Keywords

Introduction

Educational governance in Sweden has changed in radical but sequentially consistent ways since WW2. I will here do a periodization of this process lasting for more than 50 years. My main interest is in the performative turn taken at the turn of the millennium, where the governing should be based on the results of schooling – by means of student performances on tests and other kinds of indicators of how well schools are performing. What are the rationales for and expectations on this kind of governing? What are the premises for its functioning and what are the results of governing by results? I will try to answer these questions by a set of inquiries, starting with an identification of predominant governing models and their different characteristic features. Then I will do an analysis of the premises of governing by results in a communicative education system (cf. Luhmann 1995; Luhmann and Schorr 2000) and identify important results of schooling during the last two decades. Do results matter in the governing of education by results?

A Periodization of Governing Models

The history of welfare state education is full of references to governing, due to the simple fact that education governing is a recurrent irritation for politicians, administrators, as well as education professionals. Education governing is a way to talk about education – its scope and limits – and to put forward hopes in how a political system might penetrate the education system and to implement desirable changes. In the discourses on education, it is possible to identify different governing models. In Table 4.1 I present that as ideal types. The actual dates in this periodization are broadly defined, due to the complexities in the establishing and decisions concerning changes in governing.

Table 4.1 A periodization in different governing modes in the trajectory of Swedish welfare state primary and secondary education 1950–2016
Full size table

A short note about context at work in relation to the periodization: It is constructed in order to make different emphases in governing model visible, but it is as such also indicating changes in the context of – and at work in – education. Thus, the centralized welfare reformation period is part of an expansive phase in Swedish welfare education carried out during a “golden financial period under the reign of socialist governments.” The following period is in a way characterized by a stabilization of the reformation period – decentralizing or deconcentration in governing. Here, governing by goals in combination with trust in policy-makers and professionals is a characteristic context. The context of the movements into a deregulating restructuring is a combination of allusions to an ongoing financial crisis and a growing international competition and references to globalization in policy-making including transnational organizations and policy traveling. This is further accentuated during the latest period where we note the vital importance of transnational networks and actors giving the politics of education a technical-administrative face – which does not mean that education policy-making has stepped back. Instead we note the introduction of new regulating technologies and fields of governance in education.

Centralized Reformation and Governing by Detailed State Regulation

The first three decades after WW2 were considered to be golden years for the Scandinavian welfare state model in governing social progress. Education was here given special attention – as “a spearhead towards the future” (stated by the former social democratic prime minister Olof Palme in a much quoted speech from 1962). The Swedish model in education was based on centralized decision-making in consensus and standardized solutions to social problem. It was formed by the idea that the state was able to conduct (or conduct the conduct of) organized activities in desired directions. Directives and procedures were produced by the center, where the then National Board of Education played a central role, e.g., in the voluminous production of syllabi by civil servants (often former teachers) or working groups. Notions of consensus and careful preparation of political decisions were conceived of as characteristic for this period (cf. Lindblad and Wallin 1993).

Decentralizing Realization and Governing by Goals

The governing of primary and secondary education turned out to be a recurrent irritation during the 1970s and resulted in a number of state commissions such as “The School, the State, and the Local Communities.” The period of centralized governing from behind came to an end around 1980, when a new national curriculum for the comprehensive school was implemented, asking for local work plans on how to realize the national goals under local preconditions and involving actors at local levels in the making of education in order to increase “school democracy.” (Governing by goals was further developed in the next national curriculum, but then in another context.) In this decentralization the existence of common national goals was regarded as a way to preserve the direction of the primary and secondary education in Sweden (Lindblad and Wallin 1993).

Deregulating Restructuration and Governing by Markets and Evaluation

As presented by, e.g., the Swedish Commission of Power, the clients of welfare state institutions were conceived of as having too little of influence on their own situation in these institutions (according to the state commission Democracy and Power in Sweden, c.f. Petersson, 1991). This kind of irritation was related to different discourses on governance of schools – to what extent and in what ways should consumers or citizens or clients and users of education – and different ways of dealing with this financial deficit. In conjunction with an increasingly problematic financial situation, economical issues war conceived of as being essential in education considered as a cost rather than a resource. Different ways of calculating the efficiency of education were the outcomes were in focus.

