Academic freedom in Scandinavia: has the Nordic model survived?

ABSTRACT

The Scandinavian states have international reputations for promoting social democratic ideals, which have long been manifest in Nordic universities, e.g. legal protection for academic freedom and university studies free of charge. However, Nordic governments have made new h.e. laws, thereby changing university autonomy and management structures, leading to greater involvement by external personnel in university governance, and reduced academic freedom for academic staff. Utilizing legal data, and a survey of+5,000 academics from the EU states, this paper compares the protection for academic freedom in Denmark, Finland, Norway and Sweden. The de jure analysis shows that a single Nordic grouping is not apparent, but that there are two distinct Scandinavian cohorts. The de facto analysis reveals differences between the Nordic nations and other EU states but does neither confirm the Nordic model nor substantiate two regional cohorts identified in previous research and in the de jure analysis.

KEYWORDS:

• Academic freedom
• comparative analysis
• Scandinavia

Introduction

The fortunes of the Nordic/Scandinavian states have long been intertwined. From 1397 to 1523, the kingdoms of Denmark, Sweden (then including Finland), and Norway were joined in the Kalmar Union under a single monarchy (Gustafsson, 2006). From 1249 until 1809 Sweden controlled Finland, while the Norwegian and Danish kingdoms were united for 434 years until 1814. However, the Nordic socio-economic model emerged much later, in the early 1900s, during which Finland gained independence from Russia in 1917, and the Norwegian and Swedish union was dissolved in 1905.

Social, economic and parliamentary reforms, forged in the wake of these moves to independence, created a turbulent political struggle leading to compromise settlements between the two sides of industry, and between the agrarian and workers’ parties (exemplified in Finland by the Agrarian League and the Social Democrats), and which laid the foundation for the Nordic socio-economic model. The resultant model, as Dølvik (2014, 9f.) describes, was characterized by the coordinated interplay between the following three pillars: “Prudent economic steering” built on strong states pursuing active, stability oriented macro-economic policies, … “Encompassing welfare states” built on universalistic principles … “Organized working life” built on interaction between legislation and collective agreements. Strengthening these common socio-economic pillars were linguistic similarities (Swedish, Danish and Norwegian are mutually comprehensible, and Finns learn Swedish at school), cultural commonalities (there are Nordic and Scandinavian associations, covering interests such as culture, and activities like folk dancing), and moral sensibilities (the ideals of the Lutheran church, in terms of personal propriety and moral compass).

These socio-economic, linguistic and cultural linkages were given formal recognition in bodies like the Nordic Council, which was founded in 1952, and recent years have seen calls for some form of re-integration via a United Nordic Federation (Wetterberg, 2010). However, Strang’s comprehensive analysis of the workings of the Council revealed that ‘strategic security concerns and economic interests made more binding forms of cooperation, let alone a formalized supranational Nordic federation, impossible’ (Strang, 2021, p. 103). Moreover, over emphasizing the degree of symmetry between the Nordic states fails to take account of the large and on-going differences between them. For example, the corona virus pandemic was handled very differently in the individual Nordic states. Currently, Sweden has recorded 1961 deaths per million of population from Covid-19, the comparable figures for the other states are Finland = 1154, Norway = 776, Denmark = 1242 (Johns Hopkins University Corona Virus Resource Centre, 2022). As Pashakhanlou (2021) relates: ‘Sweden pursued a rather unique strategy in tackling the coronavirus pandemic. It allowed bars, restaurants, schools, and shops to stay open when most Western countries opted for a lockdown’. A cross-national evaluation of the different covid strategies, commissioned by the Nordic Council of Ministers, reported that ‘representatives of Nordic institutions who were interviewed for this study generally felt disappointed with the fact that Nordic cooperation was not considered a tool for coping with the pandemic crisis’ (Creutz et al., 2021, p. 46).

However, along with many other academic analysts, Vartiainen (2014, p. 5) has lauded ‘the remarkable success’ of the Nordic Social-Democratic Model which has ‘demonstrated good results in terms of growth, employment, gender equality, competitiveness, living conditions and egalitarianism when compared to other countries’. For example, the 2018 OCED (2019) PISA study of Student Performance reported higher mean scores for the Nordic states in Mathematics (505), Reading (507) and Science (501), compared with the mean scores for all OECD states (489, 487, 489) and the mean scores for the other EU states (495, 488, 490). Moreover, these societal values have also been manifested in the h.e. sectors of the Nordic nations – for example, free university studies to doctoral level, and legal protection for academic freedom.

In terms of the exercise of governmental power by different parties, there has been some symmetry between the Nordic states. Examination of electoral statistics for the period from the end of the Second World War reveals the dominance of the Social Democratic (or left wing equivalent) party. Hence, in Denmark, during the period 1945–2022, Social Democratic Prime Ministers were in office for 56% of the time, the comparable figures for the other states were Finland = 41%, Norway = 66% and Sweden = 80%. Moreover, some Social Democratic premiers held office for extended periods – in Sweden, Tage Erlander was in office for 23 years. However, cross national party affinities have been limited – Leruth et al. (2020, p. 94) make the point that within the Nordic group, ‘the social democrats do not appear to share a common view on European cooperation’ while ‘conservative parties are far more united on questions related to the EU and European integration’.

However, academic affinities between the Scandinavian universities pre-date the 20^th century Nordic socio-economic model. As Dhondt (2010, p. 27) describes, Uppsala University (Sweden’s first university, founded in 1477) was the model for the Royal Academy of Åbo/Turku (1640, and which became Helsinki University); Copenhagen University (1479); and the Royal Frederick University of Christiana (1811, in the city whose name reverted to Oslo in 1925). Dahllöf (1996, 199f.) identified the following common denominators shared by Nordic universities:

• the dominance of public provision, and the resultant paucity of private higher education institutions;

• the shared belief that higher education and research are public goods of national economic importance and cultural significance;

• a privileged position in national budgets, rendering them largely immune from expenditure cuts;

• sustained expansion in student numbers since the 1960s, at geographically dispersed institutions;

• free instruction and generous financial support for students;

• concentration of research among larger universities and specialized research units;

• increased internationalization of research culture through collaborative links;

• emphasis on lifelong learning, in line with the old Finnish proverb ‘Oppia ikä kaikki’ [All life is learning].

Hence, the Nordic nations, individually and in unison, have established international reputations for promoting social democratic ideals, including gender and LBGT rights, a comprehensive welfare network, and equality of educational opportunities, leading Stokke and Törnquist (2013, p. 22) to note that ‘The notion of a distinct “Scandinavian model” has had a remarkably stable presence in academic and political discourse’. Moreover, Browning (2007, p. 27) makes the point that: ‘central to the Nordic brand have been ideas of Nordic “exceptionalism” - of the Nordics as being different from or better than the norm’.

