Evaluation in Local School Governance: A Framework for Analysis

Abstract

This article develops a conceptual framework for exploring the role and consequences of evaluation at the local level of school governance. It also provides a frame of reference for the articles in this special issue on the role of evaluation in local school governance in Sweden. It consists of key concepts, three models of decentralised governance (state model, local government model and multi-actor model of governance) and four types of evaluation (indicator-based monitoring and evaluation systems; stand-alone evaluations; synthesis studies; and informal, occasional or everyday evaluations). Local school governance refers to governance that occurs in a municipality and in a quasi-market where local school actors govern and influence schooling and education. It includes the efforts of actors and institutions to govern and influence matters such as school policy, education, school climate and school safety. Evaluation is used as a generic term that refers to, for example, evaluation, inspection, quality assurance, ranking and to both stand-alone evaluations and evaluation systems. The article briefly demonstrates how the framework can be applied in an analysis of the role of evaluation at the local level of school governance by providing an example, and discusses the framework's advantages and limitations.

• conceptual framework
• evaluation systems
• local school governance
• functions
• constitutive effects

Introduction

Education research used to inform policy underscores the importance of teaching quality in improving student learning and performance (Hattie 2009) and of holding schools accountable for learning outcomes (i.e. performance) rather than for policy compliance (Atkinson et al. 2009; Hamilton et al. 2008). Policymakers increasingly view evaluations and evaluation systems as the main tools for improving teaching, education quality and student learning. For example, it is assumed that international and national evaluations of education systems (Carvalho 2012; Hanberger 2014; OECD 2009; Supovitz and Taylor 2005) and teacher evaluations (particularly the new generation of teacher evaluation) help improve education and learning (Hallinger, Heck and Murphy 2014; Reynolds et al. 2014). Further, evaluation and strengthened accountability are expected to enhance education and schooling quality (Ehren et al. 2013; OECD 2011). However, whether, how and in which conditions evaluation can contribute to this are insufficiently understood (Levin 2010; Ravitch 2010).

Recent research and a growing number of scholars question the value of many of today's evaluations and evaluation systems, and of focusing too much on performance and measureable results (Dahler-Larsen 2013; Lingard and Sellar 2013; Ozga 2009). Leeuw and Furubo (2008) claim that evaluation systems produce routinised information for single-loop learning and that the systems produce information that confirms rather than questions policies. Moreover, response systems to evaluations (Hanberger 2011) have created performance demonstration behaviour, and uncritical reporting of the effects of evaluation, and post-rationalisation of action taken to meet the performativity and accountability norm (Hanberger 2011). Similarly, strategic behaviour is a common response to performance measurements (Hood 2006, 2012; Propper and Wilson 2003) and standard-based accountability systems (Hamilton et al. 2008); for example, the tendency to focus disproportionally on what is measured by targets, paying less attention to what is not measured, is referred to as output distortion.

It is also recognised that teachers are subjected to too many accountability measures, which can have negative effects on professionalism and education (Ball 2003; Day 2002; Evetts 2011; Green 2013; Hamilton et al. 2008; Hargreaves 1994; Koretz 2009; Mausethagen 2013a, b). Better understanding the prospects for evaluation and evaluation's consequences for school practice calls for an examination of the interplay between evaluation and governance (Hanberger 2013) and exploration of how various local conditions and education systems affect evaluation. Which roles and functions do evaluation and evaluation systems have in local school governance? Can some kinds of evaluations promote school development in certain conditions, or do they mainly have negative consequences for school practice?

The guiding assumption of this article and this special issue is that evaluations can have positive, negative or mixed consequences for municipal school policies, schools and the daily lives of teachers and students. Further, existing evaluation systems can, singly or jointly, have indirect or constitutive effects (Dahler-Larsen 2012a, b; Hanberger 2013). For example, evaluations can help shape the conception of what is important in education, emphasising what is easily measurable, and promote a performance-oriented discourse on quality in education (cf. Propper and Wilson 2003).

