issues arising during their mandate period, including new questions and problems that were not present when they were first elected. Indeed, they will largely need to exercise their mandate independently, i.e. to use their powers to design and adjust policy in relation to new or changed contextual conditions and demands affecting the demos and its members during the mandate period. This means that representatives need information and knowledge of what is going on within their demos, otherwise they are not able to fulfil their roles as representatives and decision-makers. During their term, representatives are normally involved in a number of deliberative processes including, for example, deliberative processes with citizens, NGOs, party organisations, governmental authorities and business communities, which provide representatives with information that forms a potential knowledge base from which policies can be identified and developed.

Public officials as agents of deliberation

In modern democracies, representatives are not involved to any major degree in the everyday work of, for example, administration, management and delivery of services, i.e. tasks that the representatives have decided should be performed within the demos. There is typically a public administration (PA) at the representatives’ disposal composed of employed officials, professionals and operatives. A PA and its officials influences the setting of agendas, implementation, output and outcomes. It is not possible to find research findings that indicate otherwise (see for example Peters and Pierre 2007). It is therefore somewhat incomprehensible that researchers interested in representation often seem to overlook the question of how public administration systems and public officials influence the potential forms and content of representation. The PA, its management systems and its officials play a central and decisive role for and within deliberative processes. They often act as coordinators and disseminators of information and knowledge and might therefore be defined as deliberative agents. In addition, the contextual setting in which deliberations are performed influences roles and functions. It is therefore reasonable to assume that different political-administrative and management systems will affect interaction patterns between the three core groups within representative democracies: the demos, the representatives and the public administration and its officials, i.e. the interaction patterns will most probably affect what kind of information and knowledge constitutes the content of deliberations. So, what kinds of political-administrative systems currently surround deliberative processes within representative democracies?

New (normal) public management

Western democracies started to implement reforms inspired by New (which today can rather be construed as ‘Normal’) Public Management in the late 1970s and since then traditional public administrative systems have been involved in an ongoing restructuring process. According to the NPM philosophy, a reformation of the political administrative system was necessary based on two main arguments. First, increased efficiency of the public sector, which would be accomplished by incorporation of private management methods and techniques in public administration and by the use of private operators as producers of public services. Second, to strengthen democracy, which would be accomplished by giving citizens the ability to choose service providers based on their individual preferences and demands (e.g. Osborne and Gaebler 1992; Bouckaert and Halligan 2008). Implementation of reforms such as these alter the form and possible content of democracy, as well as the functions and roles of citizens, politicians, public administrators, professionals and service recipients (e.g. Lapsley 1999; Pollitt 2006). A more reasonable hypothesis is consequently that the notion of what democracy is and how it works will be profoundly changed (Christensen and Lægreid 2002; Peters 2010). In practice, the implementation of NPM-inspired reforms has been closely related to and dependent on the development and increased application of financial control and performance measurement systems (e.g. Power 1997, 2003; Lapsley 1999; Pollitt and Bouckaert 2004; Fryer, F.,  J. Antony., and S. Ogden. 2009; Dunleavy and Hood 1994; Lewis and Triantafillou 2012). Today performance measurement constitutes a core element in public management systems all over the world, at the local, national, regional and international level (e.g. Pidd 2012; Van Dooren, Bouckaert, and Halligan 2015).

Financial control and performance measurement

Intuitively, it is almost impossible to perceive performance measurement and financial control models as anything other than valuable tools supporting and maintaining strength and stability in a democracy. Not least since these tools produce information about policy outputs that has the potential to serve as a starting point in deliberative processes. In the late 1970s, early versions of financial control measures and performance measurement were introduced in local government systems in Sweden and in other countries. At that time, the adoption of key performance figures and costs represented tremendous innovations that made it possible for both representatives and members of the demos to see, discuss and question what actually came out of the political system and at what cost (e.g. Altman 1979; Mayston 1985). Budget documents and annual reports had previously consisted of headings and a total with no description of, for example, the number of children or old people within that total who would receive or had received care. The introduction of performance measurement figures opened up the opportunity to challenge, alter and change how and for what purpose public recourses were used, and the content of public services provided. This transparency gave the demos, the representatives and the public administration increased information and knowledge about public policy.

Nevertheless, a rationalistic view of decision-making, implementation and behaviour is implicit in performance measurement and financial control models. The logic behind measuring performance and costs within public organisations is the idea that general goals can be operationalised into indicators which can be used by different actors in the political administrative system as a support in decision-making (e.g. Mayston 1985; Smith 1995; Van Dooren, Bouckaert, and Halligan 2015). However, a huge and unequivocal amount of empirical research clearly shows that these kinds of systems systematically generate behaviour that counteracts the intention behind them. This behaviour includes tunnel vision, cherry-picking, gaming, myopia, sub-optimising, ossification, misinterpretation, irresponsibility, inaction and a fixation on measuring. Conduct that in different kinds of political-administrative systems and countries has together proven to generate, for example, output-fixation, goal-displacement and de-professionalisation, as well as inhibit innovation and strengthen public bureaucrats at the expense of politicians (Smith 1995; Perrin 1998; de Bruijn 2002; Van Thiel and Leeuw 2002; Pidd 2005; Radin 2006; Ordóñez et al. 2009; Fryer et al. 2009; Mannion and Braithwaite 2012, Johansson and Montin 2014; Johansson 2015). And although we have extensive knowledge about the multitude of undesired effects that performance measurement and financial control systems generate (e.g. de Bruijn 2002; Radin 2006; Mannion and Braithwaite 2012), our knowledge of the systematic and aggregated effects on the quality of deliberative processes within representative democracy is still underdeveloped.