Furthermore, the intertwined heritages of these nations and their universities, helped to create a Nordic higher education model, in which academic freedom had a specific meaning, significance and centrality to institutions founded on commonly held Humboldtian principles, within social democratic societies. As Hansen et al. (2012, p. 259) state, the Nordic countries are particularly interesting since they display a high degree of similarity on a number of key political and societal dimensions. At the same time they present striking differences in terms of the organisation of the state administration. Similarly, Rinne (2021, p. 161) declaims that ‘we have good reasons for naming … the Nordic university model’ (original emphasis), which he succinctly and accurately describes as:

a model in which the university sector has followed a wider educational and state policy, and has surrendered almost entirely into the hands of the nation state. Even higher learning is referred to as the institution for promoting democracy and equality among citizens in society. Universities are almost entirely publicly funded and there is very little room for private institutions. The institutions are, at least officially, homogenous and equal, and there is no educational market. A centralised administration and state management guarantee the limitations on competition. An important principle is to keep degree-level education free of charge, in the spirit of the Nordic welfare-state model. (Rinne, 2010, p. 78)

The objective of this paper advances the research agenda of Oberg and Wockelberg (2016, p. 58), who wished ‘to see whether the Nordic countries as a group stand out in contrast to other European regions and to investigate the variations within the Nordic region’, with respect to academic freedom. This will be done by addressing three questions. First, the paper will address the important question posed by Fagerlind and Stromqvist (2004), namely: ‘Whatever happened to the Nordic model’? Second, supported by data on the de jure protection for, and de facto realities of, academic freedom, and with reference to Nordic university legislation, the question as to whether universities in the Nordic nations are more likely to value academic freedom, in a manner distinct from the other EU states, will be addressed. This will be undertaken by scoring the legal protection for academic freedom in the EU states and examining the positions of the Nordic states relative to each other and the other EU nations. Following from this, by means of a large-scale survey, the opinions and perceptions of Nordic academic staff will be reviewed and compared with those of academic staff in other EU states. Third, both de jure and de facto data will be used to answer the question as to whether there are two distinct sub-groups of Nordic nations, in respect to academic freedom.

In assessing academic freedom in the Nordic states, UNESCO’s 1997 Recommendation Concerning the Status of Higher Education Teaching Personnel will be used as a bench-mark. All the Scandinavian states are signatories to the Recommendation, which defines academic freedom as: 'the right, without constriction by prescribed doctrine, to freedom of teaching and discussion, freedom in carrying out research and disseminating and publishing the results thereof, freedom to express freely their opinion about the institution or system in which they work, freedom from institutional censorship and freedom to participate in professional or representative bodies' (UNESCO, 1997, p. 30). This definition is considered definitive, such that (for example) the international academic freedom NGO, Scholars at Risk, also uses the Recommendation as its starting point for defining academic freedom (Quinn & Levine, 2014). However, as Nokkala and Bladh (2014) point out in their cross-national study of institutional autonomy and academic freedom in the Nordic context, there are both ‘similarities and differences’ between the constituent nations, and these will be explored in this paper.

There is largely universal acceptance among university academics as to the importance of academic freedom to their daily activities of university teaching and learning. Indeed, its importance can be attested by the fact that calls have been made to include some measure of academic freedom within world university ranking systems (see Quinn et al. (2021)). However, there have been very few attempts to measure the de jure legal protection for academic freedom, or the de facto operation of academic freedom in university departments. Moreover, of such studies that have been undertaken (see, for example, Karran (2009a and Karran et al. (2017)), none has attempted to compare individual or groups of nation states within an EU context. Hence, this paper is the first to undertake an estimation of academic freedom within the Nordic group of countries, when compared with the rest of the EU states.

De Jure data analysis

Previous analyses of the legal protection for academic freedom adopted ‘top down’ approaches. Early work by Karran (2009a), for example, examined compliance with the 1997 UNESCO Recommendation concerning the Status of Higher Educational Personnel, (UNESCO, 1997) while Nokkala and Bladh (2014) utilized EUA data based on opinions of participants at national rectors’ conferences. However, the more recent analysis by Karran et al. (2017), adopted a criterion referenced approach, allowing the derivation of individual scores, thereby enabling individual states to be ranked with each other. Using interval measures, derived from a ‘bottom up’ approach embracing a wide set of criterion referenced parameters, this study enabled the calculation of the average score for a group of nations, like the Nordic states.

To create this measure, legal and constitutional data was collected from government agencies in all the EU and Nordic states, including measures previously used, such as the provision of tenure, but augmented by additional measures, to provide a clearer and richer picture of the de jure protection for academic freedom. For example, previous studies examined the appointment of Rectors, but not their dismissal, and neglected the staff input into the appointment/dismissal procedures for middle managers (Deans and Departmental Heads).

More significantly, this new measurement system includes an additional important dimension – the ratification by EU states of international agreements which support the protection of academic freedom. Hence, the 1953 European Convention on Human Rights (European Court of Human Rights, 2010, p. 11) offers indirect protection for academic freedom, as Article 10 specifies that ‘Everyone has the right to freedom of expression’. Similarly, the 1966 International Covenant on Civil and Political Rights provides ‘the right to hold opinions without interference’ and ‘the right to freedom of expression’ (United Nations [UN], 1983, p. 178). Both these rights are necessary (although not sufficient) for academic freedom.

This ‘bottom up’ measurement method utilizes 37 different legislative elements (shown in Table 1) for each EU nation, to calculate a composite measure of the legal protection for academic freedom out of 100%, by summing the scores for five dimensions, each worth 20%. These dimensions are academic freedom for teaching, academic freedom for research; institutional autonomy; self-governance; academic tenure; and adherence to international agreements. Some of these measures were very detailed; for example, in terms of the state regulation of university autonomy three levels of compliance were distinguished (Karran et al., 2017, p. 218), as follows:

‘No Autonomy: the state has a high level of involvement in regulating universities’ activities. University governing bodies usually require state approval to enact some regulations and make decisions, and the state may have some decisive majority control over university governing bodies and their composition.

Partial Autonomy: university governing bodies may require state approval to enact some regulations and make decisions (or subsequent state confirmation), and the state may have some control over university governing bodies and their composition.

Full Autonomy: university governing bodies are free from state control and enact regulations and make decisions without prior state approval. The state has minimal involvement in regulating universities’ activities, but merely checks compliance with legal requirements’.

Table 1. Measures used for De Jure analysis.

Description	% Weighting
Measure (1): Legal/Constitutional Protection: teaching and research	20
Measure (2): Legal/Constitutional provision: institutional autonomy	4
Composite Measure: Autonomy: internal operation [(3)-(9)]	8
(3) Rector’s Appointment	1
(4) Internal Structures	1
(5) State Funding	1
(6) Commissioned Research	1
(7) Staff Appointments	2
(8) Student Recruitment	1
(9) Degree Accreditation	1
Measure (10): university autonomy: state regulations	4
Measure (11): Private sector constraints on autonomy	4
Measure (12): self-governance: legal provision	2
Composite Measure: Operational self-governance: [(13) - (16)]	12
(13) Existence of Collegial Bodies	1
(14) Composition of Collegial Bodies	2
(15) Senate Composition	3
(16) Strategic Decision Making	6
Composite Measure: Staff powers: appointment/dismissal [(17)-(22)]	6
(17) Dean’s/Departmental Head’s Credentials	1
(18) Appointing: Deans/Departmental Heads	1
(19) Dismissing: Deans/Departmental Heads	1
(20) Rector’s Credentials	1
(21) Rector Appointment	1
(22) Rector Dismissal	1
Composite Measure: Academic tenure protection [(23)-(27)]	20
(23) De Jure protection: contract duration	4
(24) De Facto protection: contract duration	4
(25) H.E. Legislation: contract termination provision	3
(26) Other Legislation: contract termination provision	3
(27) Provision for academic advancement	6
Composite Measure: Academic Freedom: constitutional protection [(28)-(32)]	10
(28) Freedom of speech provision	2
(29) Academic freedom provision	2
(30) Institutional autonomy provision	1
(31) Self-governance provision	1
(32) Robustness of provisions	4
Composite Measure: Academic Freedom: international protection [(33)-(37)]	10
(33) International Covenant on Civil and Political Rights (free speech provision)	1.5
(34) ICCPR Optional Protocol (UN complaints procedure)	1.5
(35) International Covenant on Economic, Social, and Cultural Rights (right to education provision)	1.5
(36) ICESCR Optional Protocol (UN complaints procedure)	1.5
(37) European Convention on Human Rights (free speech provision)	4