The purpose of this article is to develop a conceptual framework for exploring the role and consequences of evaluation at the local level of school governance in decentralised education systems. It also provides a frame of reference for the articles in this issue. Although the three models of decentralised governance depicted here are normative, reflecting different notions and principles of democratic governance, they are used here to reveal the models’ evaluation needs and evaluation function.

The article briefly demonstrates how the framework can be applied by providing an example of how to explore the role of evaluation at the municipal level.

Key concepts

Evaluation is used as a generic term in this issue, referring to both stand-alone evaluations and evaluation systems. We apply a broad understanding of evaluation, encompassing evaluation, audit, inspection, quality assurance, ranking etc.

Evaluation system refers to “the procedural, institutional and policy arrangements shaping the evaluation function and its relationship to its internal and external environment” (Liverani and Lundgren 2007, 241). The evaluation systems explored here refer to routines established for dealing with stand-alone evaluations at the municipal, school and classroom levels, and to systems that produce streams of information, such as monitoring, inspection, oversight and benchmarking systems (cf. Hanberger 2011; Leeuw and Furubo 2008; Rist and Stame 2006).

Governance refers to the new, emerging institutions for collective action that have evolved from traditional forms of government, including negotiated interaction between a range of actors and institutions (Klijn 2008; Marcussen and Torfing 2007; Pierre and Peters 2000), and to three changes in state functioning: the rescaling of political authority in the context of globalisation, the reconstitution of a “competition state and new regulatory mechanisms associated with ‘new public management’”, and moves to deliver public policy through multiple forms of privatisation (Lingard 2011, 367). Traditional forms of governing exist parallel to new forms of governance and are included in the concept.

What we mean by local school governance will be further illustrated in the articles but, briefly stated, it refers to governance that occurs in a municipality and in a quasi-market where local school actors govern and influence schooling and education. A quasi-market implies the public regulation, funding and inspection of schools, the accreditation of ‘free schools’, vouchers, companies running schools for profit, competition between school owners, and parents’/students’ active school choice (cf. Lubienski 2009). The definition includes the efforts of actors and institutions, such as the state, municipal education committees, opposition parties, school owners, school principals, teachers, parents, students and pressure groups, to govern and influence matters like school policy, education, school climate and school safety. This issue explores governance at the municipal, school and classroom levels.

Accountability has traditionally referred to external scrutiny and accounting (Mulgan 2000). However, its meaning has expanded and the term now has more connotations and is linked to other concepts, such as trust and fairness. Democratic accountability refers to the citizens holding governments, at various levels, to account for public policies, to the state holding local governments and public agencies to account for national policies and programmes, and to local governments holding municipal departments and service providers to account for implementing local policies (Hanberger 2009). Besides these forms of vertical accountability, officials and street-level bureaucrats (e.g. teachers) are subject to horizontal accountability in the wider accountability environment (including parents, the media etc.). According to Behn (2001), democratic accountability includes accountability for finances (i.e. keeping to budgets), fairness (i.e. following rules and ethical standards) and performance (i.e. monitoring the outcome and consequences of public policy). In this issue, accountability refers to both vertical and horizontal accountability, the focus being on accountability for compliance and performance. States use standard-based accountability policies (Hamilton et al. 2008) and accountability systems (Ehren et al. 2013; Trujillo 2012) to improve education through vertical accountability. For example, Swedish schools are held accountable for implementing national school policy and for the quality of their performance measured in national tests of students.

School development refers to changes in schools, teaching, and learning environments conceived as, and claimed to be, changes for the better. It refers to the process of improving schools and education, including improving teaching, learning environments, teacher and student engagement and motivation, and school and student performance. It does not refer to a predefined state or to specific characteristics or conditions in a school or education system because what is conceived as ‘development’ changes and varies with one's conception of the role and meaning of education in society. This implies that a shift from one condition to another can be conceived as, and claimed to be, either a case of school development or a change for the worse. Analysis of the effects of evaluation on school development in this issue integrates knowledge from school effectiveness and school improvement research.1

Local school development is accordingly defined as improvement in school culture, teaching, equity, and school/student performance.

