Undisclosed desires: quality managers’ normative notions regarding the implementation of quality management

Abstract

Following decades of quality management featuring in higher education settings, questions regarding its implementation, impact and outcomes remain. Indeed, leaving aside anecdotal case studies and value-laden documentaries of best practice, current research still knows very little about the implementation of quality management in teaching and learning within higher education institutions. Referring to data collected from German higher education institutions in which a quality management department or functional equivalent was present, this article theorises and provides evidence for the supposition that the implementation of quality management follows two implicit logics. Specifically, it tends either towards the logic of appropriateness or, contrastingly, towards the logic of consequentialism. This study’s results also suggest that quality managers’ socialisation is related to these logics and that it influences their views on quality management in teaching and learning.

Keywords:

• Higher education
• quality manager
• teaching and learning
• appropriateness

Introduction

Discussions on quality have long played a role in the many centuries-old higher education system, even if such debates have not always been labelled with the term ‘quality’. More recently, though, the concepts of ‘quality assurance’ and ‘quality management’ have infiltrated numerous sectors and been applied all around the world (Donabedian 1989; Beck and Walgenbach 2005). Consequently, practitioners and researchers in the field of higher education have begun to recognise the importance of this topic (Neave 1988; de Weert 1990; Lindsay 1992; Harvey and Green 1993). Based on these original pieces of research and the ever-increasing interest in management practices more broadly, a lively research community has developed investigating a wide range of issues of quality (Harvey 2018).

The approach of quality is often directly linked to its definitions, and, in the case of the present study, to the very understanding of quality and its implementation in higher education institutions (Cullen et al. 2003; Harvey and Green 1993; Owlia and Aspinwall 1996; Spencer 2001; Westerheijden 2007; Enders and Westerheijden 2014, 174). Recently, quality management—in this article, I use the term ‘quality management’ to refer to all quality-related practices, processes or measures of quality assurance and quality development in teaching and learning—and its impact has become one of the most appealing, but also demanding, topics in the literature. For example, some researchers have endeavoured to conceptualise its impact and analyse it through complex causal chains (Horsburgh 1999; Perellon 2007; Leiber, Stensaker and Harvey 2015). While further empirical evidence is required, it will be simply a matter of time before these research gaps close. Meanwhile, current research points towards a growing interest in understanding the internal processes of quality management and their effects (Leiber 2012; Stensaker and Leiber 2015).

What has widely been ignored in prior research, aside from a few case studies and summarising articles (e.g. Dynan and Clifford 2001; Kernegger and Vettori 2013), is the perspective of people who implement quality management: quality managers. Their complex embeddedness in the higher education system, with various stakeholders, has led to a multifaceted and fluid task profile. It comprises many different activities, mainly relating to accreditation processes, evaluations, internal surveys, reporting and consultation, as well as other activities referring to the internal development of statutes or study programmes within the higher education institution.

Although some of quality managers’ basic characteristics have been investigated (Harvey and Williams 2010), empirical research concerning their perceptions and attitudes is nearly non-existent. One reason for this deficit might be that surveying quality managers could result in well-known biases (Askling 1997), yet this argument does not appear to hold for studies investigating the perceptions of other actors such as academics or managers. Given the existing studies that are available on academics’ views—dealing with, for example, issues of resistance and resentments (Spencer 2001; Anderson 2006, 2008; Cardoso, Rosa, and Stensaker 2016)—analysis of quality managers’ perceptions should be able to follow the same methodological rationale.

The present article addresses this research gap, focusing solely on quality managers’ perceptions of quality management in teaching and learning. This seems reasonable because quality management is a part of the delegation chain of evaluation in higher education institutions. Quality-related activities are developed within higher education institutions in relation to several principals (supranational and national regulators, higher education institutions’ management teams or academics, and other stakeholders). However, within this chain of delegation, quality managers still have room for manoeuvre in which to formulate and implement internal quality policies. Hence, quality managers’ perceptions of what quality management should look like do matter, particularly if these perceptions are unbiased in terms of normative preferences.

Institutional theory suggests that these normative views, among others, may have significant implications for the implementation and further development of quality management in teaching and learning (Vukasovic 2014, 47, 48). They have an influence on what quality management should or should not do, and may influence its future developments, the development of institutional logics (Thornton and Ocasio 1999) or the establishment of quality cultures (Harvey and Stensaker 2008), but they may also give hints on other organisational problems, like decoupling (Meyer and Rowan 1977) or hypocrisy (Brunsson 1986). Therefore, this study asks which functions quality managers normatively prefer, how these preferences can be explained, and what implications they may have.