To operationalize the study, different national contexts required some minor compromises. Firstly, the data set excluded other non-university providers of tertiary education (e.g. the schools of higher vocational education, ‘Ammattikorkeakoulu’, in Finland). Although their roles are changing, for such providers their primary function is not academic research but vocational training, often in relation to local industries and labour markets and, therefore, academic freedom for research is less important for them. Second-tier tertiary institutions like Ammattikorkeakoulu often have specific objectives and organizational structures, and differ from universities. Secondly, this study excludes private universities. However, using EUROSTAT data, Levy (2012, p. 179) estimated that within the (then) 27 EU nations, the ‘private share is 12.0%’. Hence, the data shown in the tables represents legislation as applied to public universities, only.

The analysis of constitutional and legislative instruments was repeated for all EU states plus Norway, and scores calculated for each dimension, and then totalled. The nations are ranked in accordance with their total scores in Table 2. With respect to the Nordic nations the following points are noteworthy. First, it appears that, despite their shared heritage of social democratic ideals, and long-time policy collaborations, they do not appear as a coherent ‘Nordic’ group at top of the table, distinct and separate from the other EU states. Indeed, the mean value for the overall scores in the Nordic states (47.4%) is below that of the other EU states (53.8%). Secondly, examining the constituent measures, the Nordic average score is greater than the other EU states for institutional autonomy and job security, but lower with respect to international agreements, self governance, and legislative protection for academic freedom. Third, rather than a unified singular grouping, two distinct Nordic cohorts emerge. Halfway up the table, with above average scores, are Norway and Finland – for the sake of brevity, the name Nordic High has been assigned to this group; near the bottom of the table, with below average scores, are Sweden and Denmark (Nordic Low).

Table 2. De Jure Protection for academic freedom.

Country	Total	Academic Freedom	Institutional Autonomy	Self Governance	Job Security	Constitution/International Agreements
Croatia	69.0	20	13	14	4.5	17.5
Spain	66.5	15	8.5	12	11	20
Bulgaria	65.5	15	9	14.5	9.5	17.5
Germany	64.5	17.5	9.25	12.25	8	17.5
Austria	63.5	20	12	9	5	17.5
France	63.0	20	7	6.5	15.5	14.0
Portugal	61.0	10	9	11.5	10.5	20
Slovakia	60.5	20	8.5	12.5	1.5	18
Latvia	60.0	20	10	10.5	3	16.5
Lithuania	59.5	20	11	6	5	17.5
Italy	57.5	10	9	8	11.5	19.0
Norway	56.5	10	12	3.5	18.5	12.5
Greece	55.5	5	4.5	10.5	20	15.5
Finland	55.0	15	15	3	3	19
Poland	54.5	10	9.5	12.5	5	17.5
Romania	53.5	15	8	12.5	5.5	12.5
Cyprus	53.0	10	8	12.5	10	12.5
Ireland	52.5	15	12.5	3	10.5	11.5
Slovenia	52.5	5	8.5	11	10.5	17.5
Czech Republic	51.5	15	8	11	2	15.5
Belgium	49.25	10	8.5	7.5	9.25	14
Luxemburg	47.5	15	9	6	3.5	14
Netherlands	44.0	10	9	5.5	3.5	12.5
Sweden	39.5	5	6.5	3	8.5	16.5
Denmark	38.5	5	9	6.5	5.5	12.5
Hungary	36.0	5	2.5	9	8	11.5
Malta	36.0	0	10.5	6	8.5	11
UK	35.0	5	13.5	0	5.5	11
Estonia	34.0	0	10.5	4.5	1.5	17.5
All: Mean (St Dev)	52.9 (10.3)	11.8 (6.2)	9.3 (2.6)	8.4 (3.9)	7.7 (4.7)	15.5 (2.9)
Nordics	47.4 (9.7)	8.7 (4.8)	10.6 (3.7)	4 (1.7)	8.9 (6.8)	15.1 (3.2)
Non Nordics	53.8 (10.3)	12.3 (6.4)	9.2 (2.4)	9.1 (3.7)	7.5 (4.5)	15.6 (2.9)

The emergence of these two groups is unexpected, given that previous research by Hansen et al. (2012, p. 260) looking at state agencies and administrations, opined that ‘a significant difference between an East Nordic (Finland and Sweden) and a West Nordic (Denmark and Norway) administrative model should be emphasized’. However, looking at the institutional and organizational features of public administration in the Nordic states, the analysis by Oberg and Wockelberg (2016, p. 69) reported that ‘the East-West divide does not result in any differences in terms of managerial autonomy’. Similarly, Holmén’s recent analysis of h.e. autonomy and governance models in Finland and Sweden noted that ‘it has been claimed that the division in West- and East-Nordic models of governance are no longer applicable’ (2012, 4). Hence, in addition to assessing the validity of a Nordic h.e. model, in relation to academic freedom, the analysis will also establish the validity of regional groupings.

The position of these two diverse Nordic groups may be explained by recent legislation affecting the Nordic universities. In Denmark, the University Act 2003 changed universities from state organizations to autonomous self-owned institutions. Additionally, governing boards, of which the majority of members and the chairman were externally appointed, replaced elected university senates. The Board’s responsibilities include setting the university’s priorities, agreeing a development contract with the government, and hiring the Rector to ensure that the university’s budget reflects their priorities. These new policies were condemned by Wright and Ørberg (2008, 52f) who believe that ‘the Danish model combines the worst of both the free trade and the modernising state models of autonomy: universities, their leaders and academics are given freedom in the sense of individual responsibility for their own economic survival, whilst the sector comes under heavy political control. This is called “setting universities free”’. This legislation’s impact was profound, as Departmental Heads could direct academic staff to perform specific research activities; thus, individual academics had the nominal freedom to conduct scientific research, but this liberty was circumscribed by the University Board’s research strategic framework, as specified in the Achievement Contract drawn up with the Ministry.

Similarly, legislative changes in Sweden increased managerial control within universities and participation in university governance by external appointees, and thereby diluted staff involvement in strategic decision-making. The 2002 Higher Education Ordinance specifies that the University Board comprises the chair plus 14 others, including the Vice Chancellor, three representatives of the staff, and three students’ representatives. The remaining nominated University Board members do not need to be internally appointed academic staff, and are increasingly drawn from outside academia, notably the private sector, and appointed by the government for their business experience and expertise, rather than their knowledge of higher education. In consequence, as Musiał (2010, p. 55) points out, ‘the representation of the external stakeholders in Swedish universities has been strengthened to the extent that the academics employed at the institution may have very little say with respect to its management’. Summarizing policy changes in Swedish higher education, Beach (2013, p. 521) refers to ‘a shift away from academic values and areas that were once a major cornerstone of the university, but which have become increasingly marginal when measured in budget turnover’.