Two concepts are used to identify the role of evaluation in local school governance. Evaluation use refers to actors’ use of evaluation. Functions of evaluation can be indicated by actors’ use but go beyond actual use and include aspects such as how evaluations facilitate public accountability or legitimising the current governance model, for example.

The concept consequences is used in a broad sense and includes the effects and influence of evaluation on governance, policy, institutions and actors. A consequence can prevail as a direct or obvious effect, e.g. when a school inspection initiates an immediate change in school policy or school practice. Constitutive effects refers to tacit or indirect effects, for example, how evaluation (systems) can shape discourses, define what is important in education and school systems, and reinforce actors’ role in local school governance (Dahler-Larsen 2013; Lingard and Sellar 2013; Power 1997).

Evaluation and school governance

In recent decades, evaluation has expanded radically at all levels of governance as part of the broad doctrine of new public management (NPM) (Hood 1991; Pollitt and Bouckaert 2011). According to this doctrine, market mechanisms should be introduced to enhance efficiency and, in the context of school governance, to support competition between schools, free school choice, improved educational quality, and school effectiveness (Merki 2011; NAE 2012). Developing evaluation systems is essential to school governance characterised by results-based management and increased local autonomy (Andersen, Dahler-Larsen and Strömbæk Pedersen 2009; Elstad 2009; Merki 2011; Mintrop and Trujillo 2007; OECD 2013). Without measuring performance and assessing students and teachers regularly, today's school governance models will not work and the centre will lose control.

The growth of evaluation systems reflects expectations that evaluation is an effective way to improve education. Schools, school principals, teachers and students have become objects of evaluation, and how they perform needs to be monitored (Dahler-Larsen 2012a, 2013). Setting clear and measurable goals and regularly monitoring goal achievement together are assumed to enhance performance accountability and improve schooling. Evaluations and openly accessible performance data are also prerequisites for parents and students to make informed school choices (Musset 2012). In market-oriented school governance, “parents would exit their neighbourhood schools whenever it is feasible to obtain better educational quality at an equal or lesser cost” (cf. Hirschman 1970; Musset 2012, 7). This implies that parents and students should evaluate their schools and then take action if the schools fail in any way. Parental ability to switch schools can be conceived as a form of market-oriented governance. However, the actual consequences of school choice policies and market-oriented school governance are highly contested (cf. Ravitch 2010).

As evaluations are central to school governance, they need to be conceived and studied from a governance perspective (Bouckaert and Halligan 2006; Hanberger 2013; Lingard 2011; Ozga 2009). The role of public and independent schools in a country's education system largely depends on the division of power between levels of government and on the discretion given to local governments and independent schools (Holmgren et al. 2012). This discretion varies over time, depending on changes in national school governance and other factors, such as being subjected to global education governance as a member of the EU and the Organisation for Economic Cooperation and Development (OECD).

How evaluation systems are set up depends greatly on the governance structure and model applied (Foss Hansen 2013; Hanberger 2013) and supposedly reflects the knowledge needs of the governing actors. National evaluation systems serve national actors’ knowledge needs, whereas independent schools’ evaluation systems serve their owners’ information needs and local municipal schools’ evaluation systems serve local government needs. Further, stakeholders in the wider accountability environment (Behn 2001) also need information, for instance, about schools, student performance, parents’ and teachers’ satisfaction with education, and local conditions. Although evaluation systems are developed to serve many purposes, they are often used for performance accountability but without meeting the necessary conditions for such use. Not all the conditions for accountability are in place; for example, there are no clear principal-agent relationships or formal rights for principals to pass judgment on the accounts of agents and impose formal consequences, making it problematic to use evaluation systems to hold agents accountable for student and school performance (Lindgren, Hanberger and Lundström 2016).