Given this, a radical break in the governing of welfare state schooling turned up in the early 1990s. Presumable inspired by the Thatcherite restructuring of education in England, a new model for governing was introduced by means of deregulation, privatization, and marketization in combination with a voucher system. The introduction of this model was, according to our studies (Lindblad and Popkewitz 2001), presented by the finance department who told the education people what was needed to be done. In a word, economy took very visibly the lead. The expectations were that such a governing system should increase the creativity and efficiency of the school system, where boundaries for school design were taken away and where informed customers should pick the best schools (Gov. Bill 1992/1993:230 “Freedom of Choice”). These ideas were put forward in a parliamentary decision on a development plan for the Swedish school (Gov. Paper 1993/1994; 183 “Development plan for education”).

Performative Regulation and Governing by the Politics of Comparisons

This governing by the market model is still at work in Swedish welfare state education. But an irritation in policy discourses were notions considering questions on quality and the validity of school merits. Thus, the model was complemented by a further development of the governing from the front, where measurements of results were given a vital position. Around the turn of the millennium, developments of networks and governance fields including data technologies dealing with large-scale assessments turned out to be very important: firstly, international study assessments like the TIMSS and PISA as tools for governance; secondly, ranking lists – more or less valid – are regarded as highly interesting news by mass media, presenting rankings of “best schools” and “most popular programs”; and thirdly, the quantitative instruments for school educators such as the Swedish databases SIRIS and SALSA were constructed as instruments for decision-making on different levels but not to be used as an instrument for ranking – which they however turned out to be. Stated otherwise a set of instruments produced to inform the governing of education are developed during the last decade. These instruments have been very successful in developing policy agendas in recent years.

Comments on the Transition of Governing Models

This history presents a series of governance models that are to a large extent overlapping and are using governing practices that are partially contradictory and partly overlapping (see, e.g., Englund (2005) on equity issues). But how are these models and practices functioning in terms of communication – or rather as a system of senders and receivers – where it is uncertain if information presented has a communicative significance (Luhmann 1995)? How significant are they in their work on governing education and what are the implications of this?

What we note is firstly recurrent irritations considering the models. Their realization is not consistent with the expectations put forward when they were implemented (as could be expected (cf., Lindensjö and Lundgren 2000). Secondly, we note an increasing complexity in the models – from a centralized to a decentralized or rather deconcentrated one with similar ideas on policy-making “inside the state” but in changing levels for decision-making.

The big break with the educational restructuring was a governing model assumed to reflect the surroundings of the welfare state institution as a combination of deregulating and marketization including a voucher system. The well-informed actors on the education market would inform schools and policy-makers about which schools that were functioning in accordance with their preferences which then the education system could act upon in order to design schools and their ways to cultivating students. The performative turn is based on a model of governing (see Ozga 2012) by means of transnational networking among an increasing amount of agents, developments of often large databases, and putting part of the decision-making outside the formal organization (Ahrne and Brunsson 2009). This governing model is in many ways a prolongation of the restructuring period, but it is using networks and technologies for comparisons that are affecting policy agendas as well as policy decisions.

On the Premises for Performative Regulation in Education

During the reformative period, education was to a large extent governed by regulated resource allocation plus directives from the center on how work should be done and procedures to follow in decision-making as well as in schoolwork. This “push model” was evaluated in relation to the directives, and sometimes the directives were evaluated themselves – their internal consistency and value in relation to the functioning of the institution pull (Hagel et al. 2010). Governing by goals or results – sometimes labeled as a “pull model” – was assumed to open up for an increased creativity among those who design education and to be sensitive for differences in context as well as for complex interaction in the process of governing and designing of education. Given the information about goals, educational activities and structures should be designed to pull in desired direction in a way that was impossible to prescribe in a general way. Governing by results is in turn based on the premises that the goals are possible to translate into valid and reliable instruments and that these measurements will show to what extent the goals are reached in a transparent way. Results achieved should then not only inform to what extent the goals are realized but also serve as a basis for analysis of reworking the design developed and to eventually revise this design.

Thus, we get a set of interrelated issues to deal within governing by results. Highly simplified these issues are assumed to constitute sets of communicative relations in different steps:

  1. (a)

    The formulation of goals in policy documents and curricula.

  2. (b)

    The design of schools and schoolwork that are expected to realize the set of goals.

  3. (c)

    The achieved results are indicated by means of certain measurements.

  4. (d)

    The measurements of results assumed to be valid relative to the formulated goals.