Similar radical changes were proposed in Finland and Norway, but subsequently mitigated. Under the 1997 Finnish Universities Act, the Senate was the university’s supreme executive body, comprising representatives from the professors and associate professors, other teaching and research staff and other personnel, and the students (each of which constituted less than half of the total membership). Additionally, up to one-third of the senate members could be external appointees. The Rector was elected by a university electoral college, and chaired the Senate. However, the 2009 Universities Act radically altered these arrangements by establishing the University’s Board as the university’s highest executive organ. The Board has 7 or 9 − 14 members and, as before, must include representatives of the professors, other teaching and research staff and other personnel, and the students. Although academic staff still have representation on the Board, at least 40% of the University Board members (including the chairman) are now drawn from outside the university community. However, unlike Sweden, the external nominees are appointed by the university collegiate body, and not the government. The legislation originally envisaged that university boards would be required to have at least 50% of their members elected from outside the university (including the chair). Following protests from the Finnish Union of University Professors and from law professors at Helsinki University, claiming that mandating an external majority was against the constitution and would affect academic autonomy, the bill was scrutinized by the Finnish Parliament’s Constitutional Law Committee, which concurred that requiring universities to have a majority of external board members would be unconstitutional, and undermine university autonomy, which is guaranteed under Section 123 of the Finnish constitution. Hence the legislation was revised, such that universities’ boards could have more than 50% external appointees, but this was not obligatory. Additionally, under the new legislation the Rector is appointed by the Board, not the university collegium, and the manner of appointing the Rector was also radically altered. Previously, the Election Collegium, which drew members from a wide spectrum of university personnel, chose the Rector. The new act gave the right of appointment to the University Board, and the chairman of the board is not the Rector, but is chosen from the external appointees. Under the new legislation, the Rector may be an external appointment, although most universities chose to continue with the practice of internal appointments – in 2021, only three of the ten Finnish multi-disciplinary universities had externally appointed Rectors. Assessing this legislation’s impact, Rinne et al. (2014, p. 229) concluded that ‘in Finland we are … surviving in the ruins of the university. Finnish university has lost many of its dreams and old thoughts of autonomy, democracy, equality and the university’s national character. The new enterprise culture, market orientation and mechanisms and managers of power have profoundly changed the place of the university and … weakened the crucial speciality of the institution of the university in society’.

In Norway, the 1995 Act on Universities and Colleges, allowed external nominees to sit as advisory members of university boards. Under the law, as Larsen explains, (2010, 325) the Board could have 9, 11, or 13 members, and comprised the Rector (as chair), representatives drawn from the academic staff, non-academic staff, and between 2 and 4 external representatives. This situation remained largely unchanged, until a ‘Quality Reform’ programme for higher education commenced in 2001, and was implemented over subsequent years. Under this programme, the government proposed that universities should alter their status from public enterprises to become independent foundations. This change met with stiff resistance from academia, and a campaign was mounted, with a petition gathering over 4,000 signatures, and a group called ‘Vox Academica’ (see at: http://foreninger.uio.no/voxacademica/), established to spearhead opposition to the proposals. Hence, as Gornitzka and Maassen (2012, p. 122) reveal, ‘The issue of the formal legal status of universities became a symbol of the Norwegian ideas and norms concerning the balance between state, market, university and society’, and because of such resistance, the Minister abandoned these proposals. Similarly, although the ‘Quality Reform’ proposals were designed to promote changes in university governance at institutional level, whether such changes were implemented was left up to the institution. Consequently, under the 2005 Act, the University Board has eleven members and comprises the chairman, three elected from the academic staff, one from the technical and administrative staff, two from the students and four external members. However, with support of at least two thirds of the members, the board may either establish a different composition (provided that the academic, technical and administrative staff, the students and external members are satisfactorily represented, with no group having a majority) or decide to have a majority of external members. The act allows the Rector to be internally elected for four years, with the Board deciding the weighting of votes (academic staff 51–71%, technical and administrative staff 5–25%, students 15–30%). Where the Rector is elected, the Board also appoints a Director, as the head of the university’s administration. Alternatively, the board may decide that the Rector shall be an external appointment for a fixed period. The appointment process must ensure the Rector has academic and administrative legitimacy, and that the opinions of the students and staff are heard. However, Musiał (2010, p. 51) reported that ‘so far, only very few public HEIs have decided to employ a rector on a contract basis’. Hence, Stensaker’s prognosis that: ‘while some indications of a more strategic leadership are emerging at institutional level, it seems that very few institutions are actually undergoing a radical transformation’ (2014, p. 46).

In summary, the analysis of constitutional and legislative protection of academic freedom does not suggest either that the level of protection for academic freedom in the Nordic states is significantly greater than the majority of EU states, or that the Nordic states form a clearly defined and distinct group.

De Facto data analysis

Constitutions and legislative instruments constitute just the bare bones of academic freedom. The practical effectiveness of such instruments is manifest in the day-to-day minutiae of university life, and the departmental culture operating within universities. Hence, in addition to the preceding de jure evaluation, a de facto analysis is needed. The de facto data was gathered via an on-line survey. The academic freedom survey was compiled after a thorough literature review of over 1000 papers on academic freedom revealed that very little empirical work had been undertaken on academic freedom, and that the research that existed was neither up to date nor comprehensive (for example, see Kerr, 1954). The initial survey was tested on 80 academics at the home institution of the one of the authors, and subsequently amended and re-evaluated.

With respect to survey distribution, the following strategy was adopted. A printed letter was mailed to every university Rector in all the EU states, explaining the study’s purpose, indicating the URL to access an online survey, and requesting an email to be sent to all academic staff at their institution, inviting them to complete the survey. Furthermore, Education International, a global union federation of teachers’ trade unions to which most EU higher educational professional associations belong, sent a letter to all their h.e. members, explaining the purpose of the study, indicating the URL for the on-line survey, and asking for an email to be sent to all academic members of their respective organizations, with a request to complete the survey. Finally, via the websites of the largest universities in each EU nation, academic staff were identified, and sent a personal invitation to participate in the survey. Following an appeal from the Forskerforbundet (Norwegian Researchers’ Association), Norwegian academics were also invited to participate. There were 485 respondents from Denmark, 355 from Finland, 593 from Norway, 386 from Sweden, and 3529 from elsewhere in the EU. The project was subjected to ethical scrutiny in the institution of the corresponding author. In addition, as EU research funding was used to gather data, the project conformed with the RESPECT guidelines, a voluntary code of practice covering the conduct of socio-economic research in Europe, (see at https://the-sra.org.uk/common/Uploaded%20files/respect%20code%20of%20ethics.pdf) As the data gathering surveys were distributed via the internet, the project followed the research ethics guidelines for internet research issued by the Norwegian National Committee for Research Ethics in the Social Sciences and the Humanities (see at https://www.forskningsetikk.no/en/guidelines/social-sciences-humanities-law-and-theology/a-guide-to-internet-research-ethics/).