What governments do in terms of school reforms and policies, including developing and implementing evaluation systems, is assumed to determine how school systems perform (Levin 2010). However, other factors, such as students’ family backgrounds, matter more. No matter how well countries or education systems perform, more can be done, as international comparisons have become the standard for judging accomplishments (Levin 2010). It is very difficult to improve performance in large education systems in a short period of time. This is mainly because measured results reflect “larger problems and inequalities in societies” (Levin 2010, 740) that schools cannot completely overcome, and because school-improvement measures are often implemented poorly. Knowledge of how to improve education practice in whole school systems is lacking, and “in most cases there are few effective levers for doing so and often too little interest by governments in that work” (Levin 2010, 740). Although this knowledge is available, policymakers in Sweden and elsewhere tend to listen more to international organisations such as the OECD, following their advice on reforming education systems because actionable knowledge for policy is what they need and request (Breakspear 2012; Carvalho 2012; OECD 2009). Moreover, when improved performance is measured using international or national student tests, the tests’ validity and reliability also need to be considered. Policymakers, however, generally express high expectations regarding evaluation and consider international and national student tests trustworthy measures of quality in education systems, using the results to advocate certain policy solutions for managing problems in today's schooling. When measuring performance, it should be recognised that there is no consensus in the research or policy community regarding what is meant by improved performance and school development.

Evaluation in local school governance

Several studies reported in the education and evaluation literature treat the role of evaluation in local school governance. In a survey conducted in 1996, Dahler-Larsen (2000) found that administrators working in various sectors of Danish municipalities largely used evaluation as management support in adapting their work and in programme improvement, with this use and function of evaluation being interpreted as “an appropriate response to the double pressure of stability and self-inspection” (2000, 86) and as an adaptation to the norms of “reflexive modernization” (with reference to Ulrich Beck).

Trujillo (2012) studied the relationship between one urban school board and the democratic governance processes under high-stakes accountability conditions in California, finding that board members were unified in their conception of individualised goals for teaching and learning and in their belief that standardised test scores reflect effectiveness and efficiency in education. However, there was a lack of democratic processes in the board as well as a lack of “reasoned, public discourse”. Local school governance was characterised as “autocratic, illegitimate policy-making and a culture of fear within the central office and schools” (2012, 346). The study also found that the board acted as a committed implementing body of centrally determined education policy. Board members “believed in the power of a centralized, result-based system” and in “outcome-oriented accountability”, which represented a way to cope with “high-stakes accountability policies, and to protect their own power to govern” (2012, 351). The study indicates that school boards can choose to operate mainly as implementation bodies for the state.

In contrast, a recent comparative study of two Midland cities’ responses to the so-called Academies Act, which allows more freedom of choice to academies (i.e. publicly funded private schools) than to schools run by local authorities, demonstrated that the two cities differed radically in their response to the Act (Smith and Abbott 2014). One of them manifests a strong culture of acceptance and the other a strong culture of rejection. The explanation for the differing responses is traced to the priority placed on local education earlier in history (indicating path dependence) and to the integration of the Act in the local context and culture. The authors demonstrated that the local context matters in terms of whether local authorities accept and implement state policy with or without resistance.

These studies note varying local school governance conditions and demonstrate that the role of evaluation can vary according to how local actors assimilate, adjust to or resist state policies. Actors at the municipal and school levels have varying freedom of choice in different education systems and can respond to external demands and accountability pressures and use evaluations in different ways. Hence, the local context matters in various ways.

Conceptual framework for exploring the role of evaluation in local school governance

Against this background, a conceptual framework is developed to support the analysis of evaluation in local school governance and to provide a framework for the articles. The articles were not developed according to this framework at the outset, so they focus on different elements of the framework, without all elements being explored in each one.

The framework starts from the key concepts defined earlier. It also pays attention to three models of decentralised governance and distinguishes four types of evaluation that can support and affect local school governance and practice. It considers six possible functions that evaluation can have in the context of local school governance, as well as the constitutive effects of evaluation and evaluation systems. The assumption is that a governance model can shape evaluation to meet the governance actors’ evaluation needs, and that governance and evaluation interact. Although a certain governance model sets up evaluation systems and develops evaluations to meet its needs, evaluation should, and can, also affect governance. The framework recognises that actors can act more or less according to their mandates, and that their responsibilities can be interpreted differently, but also that they can act beyond their mandate.2

Local school governance, as conceived in this framework, is confined by a country's political system and governance structure. Three models of decentralised governance are described and used to reflect the governance structure and prevailing governance models in a decentralised education system. These models are ideal types (Weber 1962), but are not mutually exclusive. Although they overlap and share some characteristics, the models’ basic sources of governance legitimacy differ. The models can prevail in different combinations in different communities and education systems. The next section briefly describes the models in terms of their governance core, democratic orientation,3 accountability, and evaluation needs.