The expected roles of politicians, managers and professionals in financial and performance measurement systems

Three main groups are assumed to adopt tangible and mutually exclusive roles in relation to NPM-inspired financial and performance measurement systems within political-administrative systems. The representatives, i.e. the politicians, are supposed to develop and set goals, the public officials are expected to arrange the management system and operationalise goals into indicators, while professionals are expected to execute tasks and report results (e.g. Aucoin 1990; Van Thiel and Leeuw 2002). In theory, information generated by performance management systems can serve as a basis for and in deliberative processes between elections. The systems provide voters, politicians and users of public services, as well as managers and professionals, with information that makes it possible for them to deliberate about and assess public service quality and costs (Heinrich 2002; Poulsen 2009; Nutley et al. 2012). However, in the light of the research findings referred to above, it is reasonable to assume that the design of management systems – how performance- and financial measures are produced and presented – will most probably affect what is being deliberated and how. In order to be able to advance our theoretical understanding of how and why management systems affect deliberative processes, and by extension the quality of representative democracy, it is necessary to study such processes empirically. So, how can these kinds of processes be identified and analysed?

Case and methodology

The form and content of deliberative processes between elections are related to the contextual setting where they actually take place. Case studies are consequently a better choice than surveys if the purpose is to acquire new knowledge and generate hypotheses about causal relationships affecting the form and content of deliberation processes between elections. To quote Eulau and Karps, writing as far back as 1977:

Indeed, there are today users of survey research who have never interviewed a single person in their lives. Not surprisingly, therefore, causal models are being reified as if they described reality rather than being abstractions from reality …. Theorizing involves something more than arbitrarily inverting the causal directions on the assumption that the resultant statistical structure will somehow reflect reality. It involves giving reasons and justifying the assumptions one brings into the casual analysis. It involves “going out on a limb,” as it were, and saying something substantive about the phenomena being investigated, rather than hiding behind the artifactual “findings” of a casual analysis that may be inappropriate in the first place. (Eulau and Karps 1977, 249–250)

Methodology: backward and forward tracing

The empirical findings that will be used as a basis for exploring if, how and why NPM-inspired systems affect the form and content of deliberative processes have been collected within a case study conducted in 2015 in a medium-sized municipality situated in the south of Sweden. Around 2010–2011 a political crisis arose in the education sector in this municipality. The crisis involved numerous actors and deliberative processes and was triggered by an event that in retrospect can be defined as a formative moment. There is an epicentre in every formative moment that unleashes a chain of events, and there is a sequence of events and processes that precede the triggering event. From a methodological perspective, a formative moment is highly advantageous since it makes visible structures, processes and causal mechanism, which are often concealed under normal circumstances and difficult to detect. The triggering event will thus be used as a starting point from where processes and events that took place both immediately and long before the triggering event will be traced backwards, as will processes that followed it.

The methodology has features of process tracing, a method that in recent years has become increasingly common within qualitative research (Tansey 2007). Collier defines process tracing as ‘an analytical tool for drawing descriptive and casual inferences from diagnostic pieces of evidence – often understood as part of a temporal sequence of events or phenomena’ (Collier 2011, 824). Process tracing is often used as an analytical tool in in-depth single case studies since it allows within-case inferences about causal mechanisms. Further, the method can be used either to test and develop hypotheses/theories or to explain outcomes (Beach and Pedersen 2011). Process tracing commonly uses multiple data sources to trace events and situations backwards from a certain point in time, which will also be performed here. However, in addition forward tracing will be applied. A combination of forward and backward tracing enhances opportunities to advance the analysis by placing the results from the backward tracing in a new light through comparisons between before and after.

Conceptual framework

The normative prescripts of promissory, anticipatory, and gyroscopic representation postulate that deliberations between elections are of the utmost importance for the quality of representative democracy. In addition, these processes have to meet the criteria of being ‘free, equal, and rational or reasonable’ (Mansbridge 2003, 525). Deliberations always comprise communication about an issue or a range of issues, where the communication is based on information and knowledge available to those deliberating. A common criterion promoting good conditions for deliberations often pinpointed in the literature is that concerned parties must have opportunities to be heard, another common criterion is that concerned parties need to have access to arenas where deliberations are taking place (see for example Räftegård 1998; Bickerstaff and Walker 2005; Agger and Löfgren 2008). However, theoretical deliberations that take place within a representative democracy between elections cannot be understood as equivalent with deliberative democracy or participatory democracy. Accordingly, the main aim is to explore if and how NPM-inspired management and financial models affect the form and content of deliberative processes within a representative democracy between elections.

