The decollegialization of higher education institutions in Sweden

ABSTRACT This article surveys current management ideals of higher education institutions, and our analytical focus is on the balance between line management and faculty self-governance. It presents an empirical study of evolving governance structures including all 31 public sector higher education institutions in Sweden. The point of departure is the Autonomy Reform of 2011, which resulted in a deregulation of the Higher Education Act, and a loss of constitutional support for collegial bodies. To assess the consequences of the reform, we have examined collegial bodies and academic leadership posts before the reform (2010) and after (2020). Our findings show escalating line management in the appointment of academic leaders, a diluted role for collegial expertise, and a loss of decision-making authority for collegial bodies. What we observe is the decollegialization of higher education institutions. Our study contributes to the existing literature with an unusually comprehensive and fine-grained analysis of the consequences of new managerial ideals at the local institutional level.


Introduction
There is ample evidence that public services today are dominated by administrative ideals derived from New Public Management (NPM) (Ahlbäck Öberg & Widmalm, 2016;Hood & Dixon, 2015;Pollitt & Bouckaert, 2017;Pollitt, 1990;Waring, 2017). This is also true for public sector higher education institutions (HEIs). Changes in management ideals can, moreover, greatly affect the local decision-making structures at HEIs concerning academic core activities such as teaching and research, and the recruitment of academic staff. However, thorough empirical studies of how the introduction of NPM ideals affects the internal governance structure at HEIs are still lacking. Through an analysis of the effects of the Swedish Autonomy Reform of 2011, this study contributes to the existing literature with an unusually comprehensive and fine-grained analysis of the consequences of changing management ideals at the institutional level.
The empirical study focuses on public sector HEIs in Sweden, where the general trend in public administration since the 1980s has been towards increasing managerialism through the introduction of different types of NPM reforms. The kinds of managerialism implemented include management by objectives and results, recurrent external evaluation, and monitoring schemes (Ahlbäck Öberg & Widmalm, 2016;Chong et al., 2018;Holmén, 2022;Hood, 1991;Kok et al., 2010;Pinheiro et al., 2019;Pollitt & Bouckaert, 2017;Wedlin & Pallas, 2017). These management ideas have made significant inroads at universities, creating new conditions for the balance between line management and faculty self-governance. Different reforms targeting the higher education system have been added to this, and the Autonomy Reform of 2011 is especially important in this regard (Government Bill 2009/10:149). The purpose of the reform was ostensibly to give universities and university colleges more power to regulate their own internal governance structure. In point of fact, it meant a deregulation of the organizational manifestation of faculty selfgovernance (faculty boards and special bodies). This deregulation was not the starting point for the greater emphasis on managerialist ideaswhich followed from the NPM reforms cited above -but the Autonomy Reform gave greater legitimacy to them, and compounded their effects. The issue at stake, then, is the relationship between collegiality (faculty self-governance) and management ideas influenced by the private sector. The purpose of our study is to provide an account of how the balance between two governance structures -faculty self-governance and line management -has changed as a result of the Autonomy Reform in 2011. The account focuses on decisionmaking powers related to the content and quality of the core activities of an HEI. In contrast to other studies of effects of changes in management models, we base our analysis on formal documents on how the HEIs' decision-making processes are organized rather than survey data, which offer perceptions of steering practices. Our analysis indicates a distinct shift towards empowering line-managers such as Vice-Chancellors, Deans and Heads of Department, which is replacing collegial decisionmaking based on specialized expertise. This also means an increasingly top-down model of university governance. We use the term decollegialization to capture this shift, that is, the centralization of powers to line managers in combination with the weakening of decentral collegial decision-making bodies.
Most HEIs in Sweden are public agencies subject to regulated legal frameworks for governance, control, and organization. The legal status of public sector HEIs requires a balance between faculty selfgovernance and line management if HEIs are to maintain their specific character and mission. An increasing emphasis on line management which might follow from the Autonomy Reform of 2011 (since collegial bodies were deregulated from the legal framework) may challenge fundamental academic principles insofar as academic norms and grounds for decision-making are replaced by managerial ones (cf. Kok et al., 2010). Since HEIs first and foremost exist for the production, preservation, and dissemination of knowledge, expertise, and understanding, the shift in the grounds for decision-making might well undermine the alleged aims of the reform: ensuring quality in research and education. And more importantly, universities have traditionally been considered as types of organization that demands considerable institutional autonomy (see, e.g. Bleiklie et al., 2017;Kok et al., 2010). If public sector HEIs lose the right of faculty self-governance when it comes to academic core-activities, then they will become like other public sector-organizations without guardrails protecting them from political and/or market steering.
This article is organized as follows. The next section presents different university governance ideals and the components of these competing models. These ideals provide us with analytical tools not only to distinguish the similarities and differences between different governance structures but also to review whether such structures are on a par with the HEIs' legitimate demands of farreaching independence from government and other external forces. Thereafter follows an empirical analysis of the development of governance structures within all public sector universities and university colleges in Sweden. Finally, the article presents its conclusions.