Governing by school performances is based on the assumption that each step is a valid translation of the previous step. However, from a communication point of view (Luhmann 1995), a valid translation is as such unlikely to occur in each step. For instance, information about goals are unlikely to be significant in designing schools. Thus, there is generally a need to reduce the complexity in communication, e.g., by training of school designers and teachers and by repetitive exchanges of information such as formative evaluations. Of specific interest when governing by results is how goals are translated into measurements. Is the eventual communication resulting in valid measurements of the realization of the goals that are expected to be fulfilled – what goals are eventually communicated and how valid are their translation? To measure educational outcomes is by no means a trivial action and so is the interpretation of results, for instance, in policy-making and in work to redesign education.

In sum, these relations present complex demands for communication in governing by results. To my understanding what is crucial here is that results matter during performative regulation of education. Thus, I will focus on (c) achieved results and their significance. Firstly, what meaning is made of these results in relation to (a) goals and (b) designs? Secondly, are these results implying actionable knowledge (McLaughlin and London 2013). The second point is most important when dealing with governing by results, since such a governing implies that information of results is of significance when preserving or altering goals or design.

The idea with this study is not to accept or reject governing by results as a workable model in the governing of education. Instead I will put forward questions about how it is governing. How, and in what ways – are school results, as displayed and analyzed by different performance indicators, significant in communication about education matters – e.g., in terms of revision of the design of education. Thus, the main question is: How do school results and performance indicators work as actionable knowledge in Swedish education?

Governing by Results During a Re-regulating Period?

First a snapshot of organizing principles is presented in models for governing by results: Part of governance by results is decentralized decision-making and responsibility to evaluate school qualities to the national authorities (cf. Quennerstedt 2006). This is based on a rationalistic model of governing, where the politicians formulate the goals and administrators decide on how to reach these goals by means of subgoals and distribution of tasks and resources. In turn, the results achieved should be collected and conceptualized by the administrators and delivered to the policy-makers (cf. Sundström 2004; Jarl 2012). In Sweden at the state level, the Swedish School Inspectorate is responsible for supervision of local authorities and schools in relation to laws and regulations, while the National Agency for Education is responsible for evaluation and aggregation of statistical data. The municipalities are producing “evaluation of quality” – reports to the NAE (Jarl 2012). The schools – in turn – are responsible to plan and evaluate their performances, including to follow up student performances. This is the responsibility of the principal in accordance with the implementation of principles of New Public Management in Swedish education.

We have noted a special twist of the performative turn, where international and supranational organizations are at work. This twist is somewhat in contrast to our conception of education as a project closely related to the national state and national policies in education. It is not only governing by results, but in doing so plays active part in policy-making in technical and administrative networks collecting and analyzing different kinds of performance and quality indicators, often derived from the OECD and the EU.

This way of doing policy is now being studied in different ways over national boundaries pointing to transitions from professional and political to a technological accountability regime (e.g., Ozga 2013) arguing that:

… accountability in education is increasingly defined as technical accountability through national and international comparative measures of performance. (p. 292)

This research portrays a regime getting a hard grip on education management and performance at all levels of education in governing through different networks of new actors and partners in education. Such a regime built on contracts and measurements of results steered by transnational organizations such as OECD and EU.

We get a technological accountability that is combined with policy-making where performances in the politics of comparisons are vital (cf. Ozga 2012). The policy-making in education is translated into administrative turns. Thus, for instance, education policy-making in Sweden is to a high extent framed by results on PISA and TIMSS – in the last election in cooperation with the McKinsey reports, which present strategies to improve educational systems as defined by success in PISA – comparisons (Lindblad 2011). See Coffield (2012) for a critical analysis of the validity of the McKinsey reports. The debate on education in mass media is dominated by a similar framing of educational performances in combination with school inspection reports presenting quality problems in certain schools and decisions to turn down schools (see Segerholm 2009, on the decision by the school inspection to close down one of the few Swedish boarding schools).

How Does Governing by Results Matter in the Context of Educational Policy-Making

In one way the results in Swedish schools are improving as well as deteriorating according to available statistics. We find that the average grades are getting higher – indicating improved education performance – especially in private schools – during the last 15 years (Skolverket: Beskrivande Data 2012, p 38). However, during the same period, the results are going in the opposite direction when analyzing international student assessment exercises such as PISA and TIMSS. In accordance with the current policy agenda, educational policy statements to a large extent are referring to results from international assessment exercises – on low performances on, e.g., PISA – and demand to improve the situation. According to international and national statistics, the progress is not alarming (Skolverket: Beskrivande Data 2015, especially p. 168 ff). Instead, results based on international comparisons have a dominant downhill tendency. In international ranking lists, Sweden is sinking.