In order to explore how NPM-inspired systems within the municipality structured and affected deliberations before and after the crisis, the empirical data will be illustrated through four analytical categories a) events that generated reactions among concerned parties, b) issues deliberated, c) actors deliberating and d) deliberation results. Further, to improve the prerequisites for contextualising deliberative processes and thereby also the prerequisites for developing causal hypotheses, temporal sequences will guide the empirical presentation, and the analytical categories will consequently be interwoven in a coherent narrative.

Empirical data

Empirically, the case study includes data from multiple sources: interviews with all senior managers in the public education sector in the municipality, policy and performance documents as well as proceedings and protocols from political committees and boards in the municipality. In addition, performance and decision documents from national authorities are included, as well as media reports. The empirical data was collected in 2015 within a research programme focusing on changing values and roles among public officials.¹

General and specific contextual setting

In case studies where process tracing is used as an analytical tool to identify causal mechanisms, the general context within which activities take place is important. Contextual knowledge improves the prerequisites for understanding how and why processes unfold as they do. For those who are not acquainted with the local government system in Sweden, before presenting the case, it is therefore necessary to briefly outline a few general characteristics of the local government system and Swedish education policy in order to contextualise the case and make it comprehensible.

The general context of the case: the local government system in Sweden

Although Sweden is a unitary state, compared to municipalities in other local government systems the municipalities have a high degree of self-government, which is also enshrined in the constitution. Swedish municipalities are obliged to comply with national legislation and they cannot enact laws. However, most laws relevant at municipal level are framework laws that provide the municipalities with room to manoeuvre and the power to adjust and adapt local government activities in relation to local preferences. That apart, municipalities are free to institute measures in relation to any issues they want, unless prohibited by national law. Further, national policies developed by national authorities are only compulsory if enacted in law, which they seldom are. Each of Sweden’s 290 municipalities is governed by a directly elected local parliament that appoints members to political executive boards in different fields. All of Sweden’s inhabitants are members of at least one municipality and, while membership of a municipality is a necessary condition for voting rights, citizenship is not. The municipalities are responsible for the bulk of welfare provision such as childcare, education, health, social and elderly care, but are also responsible for issues such as planning, waste management and public transport. A Swedish municipality is usually the largest employer within its own territory, even in comparison with large private enterprises. Income taxes are mainly decided by and used by the municipalities. In short, in many respects Swedish municipalities resemble countries or counties in federal states. They are one of the most powerful entities in the Swedish political-administrative system.

The specific context of the case: education within the local government system in Sweden

During the last 40 years, Swedish education policy has undergone three interrelated phases (Lundahl 2015). The first phase, (1970s to 1990) focused on increased municipal autonomy, with the education system being municipalised by the end of the period. The second phase (1991 to 1998) focused on NPM-inspired reforms, and a quasi-market system was introduced. In the third phase (late 1990s to early 2000s), a centralisation process was initiated as a response to increasing differences between schools and municipalities, mainly through the application of economic instruments (Lundahl 2015). While municipal responsibility for education remains unchanged, since the mid-2000s the state has intensified its effort to promote equality between schools and municipalities through other means than economic instruments, including implementation of a new Education Act, increased governmental supervision through inspections, a new grading system, changes to teacher training, compulsory primary education and teacher certification (Johansson, Lindgren and Montin 2018). In spite of this, the municipalities in Sweden are still in charge of the education system both substantively and economically, even though they need to relate and adhere to expanding national regulations and policies to a higher degree then before.

The educational crisis

With the brief description of the general and specific contexts in mind, it is now time to present the case. The tracing will commence with a description of the event that triggered the educational crisis, and from there the analytical categories – events, issues, actors and results – will be interwoven in a coherent narrative of temporal sequences.

The triggering event

The educational crisis within the municipality comprised a single event that constituted its epicentre, i.e. a clearly distinguishable event from which both past and future events and processes can be traced. In the quotation below one of the public managers involved vividly describes the event that unleashed the crisis.

There were journalists at the meeting. They were standing at the back, but we were not aware that they were there. They had come prepared with cameras. This parent who was the most active, the most aggressive and opinionated, was not sitting in the front row. He was at the back among the journalists. Then, when my senior manager had started and had only been talking for a few minutes, we saw this big man, because he is a big man, with a lot of authority, barging his way through the meeting room, like an Indiana Jones, and he walks all the way up to the manager, who looks apuzzled fashion, pulls the lead out of his (the senior manager) computer, and connects his own computer, and on that computer there it is a picture of a burning oil platform, and he says: THIS my friends, THIS is the school in the municipality of X (the name of the municipality) (Senior Manager for of primary schools)

At the time of the protest, the parent who tangibly and physically demonstrated against school policy had a child in year nine at one of the municipalities’ compulsory schools. Together with three other parents he had previously demanded on several occasions that the principal of the school take action in order to bring stability to the school. The parents claimed that a high turnover of teachers, mainly in mathematics, had created a high degree of unrest and agitation in the class, which was potentially putting the pupils’ final grades at stake. The meeting that took place in 2010 had been arranged by the municipality with the intention of initiating a dialogue about how to deal with the current situation. Every parent that had a child at the school had been invited, and as well as the head of education and the head of primary schools, politicians from the local executive school board were also present. The incident described in the quotation above led to the meeting breaking down and the municipality’s intention to start a dialogue failing.