University governance ideals
The governance structure of universities and university colleges (henceforth universities) has interested scholars and government officials for several decades, especially the balance between line management and faculty self-governance. In the literature, the latter type of governance is referred to as collegiality or shared governance. Collegiality, as a form of governance, is based on a continuously developing conception of what constitutes good scientific practice and how this practice should shape research, education, and collaboration (Sahlin & Eriksson-Zetterquist, 2016). According to Bennett (1998), the basis for collegiality is a professional community upheld through active collegial modes of work, such as peer review and seminars. Hence, professional arguments should form the foundation for decisions concerning the development and quality of research and education. In relation to the line management structure, the collegial governance structure can be seen as a way of separating powers. The task of top down line management is to assure that the institutions follow regulations set down by government on how to handle public finances, employment, and other centrally decided goals. In relation to the former, the collegial decision-making system expresses an attempt to structurally uphold academic freedom by allowing neither political or other ideological organizations nor private business to dictate decisions regarding the core activities of universities. Historically, this separation of powers has been taken for granted, and sometimes even established in legislation (Ahlbäck Öberg & Sundberg, 2016;Birnbaum, 2004;Deem, 1998;Tapper & Palfreyman, 2010;SOU, 2015).
A study of collegiality within the context of higher education is motivated by the fact that this is where it has reached its most developed expression, to the point that it has been widely perceived as intrinsic to the very idea of a university (Rider et al., 2014;Tapper & Palfreyman, 2010;SOU, 2015). It should be noted, however, that the collegial model is not limited to HEIs but can also be found in other types of professional settings. In his book Lazega (2001), Lazega describes a law firm's governance structure, rewards, control systems, workflows, roles, and status system. As partnerships, law firms embrace ideals of collegiality and equality that obtain in structures of collective responsibility. Lazega's analysis is fundamentally structural, and the point of departure is that organizations performing non-routine, innovative, and often knowledge-intensive tasks need a rather flat, collegial, and non-bureaucratic structure. Complex tasks that cannot be routinized define professional and knowledge-intensive work, and when such tasks are carried out by collective actors (such as professional committees or workgroups), the cooperation and mutual adjustments by these actors cannot be accounted for by models such as Weberian bureaucracy. There is also a division of power-theme in Lazega's analysis: 'Collegiality serves both to deny an authority system and to limit its strength'. (Lazega, 2001, p. 269) Moreover, it should be noted that Lazega is not only interested in organizations that are fully characterized by collegiality, but he also points out the existence of 'pockets of collegiality' that can be found in bureaucratic organizations (cf. the discussion of 'islands of collegiality' in Sahlin & Eriksson-Zetterquist, 2016).
A contiguous discussion can be found in Evetts (2006), who distinguishes between two different and contrasting forms of professionalism in knowledgebased work: occupational professionalism and organizational professionalism. Occupational professionalism as an approach is controlled by the practitioners themselves, meaning that they are selfregulating, subject mainly to informal collegial control. This self-regulation is based on shared education and training, a strong socialization process, shared work culture and occupational identity, and shared codes of ethics that are monitored and operationalized by professional institutes and associations. This traditional form of professionalism presupposes a delegation of decision-making power to the profession, in order to make use of its professional expertise. From the contrasting perspective of organizational professionalism, professional conduct is governed at a distance and includes a substitution of organizational values, such as bureaucratic, hierarchical, and managerial controls, for professional values (Deem, 1998;Reed, 2002;Ahlbäck Öberg & Bringselius, 2015 refer to this as new managerialism).
These managerial ideals have entered the higher education sector, just as they have within the public sector at large (Agevall & Olofsson, 2019;Ahola et al., 2014;Chong et al., 2018;Ekman et al., 2018;Kok et al., 2010;Locke et al., 2011;Musselin, 2013;Pinheiro et al., 2019;Poutanen et al., 2022;Reed, 2002;Tapper & Palfreyman, 2010;SOU, 2015;Waring, 2017;Wright et al., 2019). Historically, universities have been perceived as communities of scholars, researching and teaching together in collegial ways, and those running universities were regarded as academic leaders rather than as managers or chief executives (Deem, 1998;Reed, 2002). The collegial model originally privileged professors, and later on came to include other academic teachers. In the 1960s, student movements pushed for a democratization of the decision-making structures, and as a result the higher education reform of 1977 implemented aspects of organizational democracy into the pre-existing collegial model, dividing the academic constituency into three groups: academic teachers, other staff and students (see Boberg, 2022;Ruin, 1982; for similar developments in the other Nordic countries, see Kyvik & Ødegård, 1990;Poutanen et al., 2022).
But the expansion of the higher education sector in most Western countries has led to increased demands on HEIs to justify the expenditures of public funds and to demonstrate 'value for money'. In this discussion, ample critique of the collegial system has also surfaced, where faculty involvement in shared governance is accused both of slowing down the decision-making process and of being inept at making difficult decisions when, for example, resources are declining (see, e.g. Ekman et al., 2018;Gonch, 2013;Pearce et al., 2018;Shattock, 2010). Those in support of collegial decision-making, however, agree that faculty involvement might slow down decision-making, but argue that it also assures more thorough discussion and provides the institution with a sense of order and stability (Kerr, 1963). Moreover, defenders of the collegial system also argue that the effectiveness of normative institutions is not measured by efficiency and speed but by reliability and trust (Birnbaum, 2004).
To analyse the balance between the collegial and the line management model in Swedish HEIs, we will adopt a framework building on Lazega's and Evetts's discussions of contrasting structural arrangements in understanding knowledge-intensive organizations (summarized in Table 1). The line management model is characterized by the governance structure that follows from the fact that the studied HEIs formally are public agencies under the government, and here the bureaucratic organization is prominent. But as previously stated, the higher education sector has also been a target for the management-influenced ideas that have reached the public sector in general.
In the main, the managerial model has not replaced, but rather been embedded into, an already existing bureaucratic structure (Björck, 2013;Kok et al., 2010;Sahlin & Eriksson-Zetterquist, 2016). This later development has much in common with Evetts's (2006) concept of organizational professionalism, and also with what other commentators label 'new managerialism'. The collegial model, on the other hand, is characterized by Evetts's concept of occupational professionalism. As this framework shows, the organizational hierarchy of the line management model imposes constraints on collegiality. In the contrasting, collegial, model, professional authority and peer review will characterize the organizational design, which would typically be expressed in a rather flat (decentralized) organization. When analysing the decentralization of the organization, we will furthermore make a distinction between collegial and democratic structures. Even though our main interest rests with collegial decision-making in contrast to more managerial systems, it is important not to confuse collegial forms of decentralization with what is rather democratization efforts. The line management model implies a more hierarchical structure of authority, which in recent decades has been strongly influenced by a belief in generic leadership skills where managers are typically recruited as 'experts in management'. This can be contrasted with managers who have assumed the role after a distinguished career within the profession, whose members they now are supposed to manage (primus/prima inter pares) (e.g. Pollitt, 1990, p. 438). In the collegial model, professional practice will be organized through a high degree of self-governance in matters concerning the core activities of the profession. In the line management model, on the other hand, standardization of work practices will be sought and based on rationallegal forms of decision-making (rule interpretation). How, then, can the presence and balance of these two management models be empirically evaluated? Our analytical focus will be on such decisions that currently lack legal support, and where there, thus, is room for variation between the HEIs. First, when it comes to leadership within HEIs, the expectation would be that the appointment of the highest leadership positions in the collegial model would involve processes that ensure the trust and legitimacy of, in particular, academic faculty. In organizational democracy, this constituency is widened to also include students, teachers without a PhD, and other staff. In the line management model, such anchoring processes are not deemed necessary, but the appointment is done from above. Second, when it comes to collegial and democratic versus hierarchical structures of authority, we argue that this could be observed through the HEI's decision-making organization. For example, a tendency to centralize decision-making could be an indication of a drift towards hierarchical structures of organization, while a tendency to a more decentralized organization could indicate more scope for collegial or democratic authority. A way to investigate this empirically would be to analyse the HEI's rules of procedure (i.e. who has the powers to decide what) in matters where academic expertise is clearly called for (for example curriculum and the recruitment of faculty). Third, it is also of interest to analyse what constitutes a colleague in each setting, which is studied in terms of eligibility to elect representatives to collegial bodies. In modern versions of 'representative collegiality' in collegial boards, the entrance into the profession -that is, a doctoral degree -has sufficed. Here, we find a conflict between collegial, meritocratic modes of governance and more democratically oriented models. In the latter, eligibility is not determined by the level of scientific expertise; rather, the determining factor is organizational belonging (i.e. being an employee or a student). Collegiality and democracy are sometimes conflated when one speaks of 'bottom-up' steering of organizations but must be analytically separated in this context, since the two models are founded on different principles (cf. Sahlin & Eriksson-Zetterquist, 2016;De Boer & Stensaker, 2007; on the democratization of Swedish HEIs, see Boberg, 2022;Ruin, 1982;Kyvik & Ødegård, 1990). These kinds of elective procedures are not applicable in a more managerial organizational model.
To summarize, the collegial governance structure within universities in Sweden has typically embodied (a) a decision-making structure of its own, separated from the top-down steering that follows from being public sector institutions, (b) an appointment of academic leaders from below (typically through elections) to ensure a separation of powers, and (c) the use of peer review for positions, promotion, research funding, and decisions on curricula to ensure decisions based on meritocratic principles and professional expertise. By organizing the analysis around these dimensions, we seek to chisel out the organizational logic dominating within the public sector universities in Sweden today, and how this has changed over the last decade. By systematically analysing all public sector HEIs in Sweden, we will have an unusually rich body of data from which to draw inference.