Table 4.2 gives a short overview of the development of results considering such performances during the last 10–15 years. This picture of school results is somewhat ironic in a period where governing by results is at work and especially when international comparisons are setting the agenda for Swedish educational policy. (The interpretation of the sinking results is not infrequently – e.g., by the current government – assumed to be a result of a domination of progressive school ideologies still at work in schools and in teacher education. An alternative interpretation is that they present consequences of mistakes in current education policy-making.)

Table 4.2 Broad tendencies in comparing comprehensive school performances 1998–2012
Full size table

Given the current reception of results from the OECD and the IEA, the following conclusions are made:

  1. 1.

    High visibility of international performance comparisons in Swedish educational policy discourse

  2. 2.

    Low visibility of policy reflections about reasons behind these results

  3. 3.

    High number of educational policy measures in order to improve the results

The combination of (1) and (2) is interesting: less of reflections and more of activity in measures to improve performances in international comparisons. Such measures concern more distinct governing and control of students, to strengthen the competences of school professionals and to reinforce school inspections.

Education policy discourses are increasingly focusing on school results as measured by different kinds of testing. This is further underlined by the current analysis of education policy measures since 2007. It is argued that what matters are improved school results, which is limited by the capacity to measure these results. Given this, the point is very simple – what matters most is improvement – e.g., in terms of position in international comparisons, not the results as such. How to achieve this is, according to the commission, to get more competent teachers and school leadership – in line with current education system analyses such as the McKinsey reports (Barber and Mourshed, 2007, Mourshed et al. 2010).

How Does Governing by Results Matter in School Organizing?

How is governing by results functioning – how is the communication working at the school level – what is the significance of school performances in the organizing of school activities and programs? In our research we tried to answer such questions by means of a ten-school study. Here, Jarl (2012) made an intensive study – based on document (forms, plans etc. used to document student performances) analyses and interviews – of four schools with different merit scores but with other characteristics kept as constant as possible.

In this study it was concluded that most documents concerned identification of problems and deficits and little about good performances. Consistent with this finding is the fact that “poor-performing” schools had much more documentation. From the interviews it was concluded that the principals were more or less undetectable – they are not taking the lead in this work – and the school results are not having an impact – at least in the formal organizing of the school. These findings are consistent with other research showing that the principals cannot manage to carry out pedagogical leadership under the pressure of current constraints. Jarl points to the fact that this was a prerequisite for the governing system in Sweden and that the lack of pedagogical leadership is a huge problematic in current education: school results does not matter much in the organizing of Swedish schools. What matters instead for the principals is to keep the economy in order. Otherwise they will be replaced or the school is in danger to be closed down.

To this I would like to add the notion of triage presented by Youdell (2004) stating that in a marketized context “…practices of educational triage becomes both acceptable and necessary.” (p. 407). A triage (a concept from emergency medicine) differentiates students in the safe cases, the suitable cases for treatment, and the hopeless cases. To identify the cases that can be treated within current constraints and affordances is then a way to increase school performances at the cost of resources to the hopeless cases according to Youdell (op cit). To my understanding the concept of triage can be used to pinpoint practices of inclusion/exclusion in schooling as an institution, where an important aspects concern the saving of souls in danger – diminishing the number of dropouts.

Students’ School Performances and School Choice

What is the significance of school results for students in their identity development and choice of school careers? Several studies point to differences among students in ways of dealing with schoolwork and to respond to demands of schooling. We also learn how different youth cultures display themselves in relation to each other inside as well as outside classrooms. Here we assume that scores on tests in combination with teachers’ responses on student answers are vital not only for students’ understanding about their knowledge status but also about their understanding of themselves and their position – e.g., in the school class as a social system.

Much attention is given to student performances and how to supervise and correct students in order to improve school results during the last decade. What significance does this have from a student point of view?