Clearly, the representativeness of a sample determines the generalizability of any findings that result from analysing that sample, and both are linked to the sample size. Indeed, Martínez-Mesa et al. (2014, p. 609) acknowledged its importance, viz. ‘sample size calculation is an essential item … to reduce the probability of error, respect ethical standards, define the logistics of the study and … improve its success rates’. However, in considering sampling frameworks, Bartlett et al. (2001, p. 44) make the point that ‘a simple survey of published manuscripts reveals numerous errors and questionable approaches to sample size selection’. Moreover, a recognized authority in the field of research methods accurately (but unhelpfully) notes that ‘the decision about sample size is not straightforward: it depends on a number of considerations, and there is no one definitive answer’ (Bryman, 2016, p. 183). More usefully, Olejnik (1984, 41f.) identifies four primary factors affecting sample size: statistical criteria of significance; level of statistical power; data analysis procedure; effect size.

In terms of the statistical criteria of significance, for this study hypotheses were tested at a .05 level of significance. The decision to use a 5% (as opposed to a 1% or 10% significance level) is arbitrary but as Cowles and Davis (1982), report, a 5% significance level is invariably used in studies of this kind, and across the social sciences. The level of significance chosen is the probability of a Type I (or Alpha) error (i.e. a false positive, rejecting the null hypothesis when it is true). Second, the level of statistical power, which is the probability that a statistical technique will detect an effect where it exists, and will not find any effect where none exists. Much social research habitually sets the alpha level at .05, and the Beta at 0.2, giving a power of 0.80, which is equivalent to a 5% chance of a Type I error, and a 20% chance of a Type II error (i.e. a false negative). The third factor is the data analysis procedure, the choice of which depends on the research question, the research design and the nature of the variables (e.g. continuous, categorical, etc.) being studied. Hence, by way of an example, Olejnik (1984, p. 42) notes that ‘research studies involving quantitative independent variables, require fewer subjects than studies using qualitative independent variables’. Similarly, Bartlett et al. (2001, 48f.) indicate that the sample size will vary according to the statistics to be used, with categorical data requiring larger samples than continuous data. The fourth factor affecting sample size is effect size. This study uses the chi-squared statistic, for which the effect size can be measured by means of Cramér’s V, for which Kim (2017, p. 154) provides an interpretive table. In this study Cramér’s V is calculated in respect to academic freedom for teaching (Table 5Table 5), research (Table 6), autonomy (Table 7), self governance (Table 8) and tenure (Table 9), and the results summarized in Table 10 According to the calculations by Kim (2017, p. 154), the reported values of Cramér’s V for this study would be considered to be between small (0.05) and medium (0.13), in terms of their effects.

Cohen et al. (2018, p. 205), mention the general availability of ‘tables which, from a mathematical formula, indicates the appropriate size of a random sample for a given number of a wider population’ and refers specifically to those calculated by Krejcie and Morgan (1970, p. 608) and Bartlett et al. (2001, p. 48), which have been widely used. An analysis of higher education staff in each of the EU states by Crosier et al. (2017), utilized personnel data from nationally accredited agencies (e.g. the Bundesministerium für Wissenchaft, Forschung und Wirtschaft in Austria, the Direção-Geral de Estatísticas da Educação e Ciência in Portugal, etc.) to produce accurate, comparable full time equivalent data for higher education teaching and research staff in all the EU states. Using this data, in combination with the tables provided by Krejcie and Morgan (1970) and Bartlett et al. (2001) enables a comparison of the recommended minimum and actual sample sizes encountered in this study, and the results are shown in Table 3. It can be seen that, in three nations, the actual sample was greater than the preferred minimum, while in Finland it was marginally below.

Table 3. Actual and recommended minimum sample sizes in the Nordic states.

	Sample Size In the Nordic States
	Denmark	Finland	Norway	Sweden
Bartlett et al. (2001) Recommended Sample	370	370	370	370
Krejcie and Morgan (1970) Recommended Sample	375	375	370	377
Actual Sample	485	355	593	386

Table 4. What is the level of protection for academic freedom in your country?.

	All Nordics vs EU		Individual Nordic Nations
	% Nordics	% Other EU	% Denmark	% Finland	% Norway	% Sweden
Generally low level of protection:	12.4	16.7	9.5	9.0	14.2	16.4^b
Average level of protection:	37.7	39.6	32.9	38.0	42.6	36.0^a,b
Generally high level of protection:	49.9	43.7	57.6	53.0	43.2	47.6^b,c
All (n=5244)	100(n=1786)	100(n=3458)	100(n=483)	100(n=355)	100(n=570)	100(n=378)

All Nordics/Other EU: χ²=25.56, 2 d.f. [C.V.5%=5.991], p<0.001, Cramér’s V=.07.

Table 5. Academic freedom for teaching has declined in recent years.

		% Nordics		% Other EU		% Denmark	% Finland	% Norway	% Sweden
Strongly agree	4.5		7.4		3.8		3.8	4.5	6.1
Agree	19.5		21.3		20.1		22.2	20.5	14.8
Neither agree nor disagree	36.6		32.7		34.3		31.3	37.6	43.1
Disagree	30.6		29.6		33.5		31.0	29.2	28.6
Strongly disagree	8.7		9.1		8.3		11.7	8.2	7.4
All (n=5090)	100 (n=1703)		100 (n=3387)		100 (n=472)		100 (n=342)	100 (n=511)	100 (n=378)
All Nordics/Other EU states: χ^2=21.72, 4 d.f. [C.V.5%=9.488], p<0.001, Cramér’s V=0.07
Denmark/Finland: No Difference Neither Grouping Confirmed		Denmark/Norway: No Difference East-West Groups Confirmed					Denmark/Sweden: χ2=11.64 p<0.02 East-West Groups Confirmed
Finland/Norway: No Difference Low-High Groups Confirmed		Finland/Sweden: χ^2=17.80 p<0.001 Low-High Groups Confirmed					Norway/Sweden: No Difference Neither Grouping Confirmed

Table 6. Individual academic freedom for research has declined in recent years.

		% Nordics		% Other EU		% Denmark	% Finland	% Norway	% Sweden
Strongly agree	5.9^a		7.6		5.9^a		5.2^a	5.8^a	6.6^a
Agree	27.6^a		23.4		28.4^a,		30.3	30.1	20.8^a
Neither agree nor disagree	32.1^a		32.0^a		30.1^a		28.9^a	32.2^a	37.2^a
Disagree	26.4^a		27.9^a		29.4^a		25.4^a	23.7^a	27.4^a
Strongly disagree	8.0^a		9.1^a		6.1^a		10.2^a	8.2^a	7.9^a
All (n=5091)	100 (n=1709)		100 (n=3382)		100 (n=472)		100 (n=343)	100 (n=515)	100 (n=379)
All Nordics/Other EU states:χ^2=15.52, 4df[[C.V.5%=9.488], p<0.004, Cramér’s V=0.06
Denmark/Finland: No Difference Neither Grouping Confirmed		Denmark/Norway: No Difference East-West Groups Confirmed					Denmark/Sweden: No Difference Low-High Groups Confirmed
Finland/Norway: No Difference Low-High Groups Confirmed		Finland/Sweden: χ^2=12.04 p<0.017 Low-High Groups Confirmed					Norway/Sweden: χ^2=10.25 p<0.034 Both Groupings Confirmed

Table 7. Institutional autonomy has declined in recent years.