Methodology and data
A comprehensive study of the Swedish HEI landscape serves as an excellent research laboratory for an analysis of the effects of the implementation of new public management or corporate-inspired management reforms. Sweden is a small country where higher education is predominantly publicly funded, and the inclusion of all public sector HEIs in this study generates an interesting mix of different types of institutions. The focus on public sector HEIs also means that any type of reform activity from the Swedish government within this policy sector will target all included institutions.
Hence, we have examined collegial bodies and (academic) leadership positions before and after the Autonomy Reform (2010 and 2020). For these two points in time, we have carried out a detailed analysis of (a) the procedure for appointing academic leaders, (b) the role of collegial expertise in appointing members of faculty boards or equivalent, and (c) the decision-making mandate of collegial bodies, which in turn is further subdivided into (i) decisions concerning curriculum and (ii) decisions concerning the recruitment of faculty. We are interested in investigating issues that demand a high degree of familiarity with, and competence in, specialized forms of knowledge, such as those represented by collegial bodies. We have analysed all public sector HEIs in Sweden, 31 institutions. 1 In the analysis we distinguish between different types of HEIs as follows 2 : Since the total population of public sector HEIs is included in the study, our objects of analysis consist of a mix of large, mid-sized, and smaller HEIs, and also a mix of universities and technical and professional schools. To reach high reliability in our analyses, we have turned to formal documents on how the HEIs' decision-making processes are organized. Hence, the empirical material used for the analyses comprises the HEIs' procedural rules (arbetsordningar), employment regulations (anställningsordningar), and other documents on decision-making and the delegation of powers (delegationsordningar) 2010 and 2020. In addition, directives for elections and general instructions for collegial bodies were consulted, as well as instructions for the recruitment of academic leaders. We have collected about 2 000 documents, and the nearly 250 documents that serve as the sources for Tables 2-5 are listed in Appendix (supplemental material). Administrators or academic teachers have been consulted to fill in gaps where information was lacking or to sort out ambiguities, but, in the main, our conclusions are based on the above-mentioned public documents.