To address these research questions, I refer to organisation theory—specifically, the logic of appropriateness and the logic of consequences. The former describes how things are done under assumptions of natural rules and legitimacy, and the latter understands actions in terms of cost–benefit analysis, rational behaviour and somewhat instrumental logics of action (March and Olsen 2004). My theoretical perspective therefore centres on two concurring logics that can be summarised as ‘quality management for learning’ and ‘quality management for control’ (Hoecht 2006, 548). Both aspects—learning and control—are helpful in the comprehension of two types of exemplary behaviour of quality managers in a still-developing field.

Empirically, I refer to a survey conducted in the higher education sector in Germany in 2015. As in many other countries, the German higher education sector has undergone significant changes with regard to quality management in teaching and learning. Nevertheless, the country has been described as a latecomer in terms of adopting business-like ‘new public management’ practices in academia (Schimank 2005, 369; Wolter 2004, 87). However, this mixture of being late and ‘comprehensive reforms’ (Krücken, Blümel and Kloke 2013, 419) makes Germany a promising case for empirical research, as it creates a research environment that allows for post-tests on existing research.

Internal quality management in German higher education

Nowadays, quality management in the higher education sector seems to be an integral part of new public management reforms. During the late 1990s and early 2000s (i.e. later than in other countries), universities and polytechnics in Germany were confronted with the challenges of reorganising internal structures in order to implement quality management (Nickel 2007). Some organisational features were already in place before the ‘new public management hype’ reached the higher education institutions, like evaluation and peer review (Anderson 2006, 163), and even these would be transformed (Gornitzka and Stensaker 2014, 184, 185). Other processes of self-evaluation and quality management have since become crucial facets of teaching and learning (Yorke 1995), and many new administrative units and positions have now been established (Krücken, Blümel, and Kloke 2009, 18–19). To a large extent, external expectations in terms of quality assurance practices and legal obligations had pressured many higher education institutions to develop internal approaches of quality management, while others implemented them voluntarily (Seyfried and Ansmann 2017).

In addition, the temporal differences with regard to the implementation of internal quality management relate to forms of external quality assurance (Stensaker 2003, Harvey 2006). For example, in Germany, some universities or polytechnics have been working with quality management or functional equivalents for years, and, in the beginning, were not influenced by existing regulations or external quality assurance (QA). In contrast, other higher education institutions are still developing their quality assurance systems, which suggests that they are doing so to obey the guidelines and rules of external quality assurance policies (see the phase model of Steinhardt et al. 2018, 8–10). Such changes can be considered in terms of three dominant organisational models: quality management departments as staff units, as line departments of the higher education institutions’ administration, or as functional equivalents (committee, commission, commissioner, etc.) (Pohlenz and Seyfried 2014, 148–151). Moreover, these developments correspond to research debates on the bureaucratisation of universities (Gornitzka, Kyvik and Larsen 1998, 28; Schneijderberg 2017, 3) and their shift towards being complete organisations (Seeber et al. 2015, 1451–1453), but also on decoupling and rational myths (Meyer and Rowan 1977, 343–345).

In sum, internal quality management in higher education is not only inspired by new public management approaches but also by obligations of external quality assurance. While the former follows the idea that greater autonomy of higher education institutions requires internal measures of quality assurance, the latter applies common standards that need to be met by higher education institutions. The motivation behind both approaches is to enhance quality and improve as well as develop teaching and learning. Quality management may provide information that enables higher education institutions to formulate evidence-based policies to improve academic life and mitigate ‘garbage can’ decision situations in ‘organised anarchies’ as well as transform a ‘loose coupling’ (Cohen, March and Olsen 1972, 1; Weick 1976, 2–10) into organisational forms that allow for more structure and clarity. As a result, quality management in teaching and learning is largely a matter of implementation, which mainly stems from concrete actions of and interactions with quality managers.