However, the demonstration was just one event among a network of events and activities related to the situation in the school, in which a large range of actors had been and would come to be involved. In addition to parents, local politicians and local municipal officials, the National Ministry of Education, the Swedish Schools Inspectorate, the Swedish Work Environment Authority, Dagens Nyheter (the largest national newspaper in Sweden), local newspapers, national television and the Swedish Association of Local Authorities and Regions (SALAR) were some of the actors that would come to participate. One event led to another and for a long period of time the municipality was stuck in a situation that would be described years after as, ”a media hurricane”, ”chaotic”, ”peculiar”, ”turbulent”, ”dramatic”, ” anguished”, ”frightening” by managers and senior managers who were interviewed.

Furthermore, the demonstration was not the only action the parents had taken. In conjunction with the meeting, they had also sent a letter to the Swedish Minister of Education in which they requested that the school be held in receivership by national authorities (which the law does not actually permit). At the time of the meeting, the managers and politicians were unaware that a letter had been sent. The Ministry of Education was subsequently to send the petition onward to the Swedish Schools Inspectorate, which would initiate an inspection, as would the Swedish Work Environment Authority as well. The demonstration and the petition generated a multiplicity of reactions among other actors. For example, SALAR contacted both managers and politicians in the municipality wondering what was going on there.

I know that SALAR was in touch with our senior councillor (comparable with a mayor in other local government systems) and they were wondering: what was really going on in X. (Head of Strategic Educational Management and former Senior Manager for primary schools)

And SALAR called from Stockholm (capital of Sweden) wondering what on earth we were doing, and you have to do something. (Senior Manager, education)

Over a period of about one year the managers and the politicians in the municipality managed and responded to numerous events and processes that followed the incident at the meeting. However, with no prior understanding of the local contextual setting in which the incident occurred, it is almost impossible to analyse and interpret the processes that resulted from the incident and the consequences they generated.

The local contextual setting: performance, financial and quality indicators

At the time of the educational crisis, the financial performance results in the municipality were excellent. Local taxes were the second lowest in Sweden (Statistics Sweden www.statistikdatabasen.scb.se) and the municipality did not even have to borrow money to make investments (municipal annual reports 2009 and 2010). In addition, according to national and local performance indicators, the quality of services was among the highest in Sweden, not least within the education sector (SALSA ‘open comparisons’ www.skl.se, municipal annual reports 2009 and 2010, Swedish National Agency for Education http://www.jmftal.artisan.se etc.) Accordingly, nothing in the performance management and indicator system at local level (or at national level) at the time of the incident indicated that there were any problems at all within the municipality’s school system. The finances were sound and the results more than satisfactory; in actual fact, the performance results were outstanding. Furthermore, the school attended by the children of the protesting parents scored very highly on almost every performance indicator, for example, the pupils’ final grades and their results in national tests were way above average in both local comparisons and a national perspective. At first glance, this combination of circumstances makes the parents’ worries seem somewhat incomprehensible to an outsider. Why take these drastic measures when every performance indicator was favourable?

In fact, not everything was perfect, there were problems in the school which emerged from beneath the surface, processes that were linked and which reinforced each other and which eventually culminated in the educational crisis in the municipality. So, let’s return to the period in which the incident occurred. What problems were present and when and why had they emerged?

Problems were recognised, but not taken seriously

The interviews clearly indicate that the school had had problems for many years before the incident. However, since the financial control and performance measurement systems were primarily composed of indicators related to costs and pupils’ grades, these problems were not highlighted.

You could say that we were taken by surprise because … YES, there were shortcomings, but the results hadn’t declined. I feel that the incident was like an alarm bell for us (Senior Manager, Administration).

What were those shortcomings that the performance management system and its indicators did not capture? According to the managers interviewed, in hindsight the principal problem was not the pupils’ grades, it was the work environment that was problematic, especially for principals and teachers.

I would say that it was the work environment that was the problem (Senior Manager, Education)

A lot of it concerned the work environment, teachers who felt uncomfortable etc. (Senior Manager, Administration)

As the quotations indicate, the problems associated with the work environmental were not unknown to the managers, they were aware of their existence. So, the question is: how is it possible that shortcomings that were recognised at managerial level were not addressed and dealt with, if not by the performance management system, by other means? In order to answer this question, it is necessary to briefly outline the organisational processes and factors that led to the poor working environment for principals and teachers. A couple of years before the incident, the primary-school system in the municipality had been reorganised due to external and national reforms and auditing processes.