The research context
Almost all HEIs in Sweden are public sector organizations and therefore regulated as government authorities (myndigheter). It should also be noticed that in contrast to countries with ministerial rule, in Sweden there is an organizational divide between the government ministries and central government agencies (Ahlbäck Öberg & Wockelberg, 2015). This executive dualism consists of two parts: Firstly, since the Swedish Constitution prohibits ministerial control, that is, when individual ministers influence or instruct administrative agencies, the Cabinet of ministers acts and is accountable collectively. Secondly, the Constitution guarantees the independence of the state administration when exercising public authority in particular cases. However, general steering through legislation, ordinances, government budget, appropriation directives, etc., is allowed. In short, all public agencies that are government authorities, including HEIs, follow applicable general rules on governance, control and organization.
But HEIs are also subject to the Swedish Higher Education Act and Higher Education Ordinance (SFS 1992(SFS :1434(SFS and 1993 ). This legal framework has been subject to a number of reforms over the decades, which have opened up collegial structures dominated by faculty to greater influence from students and representatives of society at large (Ahlbäck Öberg & Sundberg, 2017;Holmén, 2022;Pinheiro et al., 2019;Sahlin & Eriksson-Zetterquist, 2016). The effects of these reforms will be presented below, where we will primarily focus on the transformation of HEI management structures and governance systems in Sweden, utilizing the analytical categories presented in the previous section. Although the institutional set-up of public sector HEIs in Sweden is similar to that of other government agencies, the managerial structure is distinct. Recognizing that universities and university colleges are not merely government agencies, a dual internal governance structure has historically been applied in acknowledgement that the core activities of HEIs should not be steered by politicians, bureaucrats, or administrators (cf. UNESCO, 1997).
In other words, the internal governance structure of public sector HEIs has divided authority between the top-down public management that follows from formally being a central government agency, and collegial boards that were formerly legally prescribed to ensure collegial decisionmaking in matters regarding the content and quality of research and education. This division of powers in the internal governance structure can be seen as a guard rail for academic freedom, and the deregulation of the faculty boards, as collegial decision-making bodies, is the very focus of our investigation. Thus, the Autonomy Reform abolished the regulation requiring there be a faculty board at every institution, resulting in a debate on whether collegiality is at risk of elimination at Swedish universities (Pinheiro et al., 2019, chapter 3).
The Higher Education Act provides a framework for the organization and governance of the HEIs, and states that every HEI must have a Board of Governors and a Vice-Chancellor (rektor). Boards at Swedish HEIs are composed of two internal groups (student representatives and faculty-appointed members) and an external group whose members are appointed by the Government (Edlund & Sahlin, 2022). Since 1988 the latter group holds the majority, and since 1998 one of the external members holds the chair (until then the Vice-Chancellor was chairperson of the Board) (Ahlbäck Öberg & Sundberg, 2017;Engwall, 2017). The Government also appoints Vice-Chancellors at Swedish HEIs based on proposals from their respective boards. Before the Boards submit their proposal to the Government, faculty, other staff and students are consulted in the manner decided by the boards (Ahlbäck Öberg & Sundberg, 2017). The Pro-Vice-Chancellor (prorektor) of the HEIs are, however, appointed by the HEIs, and so are other top-management positions within the HEIs. 7 Swedish HEIs vary in size and character, and so does their internal organization. 8 At the universities, there are three basic levels of decision-making: The Board of Governors, the faculties and the departments. At the central level, there is a division of power between the Vice-Chancellor and the Board of Governors; at the faculty level, between Deans and faculty boards; at the department level, a few universities have department boards (UU, LU, SU, and LiU), and in those cases there is a division of power between the Head of Department and the board, whereas at other HEIs there is only a Head of Department. 9 At the smaller HEIs, the university colleges and art colleges, there is typically only one central collegial body, equivalent to a faculty board (below the Board of Governors), and the departments (or equivalent units such as schools or academies etc.) are led by a Head of Department. In Sweden, the traditional fora for collegial decision-making have been faculty boards and to some extent department boards (Boberg, 2022).

The autonomy reform of 2011
We have stressed the tension between decisionmaking and governance in a collegial model, on the one hand, and in a line management model, on the other, in particular with regard to fundamental aims and the means employed to achieve these. The managerialist school has great faith in strong leadership to increase the efficiency and quality of public organizations. As mentioned earlier this has not meant that the bureaucratic model has been replaced, but rather that contemporary management ideas have been embedded in the already existing bureaucratic structures. The resulting organizational form has entailed an incremental process of decollegialization of HEIs (see e.g. Ahlbäck Öberg & Sundberg, 2016;Engwall, 2014Engwall, , 2017. This is not a recent development, but one that has evolved over four decades. But the government's Autonomy Reform of 2011 fortified and legitimized this development, since the deregulation primarily targeted the collegial bodies at HEIs, which lost their constitutional base in the Higher Education Act (Government Bill 2009/10:149).
Before the reform, the Higher Education Act required and guaranteed that every HEI had boards elected by and representing the faculty; these boards were responsible for decisions concerning research.
The faculty boards also had jurisdiction over the form and content of instruction at both the undergraduate and graduate levels. The Autonomy Reform gave the HEIs as such more power to make decisions about their internal organization, rules of employment, career advancement, and so forth. This increased the authority at the level of central management, at the expense of the decision-making powers of collegial bodies. Furthermore, the protocol for the election of academic leaders was deregulated. Consequently, as the constitutional support for collegial decisionmaking bodies was abolished, the powers of upper management (the Vice-Chancellor and the Board of Governors) were enhanced. It should be noted that while the reform does not authorize or require the existence of collegial bodies, nor does it prohibit such arrangements. It is the prerogative of the institution, which is to say, management, to decide. Below, we analyse the organizational response from Swedish HEIs to the Autonomy Reform, i.e. if there has been a propensity towards more or less collegiality.