Quality management and the logics of appropriateness and consequentialism

Assuming that the implementation of quality management is largely a matter of degree, normative implications should play a correspondingly significant role. Specifically, quality management may follow two different and concurring logics, known as the ‘logic of appropriateness’ and the ‘logic of consequences’. March and Olsen (2004, 4) discussed the former as follows: ‘The term ‘logic of appropriateness’ has overtones of morality, but rules of appropriateness underlie atrocities of action, such as ethnical cleansing and blood feuds, as well as moral heroism. The fact that a rule of action is defined as appropriate by an individual or a collectivity may reflect learning of some sort from history, but it does not guarantee technical efficiency or moral acceptability.’

The authors note that this logic is identity based with experience and expert knowledge, and considers ‘how things are done’ (6), implying that efficiency, effectiveness or performance are not the only relevant criteria for taking action. Alternatively, the logic of consequentiality is based on the selection of alternatives and on rationality, whereby individuals calculate costs and benefits and act ‘in conformity with rules’ when it does not violate their expected benefits. Action is thus ‘motivated by incentives and personal advantage’ (March and Olsen 2004, 5). This logic emphasises individual preferences, substantive results, and objectives. It is used for the justification of decisions and may overpower rules.

Obviously, both logics may explain actions when implementing quality management in teaching and learning. On the one hand, quality managers may be sensitive to the particularities of higher education and aware of the implications for quality management. On the other hand, they may develop and emphasise their own interests; for example, as student advocates with a stronger consumer-oriented view. The more quality managers pursue their own interests and rationality, the stronger they adhere to the logic of consequentiality, and the more they directly prefer to influence academics from a normative point of view.

H1: The more that quality managers follow their own interests, the more they prefer to exert influence on academics.

However, the self-interest of quality managers is only one motive that may influence their interactions with academics. Lucas (2014, 220) observed forms of mutual distrust and tensions between administrative and academic staff. Even if misunderstandings and a lack of knowledge regarding what others are doing may explain such attitudes, they can still have serious implications for the implementation of quality management (e.g. in terms of monitoring or control) (Seyfried and Pohlenz 2018, 6). For instance, not trusting that lecturers intend to realise improvements in the quality of teaching and learning corresponds with stronger notions of controlling and sanctioning, which in turn relate to the theoretical focus of the logic of consequentiality on objectives and results. However, ‘control-based quality systems… can undermine the intrinsic motivation of the very people that deliver the service quality so desired’ (Hoecht 2006, 550). Therefore, in this study, I assume that, the more quality managers cooperate with lecturers, the less they normatively prefer controlling and sanctioning as adequate forms of action.

Contrastingly, if daily practices are restrictive and follow conceptions that are close to a logic of consequences, quality managers may favour measures that control and sanction lecturers (although this remains rather unusual in the contemporary German higher education system). For practical reasons, this also means that most of the current instruments do not go any further than reporting (e.g. for the provision of legitimacy) (Power 2003, 390), even though sanctioning is considered by some to be ‘very powerful in shaping operational behaviour’ (Hoggett 1996, 19). In this situation, quality managers ‘bark’ but do not ‘bite’.

Such considerations lead to further hypotheses. Note that the first three (H2–H4) refer to the logic of appropriateness, and the remaining two (H5, H6) incorporate the logic of consequentiality.

H2: The more quality managers perceive reporting as a core function of quality management, the less they favour exerting influence on academics.

H3: The more quality managers perceive the provision of evidence to academics as a core function of quality management, the less they favour exerting influence on academics.

H4: The more quality managers perceive dialogue as a core function of quality management, the less they favour exerting influence on academics.

H5: The more quality managers perceive the provision of evidence to higher education institutions’ management as a core function of quality management, the more they favour exerting influence on academics.

H6: The more quality managers perceive control as a core function of quality management, the more they favour exerting influence on academics.

In addition to daily practices, contextual factors also matter in the implementation of quality management in teaching and learning. Therefore, it is necessary to control for contextual factors, as well as considering their theoretical implications for the overall model. The contextual factors considered in the present study are: (a) interactions, in terms of resistance; (b) type of higher education institution and (c) the socialisation of quality managers. Prior literature has already addressed academics’ resistance to quality efforts (Anderson 2008; Lucas 2014), but it should be reiterated here that conflicts between quality managers and academics have at least two implications for quality managers. First, such conflicts lead to unstable relations, which pose a risk for quality managers (rather than academics) because they can harm their output legitimacy. Second, quality managers and academics can become trapped in so-called ‘repeated games’, which may negatively influence future cooperation between them.