		% Nordics		% Other EU		% Denmark	% Finland	% Norway	% Sweden
Strongly agree	10.3^a		11.9^a		16.5^a		11.7^a,	6.8	6.3
Agree	36.7^a		30.1		40.0^a		42.5^a	38.2^a	25.6
Neither agree nor disagree	33.7^a		31.2^a		30.8^a		28.7^a	32.4^a	43.5
Disagree	16.0^a		20.9		12.0^a		15.0^a,	18.1	19.0
Strongly disagree	3.2^a		6.0		0.9^a		2.1^a,	4.5	5.5
All (n=5083)	100 (n=1701)		100 (n=3382)		100 (n=468)		100 (n=341)	100 (n=513)	100 (n=379)
All Nordics/Other EU states:χ^2=50.84, 4df [C.V.5%=9.488], p<0.001, Cramér’s V=0.100
Denmark/Finland: No Difference Neither Grouping Confirmed	Denmark/Norway: χ^2=38.10, p<0.001 Low-High Groups Confirmed						Denmark/Sweden: χ^2=62.66 p<0.001 East-West Groups Confirmed
Finland/Norway: χ^2=12.11 p<0.017 East-West Groups Confirmed	Finland/Sweden: χ^2=39.28, p<0.001 Low-High Groups Confirmed						Norway/Sweden: χ^2=18.56 p<0.001 Both Groupings Confirmed

Table 8. Self governance has declined in my university in recent years.

		% Nordics		% Other EU		% Denmark	% Finland	% Norway	% Sweden
Strongly agree	18.1^a		15.2		24.4^a		16.3	17.4	12.9
Agree	32.6^a		25.3		33.5^a		36.6^a	35.9^a	23.5
Neither agree nor disagree	33.3^a		32.1^a		30.7^a		32.3^a	29.3^a	42.7
Disagree	13.5^a		21.0		10.0^a		11.3^a	15.1^a	17.7
Strongly disagree	2.5^a		6.4		1.5^a		3.5^a	2.3^a	3.2^a
All (n=5087)	100 (n=1711)		100 (n=3376)		100 (n=472)		100 (n=344)	100 (n=516)	100 (n=379)
All Nordics/Other EU states: χ^2=97.26, 4df [C.V.5%=9.488], p<0.001, Cramér’s V=0.14
Denmark/Finland: χ^2=10.72 p<0.03 Both Groupings Confirmed	Denmark/Norway: χ^2=12.36 p<0.015 Low-High Groups Confirmed						Denmark/Sweden: χ^2=41.94 p<0.001 East-West Groups Confirmed
Finland/Norway: No Difference Low-High Groups Confirmed	Finland/Sweden: χ^2=22.12 p<0.001 Low-High Groups Confirmed						Norway/Sweden: χ^2=26.60 p<0.001 Both Groupings Confirmed

Table 9. Tenure for academics has declined in recent years.

		% Nordics		% Other EU		% Denmark	% Finland	% Norway	% Sweden
Strongly agree	16.6^a		25.7		24.6^a		22.7^a	8.5	12.2
Agree	27.6^a		30.2^a		33.7^a		32.4^a	24.6^c	19.9^c
Neither agree nor disagree	29.6^a		22.3		24.2^a		25.1^a	32.4^c	36.6^c
Disagree	21.0^a		16.9		15.7^a		16.6^a	26.4^c	24.4^c
Strongly disagree	5.2^a		4.9^a		1.9^a		3.2^a	8.1^c	6.9^c
All (n=5086)	100 (n=1708)		100 (n=3378)		100 (n=472)		100 (n=344)	100 (n=516)	100 (n=379)
All Nordics/Other EU states: χ^2=77.96, 4df [C.V.5%=9.488], p<0.001, Cramér’s V=0.12
Denmark/Finland: No Difference Neither Grouping Confirmed	Denmark/Norway: χ^2=83.84 p<0.001 Low-High Groups Confirmed						Denmark/Sweden: χ^2=63.06 p<0.001 East-West Groups Confirmed
Finland/Norway: χ^2=54.31 p<0.001 East-West Groups Confirmed	Finland/Sweden: χ^2=40.08 p<0.001 Low-High Groups Confirmed						Norway/Sweden: No Difference Neither Grouping Confirmed

Trying to directly compare the national samples with their populations would be a lengthy and costly process. However, Ramsey and Hewitt (2005, 73f.) provided a methodology for assessing sample representativeness by posing five questions:

• Are the question, population and confidence clear defined?

• Was sound scientific sampling theory used in the design of the sampling plan?

• Were there correct sampling tools used properly?

• Was the sample collection random?

• Are the sample errors within the expected range for the study?

Assessing the process of survey creation, sampling framework and data collection against these parameters attests to the study’s robustness and is indicative that the national samples are representative of their respective populations.

For the Nordic states to be a distinct group, there should be a difference between the answers from survey respondents in all the Nordic states and answers from respondents elsewhere in Europe. Chi squared (χ²) statistics were used to determine if such differences exist, and if they are significant (for a summary see, Onchiri, 2013). As a further check, pair-wise differences were evaluated using post hoc Bonferroni multiple comparison tests (for a summary, see Armstrong, 2014). When the sample sizes, or the dimensions (rows and columns) of χ² tables differ, the χ² statistics cannot determine the relative size of the differences. However, Cramér’s V, which has a maximum value of 1, corrects for different sample and table sizes, thereby enabling such comparisons to be made. χ² tests were calculated and only those significant at the 5% significance level are reported. Values for Cramér’s V were extracted to test whether differences exist between responses from the Nordic states, as a group, and the rest of the EU, and whether differences exist between the nations within the Nordic group. The academic literature suggested two sub-regional groups – West-Nordic (Denmark and Norway) and East Nordic (Sweden and Finland) while the de jure analysis revealed two discrete Nordic cohorts (High and Low). In respect to the substantive (freedom to teach and freedom to research) and supportive (autonomy, self-governance and tenure) elements of academic freedom, six χ² tests between the individual nations were calculated for these five elements to see if the two different Nordic Low (Denmark/Sweden) and Nordic High (Finland/Norway) groupings, identified from the de jure analysis, are reflected in the de facto statistics, and similarly to evaluate the validity of the East and West Nordic groupings.

Table 4 provides a breakdown of responses, on a three point low to high scale, of respondents’ opinions on the level of protection for academic freedom within their institutions. The χ² value reveals a clear difference between the Nordic group and the other EU states, with a greater proportion of academics in the EU reporting lower levels of protection than their Nordic counterparts. More specifically, post-hoc tests show that there is a significant difference between the Nordic group and the rest of the EU for low levels of protection and high levels of protection, but there is no significant difference between the two groups with respect to the average level of protection. Having considered the de facto data on the level of protection for academic freedom, the constituent elements of the concept will now be examined individually, to test for the existence of a Nordic group, and examine the case for Nordic Low (Denmark/Sweden) and Nordic High (Finland/Norway) and Nordic East/West regional groupings.