Findings
In this section we present our empirical findings, which are subdivided into three subsections: (a) the procedure for appointing academic leaders, (b) the role of collegial expertise in appointing members of faculty boards or equivalent, and (c) the decisionmaking mandate of collegial bodies, which in turn is further subdivided into (i) decisions concerning curriculum and (ii) decisions concerning the recruitment of faculty.

Appointment of academic leaders: increased line management
The appointment of academic leaders elected by their peers has been a keystone of the collegial system (cf. UNESCO, 1997). Such an appointment process is thought to have the advantage of selecting leaders who understand academic work and subscribe to shared academic norms, which gives them legitimacy within the collegium (Goodall, 2009). As a result of the Autonomy Reform, there has been various degrees of displacement from earlier arrangements, when elections to collegial bodies are held. Prior to the reform, Deans were formally appointed by the Vice-Chancellor, but elected by the faculty (SFS 1993:100, Ch. 3, 4 §). As of today, the direct involvement of the Vice-Chancellor and other managers in the line organization has become the norm (see Table 2). We provide a full analysis below.
With regard to the process that leads up to this appointment, however, there is today great diversity among Swedish HEIs. Although election procedures are no longer regulated in law, about half of HEIs still practice elections. There are also clear differences between the different categories of institutions. At old universities and specialized HEIs, elections are still the default position. Among younger institutionsyoung universities and university colleges -elections are still frequent, but now occur in a minority of institutions. Finally, among the art colleges no elections are carried out.
However, even at the institutions that still hold elections, the direct involvement of the Vice-Chancellor in the process has become the norm. Across the board, we see that nomination committees are now required to consult the Vice-Chancellor before suggesting candidates. This is an expression of the managerial idea that subordinate leaders in the line management structure must share a vision similar to that of their superior -the Vice-Chancellorand be able to cooperate to fulfill the vision of HEI leadership.
Another important trend that cuts through the whole HEI system is a democratization of the electoral process. Earlier, only academic faculty with research competence had the right to vote. Today, this is only the case at UU and SLU. At all other institutions where elections are held, we see a replacement of the collegial principle of competence with a principle of democratic participation. Most commonly, the right to vote has been extended to teaching faculty without a PhD (junior lecturers). In a few cases, the right to vote has even been extended to all employees at the faculty-level, which represents a complete replacement of collegiality with organizational democracy.
Among half of the Swedish HEIs, elections have been replaced by other appointment procedures. Most commonly, the collegial election has been abolished, and nomination committees or recruitment groups have been given a more prominent role by proposing a candidate that is directly appointed by the Vice-Chancellor without any intermediary collegial election process. At a few HEIs, the Vice-Chancellor selects the chair of the board from among its members (after consulting the board) or among the professors at the faculty.
In a third group of institutions, the role of top management is pronounced in yet another way. At KI, KF, and KKH, the Vice-Chancellor or Pro-Vice-Chancellor himself or herself is the chair of faculty boards. Since 2019, KI has one faculty board, headed by the Vice-Chancellor (KI, 2019). This situation is exceptional in the Swedish case, and the closest parallel is the situation at the small art colleges KF and KKH, where the faculty boards are also chaired either by the Vice-Chancellor or the Pro-Vice-Chancellor. Because of the very limited size of these latter institutions, the situation is not, however, comparable to that at KI.
To sum up, our analysis shows a strengthening of line management's top-down control of the appointment of academic leaders in most of the HEIs, and a deprofessionalization of the eligibility to vote. The election of academic leaders by the faculty has in certain cases been revoked, while in others it has been replaced by considerably weaker collegial kinds of influence, which represent a displacement of decision-making authority from the collegium, either in the direction of organizational democracy, or to a manager in the line organization (often the Vice-Chancellor), or a combination of both.

The role of collegial expertise: from faculty members to employees
Another major change concerns the voting rules for elections to collegial bodies. The concept of the colleague is central here. A colleague in the context of collegial decision-making refers to faculty members, not to all co-workers at an institution. The principle of faculty self-governance and the idea of academic freedom are built upon the notion that the academic professions have specific professional qualifications: the requisite expertise to decide on issues concerning content and quality in higher education and research. The Higher Education Act of 2010 stipulated that only persons with relevant competence had the right to vote for candidates to the faculty boards. At the same time, students and other employees were guaranteed seats on the board through representatives chosen by student and trade unions. In Table 3, we show changes in rules for the faculty boards (or the equivalent).
The new voting rules indicate that in more than a fifth of the cases a democratic principle of employee influence has entirely replaced the principle of collegial decision-making based on expertise. Hence, at six of our studied HEIs all employees have been given the right to vote. At another 15 institutions the right to vote has not been extended to all employees, but to junior lecturers without a PhD. So far, this major change exhibits the decline of faculty influence on their own professional practice. In sum, the bulk of the HEIs that have extended the right to vote to junior lecturers without a PhD belong to the category university colleges.
Furthermore, it should be added that at five of the institutions studied the right to vote has been given to researchers with a PhD, thereby including them in the same group as academic teachers. This latter movement does not represent a dilution of collegiality, on the contrary. Table 3 also displays that for two of the HEIs studied (KTH and BTH) there are no faculty boards or equivalent collegial bodies, and for another two institutions (GIH and KKH) the appointment to the collegial body is a process led by the Vice-Chancellor. We conclude that especially for KI, KTH, and BTH the situation now for any Dean (or equivalent) is that her or his authority rests on some backing from the university top management and/or is embodied by the top management. We also note that these institutions are within the category of specialized higher education institutions (medicine and technology), a state of affairs that will reappear in coming analyses.