H7: The more that resistance against quality management in teaching and learning is present, the more quality managers will prefer to exert influence on academics.

This hypothesis hints at a possible paradox. Bureaucratisation and formalisation are viewed as triggers for academic resistance (Cardoso et al. 2016, 952; Cardoso, Rosa and Videira 2018, 69), yet, if quality managers are acting according to a logic of consequentialism, resistance will force them to emphasise rules and objectives, which brings about more formalisation and bureaucratisation, potentially culminating in a ‘vicious cycle’ of quality management.

The second contextual factor (or control variable) concerns the type of higher education institution, which are generally viewed as ‘highly differentiated and [the] least integrated’ of organisations (Billing 1998, 145). Although this control variable simply measures if a higher education institution is a university or a polytechnic (assuming that actors in teaching-oriented polytechnics tend towards normative views that support controlling and sanctioning), it also provides implicit information on organisational structures and differentiation. Based on the contention that evaluation combines ‘evaluating’ and ‘administering’ parts (Wildavsky 1972, 513), I argue that these structures matter for the implementation of quality management and the development of policies (Egeberg 1999, 156–159). In this sense, a stronger dependency on teaching—for example, in terms of reputation and attractiveness to ensure constant levels of student enrolment—would bring the quality management into a stronger position with regard to teaching and learning. Contrastingly, a stronger dependency on research would rather weaken quality management in teaching and learning because teaching is not the sole base of reputation and attractiveness.

H8: Quality managers in universities, relative to those in polytechnics, favour less exerting influence on academics.

Finally, I posit that the socialisation of quality managers, in combination with their normative roles, may be relevant as well. Scharpf (1994, 33), for example, theorised that there is ‘every reason to use the tools of organisational design to create incentives that will harmonise duty and self-interest as much as possible’. Considering the different occupational backgrounds of quality managers (Kloke 2014), I assume that they may have diverse normative role models and self-interests. Some are academics and maintain close connections to academic circles; others have been recruited from their higher education institution’s administrative staff, which suggests that they will have no close bonds with academics and that mutual understanding may be restricted (Kloke 2014, 91). It is reasonable to propose that quality managers’ socialisations will have an influence on their daily practices (Van Maanen and Schein 1977, 29) and on their notions regarding the functions of quality management in teaching and learning.

For example, if quality managers have socialised in the academic world, they will likely have a better understanding of the particularities of teaching and research and will follow a logic of appropriateness. Conversely, if they are socialised in the administration realms, they will have a better understanding of administrative processes, which may imply that—in terms of Weberian bureaucracy—they prefer hierarchies, rule-bound procedures, and so on. In this case, quality management wouldn’t be considered a ‘core academic interest’ (Vukasovic 2014, 61), which could in turn result in the application of a logic of consequentiality. Thus:

H9: The more that quality managers are recruited from higher education institutions’ administration teams, the more they will prefer to exert influence on academics.

Table 1 summarises the main arguments of the theoretical section, and reveals how both logics relate to the different criteria presented in the hypotheses above.

Table 1. Summary of the study’s theoretical arguments.

Argument		Normative preferences follow the logic of consequences	Normative preferences follow the logic of appropriateness
H1	Quality managers follow own interests	Self-interest	No self-interest
H2	Quality managers perceive reporting as a core function	No	Yes
H3	Quality managers perceive quality management as related to professions	No	Yes
H4	Quality managers perceive dialogue as a core function	No	Yes
H5	Quality managers perceive quality management as related to higher education institutions’ management	Yes	No
H6	Quality managers perceive control as a core function	Yes	No
H7	Resistance (current)	High	Low
H8	University dummy	Polytechnics	Universities
H9	Background dummy (administration)	Administration	Science

Data collection and analysis

The data set was collected through a survey as part of a larger research project on the impacts and effects of quality management in teaching and learning in higher education institutions. It captured information about quality managers’ attitudes and perceptions about their profession and tasks. In particular, the questionnaire collected data about participants’ institutional backgrounds, the organisation of internal quality management in teaching and learning, the perceived effectiveness of quality management, processes of quality management, internal cooperation, resistance and competencies, as well as socio-demographic data.