Three models of decentralised governance

The state model describes governance that starts with a decision made or programme developed by the government, elected representatives or administrative elite at the national level intended in this case to control, change or improve schools and education at the local level. The power of the state is codified in the constitution and in this model, the municipal, school and classroom levels are implementation levels subject to governing from the state. The democratic orientation is basically that of a representative, elitist model of democracy (Dryzek 1996; Hanberger 2009). Citizens can hold national representatives to account and influence the general direction of school policy in elections: that is, democratic control is assumed to be exercised when citizens vote either to keep the government in power or for opposition party candidates. School policy is implemented from the top–down in this model and demands strong management control to work well. It is a model of governance in which the state is in command, and that requires evaluation to monitor and control the achievement of national objectives. National agencies (e.g. the School Inspectorate and the National Agency for Education) undertake inspections and evaluations on behalf of the government and elected representatives to inform them about policy implementation, how the education system is performing, and the achievement of national objectives.

The Swedish state has delegated several responsibilities to municipalities, independent school owners, and principals, giving these actors substantial discretion to adapt schools and education to local conditions. The state is still in command and uses normative steering instruments (mainly management by objectives and results-based management) together with legal and economic instruments to govern schools. It is recognised that several school reforms have made the governing of schools difficult and complex in the Swedish education system. The state governs local school actors in two ways. It asks municipalities and school owners to create the conditions needed to achieve national objectives, and governs principals and teachers directly by demanding that they create conducive conditions and provide adequate support so that students can achieve the goals of the curriculum and course plans (Skolverket 2011). In addition, the municipalities govern public schools through municipal school policy and independent school owners govern their schools in line with their school policy. The chain(s) of governance and the division of responsibility between the state, municipalities, school providers, school principals and teachers can thus be regarded as complex and somewhat unclear (Holmgren et al. 2012; Montin 2007; Skolverket 2011; SOU 2015, 22).

The local government model is a model emerging from the discretion of local government.4 It is based on constitutional rights and the delegation of power. The discretion for local school policy and governance can be more or less limited by the state and can vary over time. The legal grounds of this model are primarily decisions made by local government. In countries where local governments wield considerable power and freedom of choice, the scope for local school policy and governance is extensive (cf. Lidström 2004; Montin 2007). The democratic orientation is the same as in the state model, but in this case the relevant political and administrative elite is located at the municipal level. Citizens can control local government school policies directly in elections and can vote for opposition parties if they want change. Municipal auditors have the commission to inspect local government regularly. In the Swedish case, they act as public reviewers with the right to undertake audits and investigations on their own initiative, acting for the local assembly and citizens.5 Local government bodies (e.g. education committees) need evaluations in order to monitor and evaluate the achievement of local government objectives. They can also use national evaluation systems and data when appropriate and develop their own evaluation systems if needed.

The multi-actor model describes governance situations in which governments (local, in this case) share power with other actors. It refers to governance situations in which public actors and institutions form networks and partnerships to resolve problems and challenges (facing schools, in this case). This model also includes multilevel governance in which power is shared between levels of government and between non-governmental actors or institutions, thereby also capturing governance in the civil and private spheres. The legal grounds for this governance model are the delegation of responsibility as well as agreements and contracts between participating actors and institutions. Its democratic orientation can vary, heading in the direction of elitist, participatory, or discursive (deliberative) democracy. This model's democratic orientation, however, has the most in common with that of participatory or discursive democracy. This governance model needs evaluation that facilitates network governance and meets key network actors’ knowledge needs. In this issue, attention is paid to the school governance roles of professionals (school principals and teachers), parents, students and citizens.