Decision-making authority of collegial bodies: diminished influence
Since the Autonomy Reform, decision-making powers involving academic questions, formerly granted by legislation to collegial bodies, have been delegated directly to the chair of the body (appointed by the Vice-Chancellor) or to a single line manager. Changes involving the curriculum and the recruitment of faculty are presented below.

Decisions about the curriculum
Well-founded decisions regarding which courses and programs are to be offered at undergraduate and graduate level are of central importance for the quality of instruction. The assignment of discretionary powers for the items listed in Table 4 has never been regulated in law, either before or after the Autonomy Reform. But before 2011, since the faculty board explicitly had general responsibility for undergraduate and graduate studies, decisions regarding, for example, course offerings and syllabi, it was natural for them to be made by that body.
The following tables specify the delegation of discretionary powers before and after the Autonomy Reform. We have selected items, such as the revision of syllabi, where expertise is central for the academic outcome. It should be noted that sub-delegations are not specified. Sub-delegations are less relevant to this study, since we are concerned with the principal question of the separation of powers within HEIs. From that perspective, the crucial issue is whether a collegial body or a line manager owns a certain question. Table 4 shows that in just a few years, decisionmaking powers that formerly belonged to the collegial bodies have shifted to line managers or to the top management of the HEIs. In general, the changes have been quicker and more radical at specialized HEIs. For this group the introduction of managerialism is conspicuous. Virtually no faculty governance regarding decisions on curriculum can be found at these institutions. As mentioned earlier, KTH and BTH have abolished their faculty boards altogether. Again, especially the group of HEIs with a technical profile (KTH, LTU, and BTH) display a radical change where decision-making on course and program syllabi has been more centralized and altered to a single line manager or top management. Similarly, the HEI with a medical profile (KI) follows this path of decollegialization, even though a few decisions are left to collegial bodies at the faculty level. Again, this raises questions about the managerial preference among certain disciplinary domains, especially since we do not see the same overly explicit development at the small specialized schools, which are dominated by arts and music. Table 4 also displays less collegial decision-making for younger universities and university colleges, even though the shift is not as radical as for specialized higher education institutions. It can also be observed that within the category of university colleges there are examples where collegial decision-making has Specialized HEIs

Young universities
University colleges been replaced with managerialism, that is, decisionmaking powers have been shifted to line managers. In these cases, there are clearly fewer decision-making powers regarding the curriculum left to faculty organs after the Autonomy Reform, which make them cases of top-down management. Moreover, in many instances where a collegial body still is in place it should be noted that it has become centralized (e.g. at LnU, HiG and HH). Centralized collegial bodies are a pattern found in this category of HEIs even before the Autonomy Reform. The specialization that decentralized collegial bodies normally hold is

Specialized HEIs
Young universities challenged if there is only one centralized collegial body to take responsibility for a wide variety of disciplines. The idea of professional expertise encompasses an idea of specialization, that a colleague might be able to make qualitative assessments of neighbouring disciplines. However, the professional expertise of teaching and research in one domain cannot easily be extended to all other domains of a university or university college. Similar centralization can be observed at the art colleges, but this state of affairs is less challenging to the idea of professional expertise, since the scope of their disciplinary domain is considerably narrower.
Finally, the major research universities have maintained collegial decision-making even after the Autonomy Reform. However, a deviant case in this group is GU, which has replaced decision-making powers for collegial bodies with that of a single manager in the line organization.
We see a similar development in postgraduate study programs. Table 5 illustrates the delegation of decision-making authority at the postgraduate level after the implementation of the Autonomy Reform. Before the reform, all of the selected items were assigned to the faculty boards through the Higher Education Ordinance, that is, all of the cells would have been shaded. In 2020, the picture is quite different.
In Table 5, we see that faculty boards (or the equivalent) at the old universities have retained almost all their earlier decision-making authority concerning postgraduate studies. These institutions display the pattern that was compulsory before the Autonomy Reform. An exception among the older universities is GU, one of the largest universities in Sweden, which has abandoned collegial decisionmaking to an unparalleled degree as compared with the aforementioned major universities. For the  younger universities and university colleges the clearest tendency of decollegialization can be found at KaU, LnU, and MaU. For all of these, decisionmaking powers have been shifted from collegial bodies to -in most cases -a Dean (appointed in systems that have strengthened the line management's top-down control). The category of HEIs that stand out also in this respect is the specialized higher education institutions. At the technical universities and university colleges no decision-making powers are left to collegial bodies; instead, these powers are displaced to a single manager in the line organization or the HEI top management. In contrast, SLU has reserved all decision-making about postgraduate studies for collegial bodies. Also, this time the art colleges demonstrate more collegial decision-making than the larger specialized higher education institutions. It should be noted, however, that in this latter category, for the arts and music colleges the decisions are taken at other HEIs.

KF
On the whole, we discern a pattern similar to the one we saw with respect to undergraduate studies: the majority of the major research universities have retained the collegial decision-making model, while some of the younger universities and university colleges have shifted to the line management model. The radical shift can, again, be found among the technical universities.