A private school is established

One of the municipality’s organisational reforms was related to the fact that a private school had been established in the municipality as a consequence of the Swedish system of free school choice (statutory since 1992). Within the free school choice system, it is the SSI, not the municipalities that issues a permit for schools to be set up, but it is the municipalities that are obliged to pay the cost for the school (www.skolinspektionen.se; Education Act SFS, 2010:800). The establishment of the private school meant that the municipality had to liquidate and close down one of its own schools and reduce municipal costs by SEK 18 million (about 2 million euros). Research consistently shows that on the vast majority of occasions layoffs generate anxiety in an organisation and for a while the work environment is usually affected negatively (e.g. Brockner et al. 1992; Kivimäki et al. 2001). However, the redundancies did not only take place at the school where the protesting parents had their children, only 5 out of the 40 employees that were dismissed by the municipality were employed at that school. It is therefore unlikely that it was solely the work environment in that school that was negatively affected. Further, it is also unlikely that during the dismissal process no one in the entire municipal organisation raised the issue of negative effects arising. Or were concerns actually raised without anyone listening? Yes, the latter seems to be the case. Concerns were raised but they were brushed aside at management level. It is now necessary to go even further back in time in order to identify potential causative factors explaining why the warning signals that were present were not addressed at management level. One such causative factor seems to be the financial control and monitoring system in the municipality and how it was applied.

An inflexible financial control system

At the time of the incident, there was rigorous financial control of the political boards and their activities. Since the early 1990s, keeping to the budget had been the main goal and that goal had been, and was still, superior to all other targets. How this goal was mediated throughout the municipal organisation is illustrated in the quotation below.

The chief executive of our municipality, the one we had then, used to say: You have four tasks if you work in x (name of the municipality): the first is, “Keep to the budget”, the second is, ”keep to the budget”, and the third is, ”keep to the budget” and then we have the fourth …”keep to the budget”, well that was the only thing they were ever interested in (Head of Strategic Educational Management and former Senior Manager for primary schools)

The admonishment to keep to the budget was combined with continuous and rigorous monitoring of the financial results. Every manager that failed to keep to the budget at all levels within the municipality had to explain him/herself in front of the political executive board and specify what measures he/she would take to overcome the budget deficit. If you have watched the series, The Wire, the following scene might be familiar.

There was a lot to, as it were, focus on, but there were also concerns. If you did not have a balanced budget, you had to come to the political board meeting and defend yourself in front of the whole executive board (Oh, who?). Well, the principal or the person responsible for any activity, any operational manager, any person who had any kind of financial responsibility. That person had to report in front of the board and was held accountable in the full glare, so to speak, and that person had to explain what measures she/he intended to take and explain why he/she had failed. (Senior Manager for primary schools).

In a context where the overarching and highly distinct goal to keep to the budget is combined with a procedure where individuals that fail to meet the target, regardless of why, are subjected to what best can be described as a public pillorying, there is not much space for voices to be raised, at the same time as those voices that are raised are often silenced or used against those who have spoken out (Hirschman 1970; McClean, Burris., and Detert 2013)

Exit replaces voice

The interviews indicate that, well in advance of the incident, many municipal employees had tried to initiate deliberations through using their voice as a strategy to point out that budgetary targets put the operational goals at stake. However, these voices had no effect, and as a consequence several operational managers, i.e. school principals, instead used an exiting strategy. Over a period of one year just before the incident, half of all the principals in the municipality left their employment. The school where the demonstrating parents had their children was one of the schools that had lost their principal, not just once but on three occasions.

During the course of six months, firstly one failed principal was replaced, and the principal who got the job after that subsequently dropped out. Somehow there was a general unrest at that school. (Senior Manager, quality systems)

So, when the parents approached the principal with their worries, the management function was relatively unstable. The latest principal had just resigned. According to one of those interviewed, that principal was no longer able to accept or manage the cutbacks continually placed on her by the managerial level.

There was a shortfall of three to four hundred thousand. You have to produce an action plan and describe the measures you are going to take…. so the principal resigned. She felt that she had done so much … she thought that she had developed an awful lot (Head of Strategic Educational Management and former Senior Manager for primary schools)

The quotation suggests that a lot of development work had taken place at the school before the principal resigned. The managers interviewed refer to these development activities as absolutely essential and simultaneously highly successful. Did these development activities have any bearing on what was to subsequently transpire, and why had they been initiated in the first place?

Improvement and development activities are initiated and resisted

According to those interviewed, the reason behind the extensive development activities can be traced back to the notion that at the time when the development measures were initiated, the school was not meeting the national regulatory documents, for example the national curriculum.

In 2007 the school was still following the national curriculum from 1980; it had not even started to work on the curriculum from 1994. The school did NOT use the national, target-based curriculum. … yes, the pupils were achieving good grades but they did not adhere to the national curriculum, they either used rating criteria from the 1980s or invented their own rating criteria, they had their own knowledge targets. (Head of Strategic Educational Management and former Senior Manager for primary schools)

In order to be able to transform the school into what the interviews refer to as a ‘modern school’, the political-administrative level initiated a development project. Two new principals were appointed and were given the task of identifying and implementing all measures they found necessary in order to bring about the transformation the political board and the management level requested. In addition, the principals were tasked with dismissing teachers.