Professionals’ role in local school governance rests on the delegation of power from national and local governments. In Sweden, school professionals are expected to implement national and local school policies, regulations, and requirements set forth in the Education Act, and to use their discretion to act as professionals according to the knowledge and skills attained through teacher and school-leader education and practice. They are accountable to the state/local government or independent school owners and to parents. These professionals need evaluations to identify problems and challenges in teaching, to monitor student learning and development, and to learn how their teaching and other school activities should work in practice. School principals and teachers have more or less the same knowledge needs when principals act as pedagogical leaders; however, principals acting as managers need monitoring and performance information to control teachers and hold them to account, for example, for student results.

Parents’ and students’ roles in local school governance are enacted when these actors make informed school choices, when they influence teachers, school principals, school administrators and local school boards, and when they vote in elections. Although citizens have a role in the state and local government model of governance, it is restricted to holding national and local governments to account in elections. Parents can also act as responsible citizens, in line with the participatory or deliberative democracy models, or as customers in a school market, reflecting a market-oriented democracy model. Parents’ and citizens’ evaluation needs are associated with the roles they play in local school governance and in developing their schools. Parents acting as customers need easily accessible performance data to support informed school choice, whereas parents acting as active and responsible citizens largely need the same evaluations as do other policy actors. Prior research into parents’ information needs indicates that they are primarily concerned with school atmosphere, pedagogical climate, safety and reputation, and do not need the information provided by school inspections, for example (Ehren, Leeuw and Scheerens 2005, 71). Although most parents can be expected to take little action beyond that directly concerning their sons’ and daughters’ education and well-being, those who make informed school choices or actively attempt to influence education or school safety, for example, are conceived as policy actors operating in the multi-actor model of governance.

Four types of evaluation

The framework also distinguishes four types of evaluation. The first type comprises indicator-based monitoring and evaluation systems, producing recurrent data on the inputs, processes and outputs/outcomes of policies and programmes and on school or student performance. Such evaluation can focus on monitoring implementation or assessing the achievement of objectives (Hanberger 2013; Rist 2006). Examples of this type of evaluation are international and national student tests (e.g. PISA, TIMSS, and national mathematics and literacy tests) and evaluation systems consisting of input, output and performance data on schools and students (e.g. SALSA and SIRIS).6

The second type of evaluation refers to stand-alone evaluation studies that generate quantitative and/or qualitative knowledge. Applying a broad understanding of evaluation, this type includes in-depth evaluations, inspections and municipal audits. These evaluations can use information produced by the first type of evaluation, subjecting it to additional and systematic inquiries, analyses and assessments. Examples of this type are the National Board of Education's thematic evaluations, evaluations of school reform implementation, the School Inspectorate's various inspections, and education committees’ in-depth evaluations of schools.

The third type of evaluation refers to synthesis studies that compile knowledge from a number of the first and second types of evaluations as well as other relevant studies. This type includes synthesis reports produced by the School Inspectorate or the National Board of Education and municipalities’ systematic quality work reports.

The fourth type of evaluation refers to informal, occasional or everyday evaluations conducted by local school actors to meet their personal knowledge needs. Such evaluations are initiated by individuals or small groups and can consist of systematic assessments and be documented or undocumented, but they are not made official or published. Examples of such evaluations are politicians’ and administrators’ own evaluations made during site visits and teachers’ evaluations of their own or colleagues’ teaching.

The above evaluation typology differs somewhat from the schema of the four streams of evaluative knowledge discussed by other evaluation researchers (Stame 2006, x). The fourth type of evaluation is not generally attended to but should be, as it can be of major significance. The four types facilitate exploration of the roles and functions of evaluations in local school governance.

Functions of evaluation in local school governance

The framework also considers six functions that evaluation can support in local school governance: governance, school development, accountability, critical learning, legitimisation and symbolism. The first three functions have been defined above and these and the other three functions are discussed further in a separate article (Hanberger 2011). These functions of evaluation can manifest themselves in different ways and at different levels.