The concept of academic freedom is normally taken to include two substantive and three supportive elements (Karran, 2009b). The substantive elements are freedom to teach, and to undertake research; the supportive elements are institutional autonomy, self-governance, and employment protection. Table 5 details responses that participants gave to the statement ‘academic freedom for teaching has declined in recent years’. The χ² statistic demonstrates a difference between the Nordic respondents, and those from the other EU states. 24.0% of staff from Nordic states agreed/strongly agreed that academic freedom for teaching had declined, as opposed to 28.7% of staff from other EU states; at the other end of the scale, 39.3% of Nordic staff disagreed/strongly disagreed, with this statement, the comparable figure for the EU being 38.7%. So, staff in Nordic states are less likely than their EU counterparts to consider that their academic freedom for teaching has declined.

With respect to the case for Nordic Low (Denmark/Sweden) and Nordic High (Finland/Norway) regional groupings, where χ² statistics do not report a significant difference between the scores for Denmark and Sweden, or between Finland and Norway, then the case for these groupings is confirmed. Similarly, the groupings are confirmed where the χ² statistics report a difference between Denmark (Low) and Finland (High). Conversely, where χ² statistics report a significant difference, between (say) Denmark (Low) and Sweden (Low) then the existence of such groupings is negated. The same rationale is used to test the validity of Nordic East and Nordic West groupings. Table 5 shows that in two instances both the Low/High grouping and the East/West grouping were confirmed, whereas in four instances they were negated.

Table 6 details the responses that participants gave to the statement ‘academic freedom for research has declined in recent years’. As can be seen, the χ² statistic demonstrates a statistically significance difference between the Nordic respondents, and those from the other EU states. In this instance, 33.5% of staff from Nordic states agreed/strongly agreed that academic freedom for research had declined, as opposed to 31.0% of staff from other EU states; at the other end of the scale, 34.4% of Nordic staff disagreed/strongly disagreed, with this statement, the comparable figure for the EU being 37.0%. So, unlike the case for academic freedom for teaching, Nordic academics are more likely to believe, when compared with academics elsewhere in Europe, that their academic freedom for research has declined. However, with respect to variation within the Nordic nations, four of the six statistical tests support the idea of high and low cohorts, as suggested by the de jure analysis, but similarly, four out of six tests also confirm the Nordic-East and Nordic West Groupings.

The first of the three supportive elements of academic freedom, institutional autonomy, is considered in Table 7, and the χ² statistic detects a statistical difference between the Nordic respondents, and those from the other EU states. In this instance, 47.0% of staff from Nordic states agreed/strongly agreed that institutional autonomy had declined, as opposed to 42.0% of staff from other EU states; at the other end of the scale, 19.2% of Nordic staff disagreed/strongly disagreed, with this statement, the comparable figure for the EU being 26.9%. So, Nordic academics are more likely to believe, when compared with academics elsewhere in Europe, that the autonomy of their universities has declined. However, with respect to variation within the Nordic nations, three of the six χ² tests support the idea of high and low cohorts, as suggested by the de jure analysis; similarly, three of the tests support the Nordic-East/Nordic-West division

Table 8 details participants’ responses with respect to self governance. As with the previous tables, the χ² statistic demonstrates a difference between the Nordic respondents, and those from the other EU states. The table shows that half (50.7%) of staff from Nordic states agreed/strongly agreed with this statement, as opposed to 40.5% of staff from other EU states; at the other end of the scale, 16.0% of Nordic staff disagreed/strongly disagreed, with this statement, the comparable figure for the EU respondents being 27.4%. So, Nordic academics are much more likely to believe, when compared with academics elsewhere in Europe, that self-governance within their universities has declined. However, with respect to variation within the Nordic nations, five out of the six χ² tests support the idea of High and Low cohorts, as suggested by the de jure analysis, while three out of six support the existence of Nordic East and West groups.

The results in relation to the final supportive element of academic freedom, tenure, are considered in Table 9; the χ² statistic demonstrates a difference between the Nordic respondents, and those from the other EU states. The table shows that 44.2% of staff from Nordic states agreed/strongly agreed with that tenure has declined, as opposed to 55.9% of staff from other EU states; at the other end of the scale, 26.2% of Nordic staff disagreed/strongly disagreed, with this statement, the comparable figure for the EU respondents being 21.8%. So, Nordic academics are much less likely to believe, when compared with academics elsewhere in Europe, that tenure within their universities has declined. However, with respect to variation within the Nordic nations, only two out of the six χ² tests support the idea of high and low cohorts, as suggested by the de jure analysis; similarly, only two tests confirm the Nordic East/West groupings.

Conclusion

This paper addresses three questions. First, the important question posed by Fagerlind and Stromqvist (2004), namely: ‘Whatever happened to the Nordic model’? In sum, the de jure data undermines the Nordic model. However, with respect to each of the fundamental facets of academic freedom, χ² tests of the de facto data reveal that statistically significant differences were found between the group of Nordic states and other EU nations. The second question was whether universities in the Nordic nations are more likely to value academic freedom than their EU counterparts. In general, as the responses concerning the level of protection in Table 4 show, academics in Nordic states profess a greater belief that they enjoy a high level of protection for academic freedom, when compared to their EU counterparts. Hence, the de facto data provides some support for the continuance of the Nordic model.

Cramér’s V statistics were also calculated, as it measures of the effect size, with a small effect where = 0.10 ≤ V < 0.30, a medium effect where 0.30 ≤ V < 0.50, and a large effect where V > 0.50. Table 10 below, gives the χ² and Cramér’s V statistics for the tests undertaken between the Nordic group and the rest of Europe. As can be seen, in the first four entries, which refer to the level of protection for, and decline in, academic freedom, along with the substantive elements of academic freedom for teaching and learning, the value of Cramér’s V is very small. However, it is significantly larger in respect to the supportive elements of academic freedom – autonomy, governance and employment protection. The reasons for this probably lie in the fact that, as was discussed above, all the Nordic nations have been subject to legislation in recent years which has increased institutional managerial autonomy, thereby diminishing individual autonomy, as well as altering governance structures, to the detriment of academic participation in decision making, along with (in some nations), a weakening of employment protection. Thus, legislative protection for academic freedom for teaching and research still exists, but its foundations have been inexorably hollowed out.

Table 10. Chi-Square and Cramer'’s V statistics.

	χ^2	Cramér’s V
What do you think is the level of protection for academic freedom in your country?	25.56	0.07
Academic freedom for teaching has declined in recent years	21.72	0.07
Academic freedom for research has declined in recent years	15.52	0.06
Institutional autonomy has declined in recent years	50.84	0.10
Self governance has declined in my university in recent years	97.26	0.14
Employment protection for academic staff has declined in recent years	77.96	0.12

The third question was whether there are two distinct sub-groups of Nordic nations, in respect to academic freedom. However, in respect to the substantive and supportive elements of academic freedom, overall, the statistics are not supportive of the Nordic nations having two cohorts, as the de jure analysis implied, or as previous academic research has suggested. For each academic freedom element, six tests comparing the four nations were undertaken, to support or negate the existence of two cohorts. For academic freedom in teaching, two out of six tests were supportive; for academic freedom in research, autonomy, self-governance, and tenure, the results were 4, 3, 5, and 2 out of six. Hence, in three out of five elements, half or more of the results support rather than negate the hypothesis of two sub-Nordic cohorts.