Two new principals were appointed and they were instructed, actually directly from the political level that: “Something has to happen, you have a free hand to do whatever has to be done” and they did. (Senior Manager, education)

The improvement and development work included the implementation of the current curriculum (i.e. the regulations from 1994), the winding up of the old lecture system and the abolition of the old so called USK agreement (concerning teaching duties) that still was in use at the school even though it had been abolished in the early-1990s. In concrete terms, the development work included: target and development plans, the introduction of teamwork at teacher level and the formation of subject links. The interviewees painted a picture of a tightly knit group of teachers, with strong informal leaders who felt threatened by the measures taken.

One of the informal leaders also prepared the schedule, which meant that it was not an educational schedule at all. Instead it was about who had to go home for dinner, who had to go out with the dog and who wanted to be free on Friday afternoons, it was not about education at all, it was about personal needs. I was informed by phone when these changes were made, and that’s why I’m telling you this, because this is partly what led to the poor work environment… I was phoned up and asked: What is going on at the school, it was the teachers’ relatives who phoned, and the principals received many, many more calls like that. For example, one teacher’s wife called and she said: “does he really need to be at the school all day?” I replied: “yes, he works full-time so that’s not particularly strange”. “But when is he supposed to come home and eat lunch”, she asked. (Head of Strategic Educational Management and former Senior Manager for primary schools)

Vested interests and/or professional resistance

If we take the picture painted by the managers as a starting point, the resistance of the teachers can be interpreted as an act of defiance in favour of their autonomy and discretion, not based on a professional interest but on vested interests. However, one of Sweden’s most praised and well-known journalists who was subsequently to analyse the problems at the school in Dagens Nyheter (Sweden’s leading newspaper), depicted the teachers’ resistance quiet differently; as unselfish behaviour in defence of their pupils and as a struggle to maintain the professional competence necessary for teachers. The explanation put forward by this journalist is actually diametrically opposite to the one given by the managers.

The principals that were in charge of the transformation process at school level were basically preschool teachers i.e. they had a university degree and were trained educationalists for younger children. The article in Dagens Nyheter makes it clear that such persons, (their university degree is never mentioned), lack the necessary competence needed in order to manage and direct improvement and development processes in a high school.

The new principals had views on mathematics. They certainly aren’t teachers in mathematics. (To tell the truth none of them were certificated to teach in any subject at all at high school level). However, they claim that the textbooks have to be changed. The mathematic teachers do not agree. However, one day when they arrive at work, the three women have brought a waste container into their room. “And there they stood throwing away 400 mathematics textbooks, our papers, our training materials, old tests. It was totally unreal. Like lunacy”. (DN 2011)

In the quotation above, the journalist portrays the principals – or the three women as he chooses to denominate them – as lunatics through the use of quotations from interviews, and as incompetent through reference to the fact that they are not certified as high school teachers. What he neglects to mention is that the person who was senior manager for primary schools and the manager in charge of the development activities at this time was a certified high school teacher in, yes precisely: mathematics (the latter is a rhetorical formulation commonly used by this particular journalist). Although gender relations are not the focus here, it is interesting to note that the hypothesis put forward by the journalist is substantiated through the (un)conscious use of perceptions about gender and status hierarchies that are essential for the maintenance of power relations (e.g. Kanter Moss 1977).

However, the descriptions of the conditions that brought forth the educational crisis are essentially the same in the stories recounted by the managers and the journalist even though they emphasise and interpret the same explanatory factors in different modes. What both narratives have in common is that the teachers perceive their working conditions as unsustainable and that the parents are concerned about their children’s results. Furthermore, both the managers and the journalist describe an alliance between teachers and parents in which a joint criticism is formulated related to the changes that have taken place at the school. But the similarities end here. In the journalist’s version, the parents and the teachers are portrayed as heroes who use all means available to try to defend the professional competence of teachers and counteract ill-suited and counter-productive organisational changes that for some inscrutable reason are implemented. In the manager’s version, a picture emerges instead of an unholy alliance between teachers and parents, where the former have taken the latter as hostages in order to defend their illegitimate discretion and their vested interests.

The national inspectorate makes demands

Why the political and managerial level initiated the development activities that came to negatively affect the working conditions at the school is not clearly elaborated, neither in the interviews nor in the media reports. In fact the development work was initiated by an external actor, SSI, with the aim of coming to terms with the school’s inability to meet national curricula and working time agreements. In a decision from SSI 2009, the authority required the school to take immediate action to comply with national regulations (SSI, Ref. no. 43–2008, 440). Among other things, SSI requested that the school take measures to guarantee that ‘the education is based on the national curriculum’. When related to regulations, requests by national inspection authorities are compulsory. The national curricula are a set of regulations, implying that the municipality had no option, it had to comply with the requests issued by SSI. It is therefore reasonable to assume that SSI’s decision was, at least to some extent, significant for the change and development work initiated at the political and managerial level in the municipality. This interpretation is supported by an article in the Manager and Leadership journal published by The National Union of Teachers in Sweden in which the previous principal (i.e. one of the principals that had resigned before the incident) at the school commented on SSI’s decision with the following words: ‘Yes, I simply felt Yes! They (i.e. SSI) saw the same things I saw’ (Chef and Ledarskap 2011). Accordingly, it is very likely that the development work was initiated as a response to the school’s and its teachers’ deficient capacity to meet national regulations. However, even so the pupils’ grades and results were well above the national average, which from a control perspective is quite interesting. What impact do regulatory-, curricula- and policy documents actually have for achievement of targets compared with professional knowledge and competence and the way professionals exercise it? Whatever the answer to that question, the development process was related to the ambition of the school to meet national regulations rather than emanating from malicious principals and homemade targets. In other words, it is not possible to draw the conclusion that the development activities were unjustifiable based on the chain of control manifest in Swedish national school policy. It is also easy to see that the teachers’ working conditions were affected and that they perceived that their autonomy and discretion were at stake. The resistance among teachers and parents was clear and pronounced, which is evident from the empirical material. It is how these conditions were understood, interpreted and deliberated upon in relation to indicators generated by the financial control and performance measurement system, by actors within the local representative democracy, that constitutes a potential and partial explanation for the chain of events and processes that led to the educational crisis.