Brief demonstration of how the framework can be applied

In an explorative study some kind of case study would be a feasible research design. The choice of cases should reflect differences in conditions that can affect evaluation in local school governance, for example, size of municipality, political majority, school performance, critique from school inspection, socio-economic factors etc.

As an example, to explore the role of evaluation at the municipal level there is a need to compile information about the municipality's school governance model, school policy and evaluation system from a close reading of websites, policy documents and minutes from Education committees. This reading helps the researcher identify who the key actors are; from a governance perspective they include politicians in power and central administrators, and politicians in opposition and auditors commissioned to audit local governments and administrations, and other actors that could be involved in municipal school governance. Interviewing opposition politicians and auditors could provide a different and more critical understanding of the role of evaluation, and thus give a more comprehensive picture of the role of evaluation in this context.

The framework and key concepts guide the interview questions and they should capture the use of external and internal evaluations, responses to transnational, national and local evaluations. Questions should be open-ended and semi-structured. They should facilitate the interviewees to share what is important for them, and their understanding of which (type of) evaluations are the most valuable for local school governance and for what (governance, accountability or school development), but also which evaluations affect them the most. Questions about how the actors conceive the consequences of their own and external evaluations and evaluation systems should also be included.

Analysis of documents will include qualitative content analysis, asking specific questions of each text (including interviews) based on specific research questions (Bryman 2012). The framework and key concepts, the three models of decentralised governance and the local government model in particular at this level, guide the analysis. Actors’ use and non-use of evaluations reflect their perception of the role of evaluation, how they respond to external evaluations and manage the evaluation web. The actors’ line of reasoning is central in this framework and so is the researcher's dialogue with the data and interpretation of empirical findings (Gottweis 2006; Fischer 2003; Hajer and Wagenaar 2003; Yanow 2000). ‘First-level interpretations’ (made by school actors) and ‘second-level interpretations’ (made by researchers) (Yanow 2000) of the role of evaluation in local school governance are analysed in relation to a given multilevel governance structure, which provides local governments with constitutional power and some discretion in the education system. The framework and the theoretical considerations it is based on assists the researcher to interpret the actual functions which goes beyond actors’ use, and more tacit and constitutive effects of evaluation in this context.

This is a brief demonstration of how the framework can be applied at the municipal level. It can also be used at the school, classroom and parents’ level of governance, and to synthesise the roles of evaluations at these levels. In this issue, the application of the framework at these four levels will be demonstrated.

Conclusion

The open framework and the definitions of key concepts reflect a broad notion of evaluation, governance and school development which can be used for exploring the role of evaluation in local school governance in decentralised education systems. An open framework is justified because research into evaluation practices in local school governance is scarce, and because governance and evaluation can evolve in different ways. Focusing on only one type of evaluation e.g. school inspection does not reflect the governance and policy context in which school inspection operates and how it is affected by, or interplays with, other evaluation systems. One downside of adopting an open framework and a broad approach is that specific evaluation systems that can have a significant impact may by overlooked or analysed too shallowly. However, it is not possible to empirically study the role and consequences of all evaluations at the same time. The framework makes it possible to represent prevailing evaluations and evaluation systems in a crowded policy space, and the interplay and dynamics between evaluation systems and between evaluation systems and governance. It can be used for comparisons between municipalities in one country or between municipalities in different countries, where the local level of governance has been delegated power and discretion.

The framework is developed from a combined bottom–up and top–down approach because evaluation and local school governance in decentralised education systems are embedded in a multilevel governance structure. A limitation of a bottom–up approach is that the significance of the state model of governance and its effects can be understated. Exploring national policy implementation and the influence of evaluations on local institutions and school actors from a top–down approach would focus on achievements of national objectives and implementation problems. This would not reflect the context in which local school actors operate, and critical responses to national evaluations and lack of compliance would be conceived as implementation problems, for example.