Decision-making about the recruitment of faculty
The recruitment process for faculty is of the utmost importance for academic quality. According to collegial principles, the appointment of new faculty members should be preceded by a process of rigorous peer review to ensure that the most qualified applicant is hired. Before the Autonomy Reform, the faculty board or a special committee was responsible for organizing and executing the application process for senior lectureships and professorships, and, on that basis, recommended whom to hire. At the start of the recruitment process, the faculty board decided on the subject area, requisite qualifications, and assessment criteria for these positions. As a rule, faculty boards delegated this work to an academic appointments committee.
For this analysis, we have chosen the recruitment of senior lecturers as our example, since it is less regulated than professorships and thus could display more local variation. For the year 2020, we observe patterns similar to those identified in the previous sections (see Table 6 in Appendix (supplemental material)). The old universities have to a greater extent kept the collegial decision-making process intact. Particularly, this is true for UU, LU, and SU. Likewise, the specialized higher education institutions, which earlier were shown to reveal farreaching decollegialization, also exhibit weakened collegial institutions with respect to recruitment of faculty. At HEIs with a medical or technical profile, centralized academic appointments committees, either directly appointed by the Vice-Chancellor or with a minority of elected faculty, are in charge of the recruitment process. Line managers also have gained a firm grip on initiating recruitment and establishing the academic profile of lectureships. Among the specialized HEIs, the agrarian university is the clear exception. As we have noted earlier, the collegial structure at this institution is similar to that found at the old universities, and it has also retained the recruitment system that was earlier regulated in law.
Individual academic leaders have gained somewhat more extensive decision-making authority at the young universities and university colleges. At a third of the institutions, line managers on different levels appoint the recruitment committees, whereas at others, only a minority of the committee members is appointed by the faculty. Line managers have a strong hold of the recruitment process at these institutions, and generally initiate recruitment, decide profiles, and in many cases even appoint evaluators. Collegial bodies are typically responsible for the process as a whole, and rank candidates based on the evaluators' reports. At the university colleges, however, these collegial bodies are centralized and cover several disciplinary domains, which is a challenge for the original idea of collegial expertise.
The art colleges represent a category of their own, quite distinct from the other HEIs investigated here. The Vice-Chancellor plays a more important role at these institutions than anywhere else in the system. The Vice-Chancellor initiates recruitment and decides the profile for lectureships. Furthermore, the Vice-Chancellor appoints recruitment committees or recruitment groups at a majority of the institutions. The bodies that are in charge of the process are centralized, as at the university colleges, but since these art colleges are heavily specialized, this is not as detrimental to the collegial model. It is safe to say that the character of the system of recruitment reflects the size of these institutions -they are equal to the size of departments at other institutions in the Swedish system.
All in all, this analysis suggests a distinct shift to a more top-down mode of governance. In terms of recruitment processes, there is a tendency towards centralization of powers to line managers, which is to say, decollegialization.