The online survey was conducted in the summer of 2015, and was submitted to 238 of 279 higher education institutions (85%) in Germany where a quality management department or functional equivalent was present (commissions, committees, staff units, etc.). The sample included universities, universities of applied sciences (or higher education institutions without a right to award doctorates), and universities of art and music. Only private higher education institutions were excluded from our sample. The survey was addressed to quality management units on a central organisational level, thereby excluding faculty or departmental levels. The study’s population totalled 639 individuals who were engaged in quality management. Approximately 46% (n = 294) of all quality managers at least partly responded to the questionnaire.

For the statistical analysis of a data set, it was important that the sample featured no distortions in comparison to the overall population (i.e. that it was representative). Therefore, I controlled for different features such as the type and governance structure of higher education institutions or the individual features of quality managers (e.g. socio-demographic data). However, I found only marginal differences for these variables between the study’s sample and the wider population, suggesting that the sample was representative of quality managers in the German higher education sector.

To analyse the data set, I used descriptive statistics and ordinary least squares regression analysis. For the measurement of the dependent variable—quality managers’ notions about the implementation mode of internal quality management—I constructed a Guttman scale featuring five items. I assume that such notions are not exogenous and predetermined, but, rather, that they develop and sharpen during one’s occupational life, daily practices and experiences—and that they are endogenous and circular in terms of influencing actors’ behaviour. Therefore, I used these notions as a dependent variable.

Quality managers were asked to agree (coded ‘1’) or disagree (‘0’) in terms of what they considered to be the main functions of internal quality management in teaching and learning. In this context, quality management was described with items on ‘publishing’, ‘reporting’, ‘steering’, ‘controlling’ and ‘sanctioning’ (Table 2), which were presented in an arbitrary order. Based on these dichotomous items, I constructed a Guttman scale, which is a form of cumulative scaling that proposes to measure the consistency of respondents’ answering behaviour regarding certain attitudes. The basic idea is that items measure a one-dimensional continuum of a concept but are cumulative in their logic and intensity. That means that the first item signals rather weak agreement with the concept, while the last item measures strong agreement. The respondents are only allowed to answer with ‘yes’ or ‘no’. The overall logic is that a respondent who agrees with stronger items also would agree with weaker items but not vice versa. For the administration of these items in the survey, their logical order was mixed in order to avoid a predetermination of answering behaviour.

Table 2. Questions and items of the survey.

Question	Items
Should quality management at higher education institutions in your opinion complete the following tasks? (Possible answers: ‘yes’, ‘no,’ ‘do not know,’ ‘refuse answer’)	Quality management… …should lay the fundamentals steering
	…should report its main results
	…should support sanctions
	…should serve as a controlling instrument
	…should publish main results within higher education institutions
How far do you perceive yourself as a representative of the quality management department? (Answers on a scale ranging from 1, ‘fully correct’, to 6, ‘not correct at all’)
How far do you personally agree that the following aspects are functions of quality management in teaching and learning at your higher education institution? (Answers on a six-point Likert-scale ranging from 1, ‘fully applies’, to 6, ‘does not apply’)	Providing reports
	Provision of relevant information for higher education institutions management
	Provision of relevant information for higher education institutions departments
	Quality control in teaching and learning
	Promotion of dialogues between actors of the higher education institution
Do the departments currently resist against quality management in you higher education institution? (Possible answers: ‘yes’, ‘no’, ‘do not know’, ‘refuse answer’)
At which type of higher education institution are you employed? (Possible answers: ‘Universities’, ‘Universities of Applied Sciences’, ‘Art and Music Universities’)
Please indicate all areas in which you have worked after completing your training and before starting your current position. (Multiple answers possible)

A Guttman scale is considered to be valid if respondents demonstrate consistent and coherent answering behaviour. That is, agreement for certain items implies agreement for others. For example, in the present study, if quality managers prefer to steer, they need some sort of reporting or they have to publish the results to different stakeholders and decision makers within higher education institutions. Therefore, quality managers who agree with the ‘steering’ item are likely to agree with the items on ‘reporting’ and ‘publishing’. Or, for other quality managers who prefer to sanction, agreement with the items relating to ‘controlling’, ‘steering’, ‘reporting’ and ‘publishing’ would be expected. All deviations from this coherent answering behaviour were coded as errors. Consequently, the Guttman scale both uncovered quality managers’ notions about the functions of quality management and gave an overview of the consistency of these notions. Two variables were thus constructed: an error variable measuring the number of inconsistent answers and an index measuring quality managers’ notions about the functions of quality management.