The framework supports two interrelated explanations concerning the actual roles of evaluation in local school governance. First, the governance structure and the applied model of governance confine and partly shape the role of evaluation at the local levels of governance. Second, how local school actors use their discretion and interpret their role in the education system, including how they respond to the accountability pressures, explain how the role is realised, and that actors at the same level of governance can develop partly different roles. Further, local school actors can have different visions of and interests in how education should be developed, and these affect how they conceive of and respond to evaluations.

The three governance models provide three maps for analysing the role of evaluation in multilevel governance in nation states with three levels of government. They help explain and enhance our understanding of why evaluations are developed and used in various ways in local school governance. The state model depicts national evaluations as a tool to provide the government with performance information about national school policy and compliance with laws and regulations. The local government model, used here solely to depict evaluation at the local level, can also be used to reflect the role of regional governments in school governance, that is, in education systems where the regional level has been delegated more responsibility (Hanberger 2009). (In Sweden, the regional or county level has limited responsibility for education.) The multi-actor model also reflects governance in which local, global and transnational actors participate. The OECD's role and the impact of PISA, for example, can be weak or strong at the local level, and the impact of global school governance can change and vary between different municipalities and countries. So far, the main OECD-PISA influence is on the national level of governance, but the influence is transformed through the state model of governance and has affected governments’ governance by numbers (Lingard 2011) and the focus on enhancing student performance, for example. This model also reflects the influence of school companies and other school actors operating in a school quasi-market (e.g. business organisations, teachers, unions and parents/parents’ organisations).

The framework incorporates the interplay between governance and evaluation, as governance shapes and prescribes various roles for evaluation, and as evaluation supports and assesses governance in various ways (Hanberger 2013). The three ideal types should be used together in analysing prevailing evaluation systems because all three models operate simultaneously. Thus, a mixed method and combined bottom–up and top–down approach seems feasible for exploring evaluation in local school governance when evaluation and local school governance are embedded in a multilevel governance structure and can emerge in different ways.

Although the key concepts and three models of governance frame the analysis of evaluation in all articles in this issue, these articles were not developed solely according to this framework, so not all elements of the framework are explored in each one. In the final article, the authors recapitulate and discuss the framework's advantages and limitations in light of the preceding empirical studies in the issue.

Additional information

Notes on contributors

Anders Hanberger

Anders Hanberger is Professor in Evaluation and director of research at the Umea Centre for Evaluation Research, Umea University. His research includes public policy and governance, evaluation, evaluation systems, the interplay of evaluation and governance, and evaluation methodology. Special interests include the role of evaluation in democracy and accountability. Email: [email protected]

Notes

1 In school effectiveness research, “student achievement” is the most important effectiveness criterion and school quality and equity are two key concepts. “School quality” refers to “the degree to which a school scores better than other schools, corrected for student intake characteristics”, and “equity” “to the compensatory power of schools”, for example, regarding gender and ethnicity (Reynolds et al. 2014, 205). School improvement research focuses on “change and problem-solving in educational practice” (Creemers and Reezigt 1997). School improvement does not occur if the “school culture” – a key concept here – is not “favourable”, that is, schools “must have shared goals and feel responsible for success”. In addition, there must be a culture of “collegiality”, “risk taking”, “mutual respect and support” and “openness” (Creemers and Reezigt 2005, 363).

2 The framework incorporates knowledge and elements of a framework exploring the functions of evaluation in democratic governance (Hanberger 2011, 2013) and three models of decentralised governance (Hanberger 2009).

3 The classification of democracy is based on three models of democracy: elitist, participatory, and deliberative or discursive democracy (cf. Dryzek 1996; Hanberger 2006).

4 For the purpose of this issue, this model is confined to the local level of government but can also be applied to the regional level of government (Hanberger 2009).

5 Individual citizens can appeal local government decisions if they think they violate the local government mandate.

6 SALSA stands for Skolverkets Arbetsverktyg för Lokala SambandsAnalyser (NAE's tool for local correlation analysis) and SIRIS stands for Skolverkets Internetbaserade Resultat- och kvalitets Informations System (NAE's Internet-based results and quality information system).