Conclusions
The analyses in this article indicate a clear transfer of decision-making authority from collegial bodies to individual line managers during the period studied. We have called this a process of decollegialization. The most striking examples of this development can be found in the category of specialized higher education institutions, and HEIs with technological profiles stand out as very highly decollegialized (and we see the same development at BTH, a university college with a technical profile). It is possible that the readiness to adopt business-style managerial models within these institutions can be related to the close relationship that these types of disciplines have to the industry (cf. Larsen et al., 2020). Also, the category of university colleges displays clear evidence of decollegialization, where top-down management has been strengthened, and collegial bodies lack specialization in terms of expertise, since they tend to be centralized (or general) with a span of control encompassing many -if not all -domains of the HEI. When collegial bodies are centralized in the way that is the case today at the majority of the university colleges, the specialization which scientific expertise rests upon for decision-making is put out of play. As a group, the younger universities have to a somewhat higher degree than the university colleges kept collegial decision-making processes, even after the Autonomy Reform. Finally, the old universities have preserved a high degree of collegial decisionmaking. These results are similar to Kok et al.'s (2010) findings in the UK that academic decisions continue to be made by academics at traditional universities, while new universities (starting from 1992 and onwards) are more receptive to a managerialistic agenda. The authors argue that this may be due to younger universities often having roots as polytechnics with bureaucratic and managerialist ties.
How, then, does this process of decollegialization play out in the dimensions studied? First of all, the way academic leaders are appointed within the different institutions indicates a move away from election by peers to appointment by unelected managers, unencumbered by the need for deliberation, justification, or appeal to standards accepted among academic staff members. Central management has gained more control over the selection of academic leaders. The line management model, in turn, relies on a high degree of loyalty from subordinates in relation to those in superior positions by virtue of their position in the hierarchy. This is at odds with the rationality of collegial decision-making, where loyalty is directed towards shared norms and principles rather than hierarchical superiors. Secondly, an evident change in the institutions examined in this article is the blurring of the distinctiveness of the meritocratic collegial system and its replacement with a system of internal representative democracy (cf. Maassen & Olsen, 2007). Here we thus discern a clear confusion between the collegial governance model and the idea of employee democracy. The former is based on an idea of expert rule (exclusivity) while the latter is based on an idea of equality (inclusivity). Thirdly, the results show that the organizational manifestation of academic freedom and functional decision-making on the basis of specific competency through collegial decision-making bodies has diminished considerably. Hence, the autonomy granted by the reform has indeed entailed a higher degree of autonomy, but for top management rather than academic staff. This means a strengthening of the institutional autonomy (cf. the developments studied in Poutanen et al., 2022). The article shows that the faculty's possibility to make decisions on the basis of professional judgement on issues involving the form and substance of research and teaching has been heavily reduced. It is striking that several of the HEIs hardly live up to the less demanding idea of 'islands of collegiality' since those organizations have little or no division of power, and the disciplinary expertise has been given a weaker -and in some cases non-existent -position in the organization. In sum, a national reform based on the explicit idea of decentralization in fact led to centralization at the local level, i.e. centralization within the individual HEIs.
We argue that a sequence of developments made this shift possible. The Autonomy Reform gave the top management of HEIs the opportunity to decide on the HEIs internal organization. Before that, students and faculty on HEI boards of governors were reduced from holding a majority to a minority of seats and votes in 1988, and board leadership was transferred from Vice-Chancellors to external chairpersons in 1998. Hence, the boards are not led by representatives of the academy itself, but by external members who do not necessarily possess in-depth knowledge of conditions for academic work or the collegial model. However, this set of conditions is similar for all HEIs. It does not explain why some institutions move further towards decollegialization than others. Other conditions may explain this divergence within the field, such as the closeness to industry (cf. Larsen et al., 2020), academic culture and traditions (cf. Kok et al., 2010), the proportions of faculty with and without a PhD (cf. Agevall & Olofsson, 2019), the preferences of people in power when a reform was enacted, etc. However, such an analysis is beyond the scope of the present study.
What is at stake? Previously, the Higher Education Act mandated use of the principle of division of power, giving constitutional support to faculty boards. Our study shows that this division of power between top management and collegial bodies has been completely or partially abandoned in several Swedish public sector HEIs. Our point of departure is that knowledge production and knowledge transfer presuppose completely open examination, intellectual independence, and neutrality, not least in relation to the demands and wishes of different stakeholders. Ultimately, securing these bases for research and education is a question of the autonomy and integrity of HEIs, which fundamentally means the autonomy of the faculty at the institutions. Academic staff must be able to exercise scientific judgement independently of political or economic interests. Faculty self-governance could be said to be the organizational expression of academic freedom, and it is noteworthy that contemporary attacks on academic freedom often involve undermining, sidestepping, or completely displacing faculty self-governance. This is a sign of the close relationship between academic freedom and selfgovernance (cf. Gall, 2021).
Are the risks to freedom outweighed by benefits of increased line management? It is clear to us that the management reforms in the higher education sector have been implemented with the intention of achieving higher efficiency, and the policy documents also use increased competition in the sector -nationally and internationally -as an argument for the need to increase efficiency (Amaral & Magalhães, 2002;Bleiklie et al., 2017;SOU, 2015:92. Pinheiro et al., 2019. However, in the long run it is not selfevident that substantial efficiency can be gained through increased line management at the expense of collegial decision-making. There are also reasons to weigh imagined gains against the costs that follow when depriving the academic profession of the opportunity for real influence over the quality and content of teaching, research, and recruitment. Many before us have described and discussed the managerialization of university governance, both for individual countries and in a comparative perspective. Few studies have, however, made a comprehensive analysis of the consequences of increased managerialism at the local level (for an exception, see Kwiek, 2015). Our study of all public sector HEIs in Sweden shows that the very idea of decisions based on expertise (collegiality), for issues such as curriculum and recruitment, in most HEIs is disassembled as soon as the right to decide on the internal organization is decentralized to the institutional level. The main effect of the deregulation of the Higher Education Act in this part was a process of decollegialization. At first glance, this is a paradox. Yet as this study has argued, it can be understood. When university line managers are empowered to organize how decisions are made, such as occurred in Sweden's reform of the HEI sector in 2011, they may give themselves decisionmaking authority and take it away from academic staff, as occurred most notably in Sweden's specialized HEIs in the wake of the reform. Hence, to secure and enhance collegiality in HEIs, academic staff involvement in crucial university decisions needs to be codified in legislation.
Finally, there does not need to be a contradiction between the HEIs' dual roles as public agencies and self-regulating communities of academic teachers and students. But having dual roles necessitates dual governance systems, where the two sides are responsible for different types of issues based on specialization and competency. The groundwork for decisions about the recruitment and promotion of faculty, as well as about research and teaching, should fall within the remit of academic staff. The HEI administration, on the other hand, should take responsibility for those issues belonging to their sphere of competence, such as accounting, human resources, student welfare, public and internal information, and other tasks issuing from the HEI's role as a public organization (Rider et al., 2014). Given that the primary role and core mission of HEIs is to provide higher education and to conduct research, they should fundamentally be collegial organizations, and have line management only where functionally necessary. A university that is mostly governed by line management -without much or any collegial decision-making -would challenge the very idea of a university. Such an organization would be just another central government agency.

Notes
1. In 2020, there were 48 HEIs in Sweden, and the public sector HEIs' share of staff, students and research funds was more than 92% (UKÄ, 2022). Those public sector HEIs that have ceased to exist since 2010 are not included in this analysis. 2. Throughout the analysis we will make the formal distinction between universities (universitet) and university colleges (högskolor) explicit. The term universitet refers to a full research university, while högskola is a teaching organization without any general right to award PhD exams. 3. The old universities were established before the 1977 reform of the Swedish higher education system. 4. Because of its academic profile, we count Luleå University of Technology among the specialized higher education institutions. 5. Malmö University College was granted full university status in 2018 and changed its name to Malmö University. 6. SFS stands for the Swedish Code of Statutes. 7. As an effect of the Autonomy Reform of 2011 the procedures for the appointment of Deans, Vicerectors (or the equivalent) etc. can vary from institution to institution: either by some sort of election within the HEI or just as a top-down appointment. This will be clarified below. 8. There are significant differences between Swedish HEIs when it comes to the number of faculty (fulltime equivalent teachers and researchers) and registered students. In 2020, old universities on average had a faculty of about 2 600 and 42 000 students, new universities had a faculty of 770 and 22 000 students, specialized HEIs had a faculty of 1 600 and 12 000 students, university colleges had a faculty of 330 and 10 000 students, and art colleges had a faculty of 80 and 750 students (UKÄ, 2022). 9. A few universities also have collegial bodies at a level above the faculties (disciplinary domains).