The aftermath of the crisis

In retrospect, the managers’ perception was that there was a lack of real communication between actors within the municipality’s management and control chain, both vertically within the political-administrative system and horizontally in relation to parents and pupils, both before and during the crisis.

At the time of the incident they (the politicians) said: “We hadn’t understood that the situation was as it was, of course we have heard things, we had received signals but we didn’t acknowledge them”. And yes, we (the managerial team) had largely reported results and statistics, and the results were positive, the pupils’ results were good. Right or wrong, but the politicians perception was that there had not been clear communication concerning what the situation was really like at the school. (Senior Manager, education)

It was not a question of results. It was about parents who were very clear and who intervened. It was about a lack of communication between politicians and management, management and the school. Well, there were a lot of ingredients, and hindsight and having all the facts to hand makes it possible to see them. (Senior Manager, primary schools)

According to the managers, a lack of communication within the municipality was a precondition that triggered the educational crisis. However, rather than a lack of communication per se, it seems to be the design of communication channels and how these channels structured communication patterns and thereby the structure and content within deliberations that prevented and hampered communication. A comparison of communication patterns before and after the crisis supports this assumption.

First, after the crisis the politicians advanced their positions and involvement in setting agendas and prioritising policies. In the wake of the crisis they no longer reacted to proposals put forward by the managerial team as they had done before the crisis.

They took a decision in the municipal executive (The political board for the municipality as a whole). They wanted to see changes in the school. They said, “We want to set the agenda, i.e. we want to provide a framework for school improvement”, and it is fairly unusual that politicians step in so clearly and have ideas about school development” (Senior Manager, primary schools)

Second, the politicians and the managers started to work more closely together. The politicians started to have meetings more frequently both with the managerial team and the schools. They started to actively collect and deliberate knowledge about changed conditions and new questions during their mandate period. The politicians no longer waited for knowledge to arrive, they actively searched for it, deliberated and assessed it.

Third, the performance measurement system was changed. The focus shifted from quantifiable results and indicators to more complex assessments of performance results. In-depth analysis and deliberations of results were given precedence over reporting simplistic results. The managerial level started to use their expert knowledge to support the politicians with a broad and multifaceted knowledge base instead of making policy on their own. The managerial team started to execute the tasks the politicians gave them in a more flexible and interactive mode.

Something positive emerged out of this, a more open climate, we can now talk about difficult things, we focus to a greater extent on analysis and why we have the results we have (Senior Manager, administration)

Fourth, the financial control and monitoring system was profoundly changed. The focus shifted from rigid financial accountability to a focus on improvement and development. Punishment was no longer the answer to poor results in operational activities, instead the main strategy was to deliberate with and support those who had difficulties meeting targets so that in the future they would be able to meet the targets. Achieving a high level of objectives within the demos as a whole became a priority goal, which if successful would potentiality strengthen the demos’ trust in the political-administrative system.

And now we never talk about money. We talk about the activities, the policy. There REALLY has been a paradigm shift in this municipality: before and after the educational crisis. And now we look at the results as they are. We dare to look at them, and we do not assess the schools in relation to other schools, instead we discuss what we can do to help a school with poor results, how we can support them. We do not condemn them as we used to before (Head of Strategic Educational Management and former Senior Manager for primary schools)

Events, issues, results and actors

Through tracing the educational crisis backward and forward in time, the conclusion can be drawn that to some extent the educational crisis was externally generated by events over which the municipality had no authority; the establishment of a private school within the free school choice system and statutory requests from the Swedish School inspectorate to take action. As a consequence, two issues were put on the political-administrative agenda in the municipality; the closure of a municipal school and the necessity of compliance with national regulation in the field of education. The political-administrative management of the issues resulted in cut-backs and development activities. These measures generated concerns among parties involved such as principals, teachers and parents, actors who in different ways tried to highlight their views and deliberate on measures taken. However, the concerns about work environment and quality issues were brushed aside by the political-administrative management. In other words, there were not many opportunities for concerned parties such as principals, teachers and parents’ to be heard and gain access to deliberative arenas. How come? What hypotheses of how and why NPM-inspired management systems affect deliberations between elections can be generated from tracing events, issues, actors and results backward and forward in time?