In toto, the results are ambivalent, rather than conclusive, but are suggestive of a shift away from a tangible and discrete bloc of Nordic states. The de jure results do not support the existence of a separate Nordic model of higher education, while the de facto data suggests that although the Nordic states provide better protection for academic freedom than their EU counterparts, there has been an erosion of the supportive elements of academic freedom, following university reforms that implemented new public management techniques. This finding tends to support the conclusion of Oberg and Wockelberg (2016, p. 78) who argue that ‘While the Nordic region stands out as distinctively autonomous … it cannot be said to represent a “model”’. Furthermore, there appears to no hard systematic evidence for de facto sub-regional groupings, either of Nordic East and West, or Low and High.

The reasons for the apparent decline of the Nordic model are not unique to its constituent states and their actions, but can be found in national, European and international trends, all of which have acted to increase the degree of homogeneity of national h.e. systems and individual universities. For example, Huisman et al. (2007, p. 574) point to (inter alia) the impact of the Bologna Process on institutional diversity in h.e., and report that ‘higher education systems show an inherent drive towards homogenization, that is less diversity’. All of the Nordic states signed up to the Bologna Process in 1999, and have been strong advocates since. Similarly, Zapp et al. (2021, p. 538) demonstrate how ‘fundamental changes in the global environment of higher education contribute to blur organizational forms and sectors across national borders’, while the metrics of global university rankings act to narrow the parameters to which universities pay the strongest attention. Hence, the situation regarding a decline in academic freedom in the Nordic states is not aberrant, but much in line with the rest of Europe. As Nielsen and Birch Andreasen (2015, p. 94) point out: ‘The recent reform agenda for management and financing of higher education institutions is not unique for Scandinavia. Similar patterns seem to appear in many countries worldwide’. A large-scale study of academic freedom within universities in Europe found that when staff were asked how academic freedom had changed in recent years, 44.7% thought it had diminished, 24% thought that it was unchanged, while only 5.6% thought it had increased (Karran & Beiter, 2020, p. 132).

However, there are facets of the Nordic states and their changing relationships, which may have also acted to diminish their desire to protect academic freedom and further increase their policy divergence. As was reported previously, national legislation in Denmark (2003 University Act), Finland (2009 Universities Act), Norway (1995 Universities and Colleges Act), and Sweden (2002 H.E Ordinance) radically altered the systems of university governance in the Nordic states. Such developments led Nature (2016, p. 315), the globally pre-eminent academic journal for the life sciences, to lament in an editorial that ‘the trend of turning universities into businesses is limiting research freedoms in traditionally liberal Scandinavian institutes’. Highlighting two high profile managerial abuses of academic freedom (Hans Thybo in Denmark and Paolo Macchiarini in Sweden), the editorial proclaimed that ‘Corporate identity might work for a university as a marketing concept – but it offers little incentive for independent minds to speak out and make conclusions’.

Additionally, the manner in which the Nordic states have addressed the process of Europeanization and international collaboration has acted to weaken Nordic ties. Indeed, Olesen and Strang (2016, p. 41) argue that ‘the main challenge for Nordic cooperation … is that the Nordic states are integrated differently into the EU’. In point of fact, the Nordic nations’ engagements with the EU project have been unsymmetrical and idiosyncratic, rather than uniform and coordinated. Denmark joined the EU in 1973 and was later followed by Sweden and Finland in 1995, while only Finland has adopted the Euro currency (in 1999). Norway was twice invited to join but declined after referenda that failed by narrow margins, in 1972 and 1994. Similarly, with respect to NATO, Denmark and Norway were founder members in 1949, but Finland and Sweden delayed considering applying for membership until June 2022, following Russia’s invasion of Ukraine. Moreover, while the EU gained power and sovereignty, the Nordic Council ‘never evolved into an arena or an instrument for the coordination of EU policies and establishing joint Nordic agenda on the European level’ (Etzold, 2020, p. 18). The absence of any shared, coherent, Nordic response to growing Europeanization had led Milijan (1977) to describe the Nordic states as ‘reluctant Europeans’, while Grøn et al. (2015) ask whether the Nordic states are ‘Still the “other” European Community’? However, assessing the consequences of such trends, Browning (2007, p. 40) argues that ‘Nordic internationalism is increasingly melding into a more general European profile to the extent that specific “Nordic positions” are no longer clearly identifiable’.

Looking ahead at the prospects for Nordic universities in successfully addressing these national and European trends, Pinheiro et al. (2014, p. 236) identify ‘four distinct yet interconnected sets of dilemmas with the potential for disrupting the dynamics of individual HEIs and their respective domestic systems’. These are (i) expansion and contraction (i.e. shifts in student numbers); (ii) legitimacy (which focuses on changes to the ‘social pact’ between HE and society); (iii) the erosion of authority (via changes to governance and national steering); and (iv) pressures on the public purse (manifested by changing macro economic circumstances and other demands on government spending). How these will be addressed will vary considerably between the Nordic nations. As well as having dissimilar national socio-economic circumstances, the governments and individual political parties within the Nordic nations have different perceptions and understandings of the role and functions of universities at local, national, and European levels, and different ways of pursuing the necessary political negotiations to successfully confront these dilemmas.

Whether these dilemmas will be detrimental to the Nordic higher education model (with its emphasis on academic freedom), remains to be seen – as the famous Danish physicist, Niels Bohr, once humourously remarked: ‘Prediction is very difficult, especially about the future’. A Nobel Laureate, Bohr was one of the most gifted scientists ever, and when the Scandinavian countries formed the Nordic Institute for Theoretical Physics in 1957, Bohr was its first chairman. As well as research in connection with the ‘Bohr Model’ of atomic structure, for which he won the Nobel Physics Prize, he is renowned for the Bohr Radius, the Copenhagen Interpretation, and the discovery of Hafnium.

Under the Danish 2003 Universities Act: The Head of Department may allocate specific jobs to specific employees. Members of the academic staff are free to conduct research within the strategic framework laid down by the university for its research activities to the extent they are not requested to address jobs allocated to them by the Head of Department. It is difficult to imagine how a talent as eclectic and prodigious as Bohr’s could flourish in such a managerial environment. Indeed, Wright (2014, p. 309) has chronicled how, following legislative changes, Danish university staff were ‘marked by very high levels of stress. … [a]mong … . members of the department … . One had collapsed lifeless in the corridor at work, … Two had experienced the same kind of collapse at home and described how they suddenly could not function at all – they could not read, mark exam papers, write reports, or do anything at all’. Albert Einstein, Bohr’s great friend and colleague, in 1933, in a speech on academic freedom on behalf of the Academic Assistance Council (established to assist academics fleeing Nazi persecution) declaimed: ‘Without such freedom there would have been no Shakespeare, no Goethe, no Newton, no Faraday, no Pasteur, and no Lister’. And, Einstein might have added, ‘no Bohr either’. Hopefully, despite the apparent decline in the Nordic higher education model, as revealed in this paper, academics in the Scandinavian states will still have sufficient academic freedom to continue to make important scientific discoveries, as did Bohr. However, the rise across Scandinavia of populist political parties such as the Danish People’s Party, and the Sweden Democrats, with their emphasis on ethnonationalism seems likely to further increase the divergence between the Scandinavian states and thereby weaken, rather than sustain, the once strong Nordic educational model.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported EU Marie Skłodowska-Curie Actions under grant PIEF-GA- 2012-327101.