Firstly, going to classroom interaction, it has to be noted that this is highly complex today with less of simple IRE triptychs (teacher interrogation – student response – teacher evaluation) in formative evaluation of student performances (cf. Mehan 1979; Hoetker and Ahlbrandt 1969). In our research we made a set of detailed analyses based on recordings of classroom interaction showing that the teachers are still in control but participating in more extended sequences with more of student responses and less teacher evaluation indicating problems to identify straightforward right/wrong responses in current classroom discourses (Hansen et al. 2015). Other studies have shown increasing classroom governmentality where students are to choose their tasks and plan their ways to manage these tasks (Österlind 2010).

Secondly, a very important stage in school careers is the transition from lower secondary to upper secondary school. This implies that students and parents have to select school as well as school program. A large set of information sources and guidance activities are provided to the students in order to support their “informed and rational choice.” Hansen and Lindblad (2011) identified a set of student identities, based on information about their school efforts, homework, etc., and choice rationalities. Given this he analyzed the significance of different kinds of information for their school choice. Irrespective of identity as well as rationality, what matters are personal and informal contacts, firstly parents and siblings and secondly “shorter work experience programs” plus face-to-face communication in an annual education fair, where upper secondary schools present themselves and try to get applications to their alternatives. What seems to be of less significance is guidance talk, teacher conversations, websites, and visits to schools.

A preliminary conclusion is that students are to less extent “pushed” through the school by teachers and teaching learning materials. They are to a larger extent expected to “pull” themselves through their school life by motivation to improve their grades and build their future. According to our studies, there is a number of students that are not part in this pulling machinery. Firstly, they are not at present engaged in the future offered by schooling. And secondly, they are not engaged in classroom discourses. The concept of triage can be used here as well, having striking similarities to the identification of so-called steering groups by Dahllöf (1967) in the Swedish school during the 1960s. This is interpreted that schooling as an institution in Swedish has a focus to push students “at risk” into more safe positions since long. It also means that a significant share of students is more or less outside the working of classroom pedagogy.

Concluding Remarks: What Matters Most in Governing by Results?

As a point of departure, a periodization based on governing models in Swedish education since the WW2 was presented. It was argued that a restructuring period since the early 1990s was moving over into a period of performative regulation where education policy-making was increasingly based on the politics of comparisons – where local, national, and international ranking lists (such as PISA and TIMSS) are of vital concern. Given other research (such as Ozga 2012), contexts for policy-making seem to be in change – toward the working of transnational fields with different partners and networks – with changes in accountability and expertise.

However, this performative turn has not been connected with flourishing results according to measurements that are considered as vital – e.g., international comparisons. The measures taken by the government point to development of the work of teachers and the control of student performances to be vital to improve the situation. Our findings from ongoing research on lived curricula and school results point in a somewhat other direction:

  • The organizing of schools and communities seem to focus on the financial situation – in making and keeping the budget. There is little room for analysis of the pedagogical situation and to develop theoretically based development strategies to pull the schools forward.

  • The schoolwork does not fit a pull model for all students. A substantial number of students show little of interest to pull themselves into a desired future. Actually the current policy measures seem to work in the opposite direction – to correct and to push these students.

  • This is combined with a somewhat diffuse accountability picture. Transnational organizations are defining the goals, while teachers and students seem to be accountable for realizing these results.

  • There is little educational analysis on the ongoing turns and their implications. Actually, governing by results is – at least in theory – inviting to retrodictions, to analyze why the current outcomes had to be in this way (see, e.g., Von Wright 1983). But actually little seems to be done of this kind of intellectual work!

Given these conclusions – governing by results is a problem that is demanding an expertise in analyzing education, in comprehending why these results occur. This is not about measurements as such – it is about putting things right in educational analysis.

Stated otherwise, there seems not to be a lack of measurements in Swedish education of today. What is missing – given the conceptual framework used here – are analyses of the working of the education system, e.g. its curricula, pedagogy, and evaluation, referring to Bernstein (2000) and its meaning in societal and cultural contexts. Current ways of dealing with school performances do not seem to be very meaningful in that respect. Given this it is today of vital importance to analyze the working and functioning of current and emerging expertise in education under a performative turn.

Based on these conclusions, it seems to be reasonable to expect an increasing irritation considering governing by results, first in terms of new measurements and analyses of patterns of school outcomes (see, e.g., the OECD report 2015) and second by analyses of schooling processes in order to capture causal mechanisms and not only patterns or correlations. And what is regarded as valid directives for action – from the inside or outside welfare state education – is presumably a matter of trust in the teaching profession or in national or international governing expertise. Or in other words – a matter of politics of knowledge.