How and why NPM-inspired systems affect deliberations between elections

A reasonable hypothesis that can be deduced from the backward and forward tracing is that, in combination with the financial control and monitoring system, performance management system effectively structured communication patterns in the municipality before and in the first phases of the crisis, i.e. who had the right to talk to whom, about what and when. Wherever deliberations were taking place, communication was clustered around performance measures and financial results.

In concrete terms, performance indicators and financial results served to set the agenda and issues that did not fit in were neglected or brushed aside. Views and arguments put forward by concerned parties such as parents, principals and teachers emanating from sources of knowledge other than indicators, and results extracted from the financial and performance management systems, were not considered as relevant by those in charge of these systems, i.e. the management. Accordingly, views and facts that were not coherent with results extracted from the management systems were not given access to the deliberation arena. In addition, the management system structured communication patterns between concerned parties so that the possibility for actors with divergent opinions to raise their voices became limited and curtailed through the use of public pillorying, which had the effect that, instead of raising issues, principals utilised the exit option. In turn, the principal’s use of the exit option hampered the possibilities of parents and teachers to express their views to someone in charge of or familiar with the situation at the school.

The performance management and financial control and monitoring systems had performative effects on who had the right to talk to whom about what and when. Those in charge of the systems, i.e. the management, were given the prerogative and power to decide how to interpret and assess reality, while those who had divergent views and interpretations were silenced. The politicians’ interpretation of the silence and the ‘excellent’ results retrieved from the systems can be described as ‘a good state of health doesn’t attract attention’.

A distinctive feature of NPM-inspired internal management techniques, like those applied within the municipality studied, is to steer and control the delivery system’s outputs (even though the intention is often focused on outcomes) rather than inputs or throughputs (see for example Hood 1995, Pierre and Peters 2000). Numerous studies have provided evidence that the application of these kinds of techniques has generated similar undesired effects and behaviours in different political-administrative contexts and countries (Smith 1995; Perrin 1998; de Bruijn 2002; Van Thiel and Leeuw 2002; Pidd 2005; Radin 2006; Ordóñez et al. 2009; Fryer et al. 2009; Mannion and Braithwaite 2012, xxxx). Accordingly, it is reasonable to assume that the findings identified and hypotheses generated through backward and forward tracing of a crisis that occurred in a Swedish municipality are also relevant and valid in other contexts and political administrative systems.

The most important and generic lesson emanating from the analysis of the educational crisis about deliberations between elections in representative democracies is that the design of management and control and monitoring systems has the potential to have a negative effect on the conditions for those deliberations between elections, and by extension the quality of representative democracy. Such systems do not only affect which actors are allowed to participate in deliberations but also when and how they can participate. Further, since the information and knowledge retrieved from these kinds of systems limits what it is possible to observe, it is difficult for divergent views and experiences not in line with the observable to reach deliberative arenas. At the same time, it is difficult for throughputs and outcomes to enter deliberative processes at all.

To conclude, a clear division of labour between the managerial and political level and in line with the philosophy of NPM limits views and creates tunnel vision and behaviour that locks politicians and public managers into behavioural and communication patterns that circumscribe their discretion, thereby preventing them from acting when action might be needed. Lock-in effects arise and are cemented in place. When representatives on behalf of the demos decide to apply and use financial control and performance measurement systems they limit and weaken their own, and after the initial implementation stage that of the managers as well, capacity to set agendas and formulate policy through deliberations, and consequently impair their own, as well as the managers’, ability to assimilate and absorb knowledge about what is going on in the demos and within public activities during their mandate period.

Bring the politicians back in!

In order to be able to fulfil their role as representatives’ and decision-makers between elections, politicians need knowledge about how concerned parties perceive, interpret and assess new, ongoing and evolving policy issues as well as knowledge about suggestions and measures to address them. Representative democracy is weakened when performance management and financial control systems hamper politicians’ abilities to increase knowledge through deliberative processes.

Representative democracies need dynamic deliberation processes between elections. Representatives need to be active in deliberations, they cannot stay at arm’s length from the managerial level and the managerial level cannot be left on its own in managing management systems. A clear division of labour between the political and the managerial level as prescribed in the NPM philosophy further locks in effects that promote tunnel vision and zombie behaviour among politicians, managers and professions. With a performance management system that keeps politicians and managers apart, the risk increases exponentially that politicians will not be able to use their independent mandate wisely and at the same time the managers will not be able to support the politicians with meaningful knowledge. It is likely that the quality of democratic decisions will deteriorate and consequently trust in representative democracy as a decision system among members of the demos will probably be weakened. Representatives in a representative democracy need to be involved in management and deliberative processes between elections – bring the politicians back in!

Acknowledgements

I would like to take this opportunity to thank the two anonymous referees for their reasonable, highly relevant and constructive comments on my manuscript. Thank you both!

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the Swedish Foundation for Humanities and Social Sciences (grant number p12-0317:1)

Notes

1. The fact that the municipality had gone through a political crisis in educational policy was not known in advance and was not a selection criterion. Instead, the municipality is one of five municipalities that have been studied in the Swedish national local government research programmes since 1980.