For this purpose questions were included concerning quality managers' normative assessments of core tasks in quality management, with the survey featuring items ranging from ‘reporting’ to ‘sanctioning’. In total, 224 participants responded to all relevant items, while some respondents left some questions unanswered. Therefore, I used a statistical approach to estimate the missing values for these respondents, applying a procedure on a median basis. If the procedure created values of either 0 or 1, respondents were included in the analysis. Otherwise, the information of respondents who replied, for example, to four instead of five items would be completely lost. If the procedure created values of 0.5, respondents were excluded from the data set, because these values were not applicable in the study's Guttman scale. This supplementary procedure made 85 additional respondents available for statistical analysis because they now had on each item of the dependent variable either 0 or 1. Thus, after the inclusion of respondents with estimated values, the overall number of cases increased from 224 to 309 for the dependent variable.

This study’s independent variables, which attempt to explain normative perceptions, are closely related to the theoretical discussion above. I measured quality managers’ self-interest as a general assessment of how far they perceive themselves as representing their own interests. Further items considered the quality management practices and functions of the university or polytechnic where the quality managers were working. All of these items were measured on a six-point Likert scale. For contextual information, such as sociodemographic data or type of higher education institution, I used simple ‘0’ and ‘1’ coded dummy variables. Thus, quality managers were asked if they faced resistance from academics in their current position (yes/no), if they belonged to a university instead of a polytechnic (yes/no), or if they had been recruited from the administrative branch of the higher education institution (yes/no). These variables are the basis for the empirical analysis (Table 2).

Results and discussion

Considering the patterns of the dependent variable, quality managers were very stable in their normative views regarding what quality management should do. The coefficient of reproducibility was 0.95, signalling that use of the Guttman scale was acceptable (Jobling and Snell 1961, 115). Although this result does not necessarily mean that all items were on the same dimension, it does indicate that the answering behaviour was consistent (Schooler 1968, 296). Overall, only 71 errors were recorded, including 22 cases with two errors and nine cases with three errors; for example, when quality managers preferred control but did not favour steering, or when they preferred reporting but not publishing. However, the vast majority of answers ranged between steering and sanctioning. Beyond this, the most remarkable result is that quality managers displayed rather closed normative notions, which may result in consistent preferences regarding how quality management should look in their personal view. This underlines why it is so relevant to explain these normative notions, as this may have serious consequences for the further directions concerning how quality management could develop.

Normative worldviews are not exogenous, but, rather, endogenous and influenced by different institutional and individual variables. I included these variables in a regression model in order to explain notions of how quality management in teaching and learning should perform. Specifically, I calculated ordinary least squares regressions including the variables presented. The dependent variable was the score of the Guttman scale, ranging from 0 to 5 (Table 3), indicating quality managers’ normatively preferred functioning logic of quality management. Accordingly, I used the other items from the study’s questionnaire to test the hypotheses presented.

Table 3. Ordinary least squares regression analysis results.

	Model I		Model II
	Coeff.	SE	Coeff.	SE
Constant	1.64***	0.56	0.93	0.73
Representing quality management’s own interests	0.13*	0.07	0.11	0.08
Quality management as reporting tool	0.08	0.06	0.09	0.07
Quality management as provider of evidence (management)	0.14*	0.08	0.15	0.09
Quality management as provider of evidence (profession)	−0.01	0.09	−0.05	0.10
Quality management as control of teaching and learning	0.25***	0.05	0.27***	0.06
Quality management as promotor of dialogue	−0.19***	0.06	−0.13*	0.07
Resistance (current)			0.22	0.17
University dummy			−0.11	0.15
Background dummy (administration)			0.37**	0.17
Number of cases	165		149
R	0.52		0.54
R^2	0.27		0.29
R^2-corrected	0.24		0.24

*p < .10.

**p < .05.

***p < .01.

Table 3 displays the results of the two models from the ordinary least squares regression analysis. Model I analyses the influence of quality managers’ assessment of the daily quality management work of their respective higher education institution on their notions (Guttman scale). Two coefficients of this model are strongly significant: the practice of quality management as a control of teaching and learning and the assessment of quality management as a promoter of dialogue between different actors within the higher education institution. Consistent with the theoretical assumptions in H4 and H6, both coefficients exhibit the predicted sign; thus, while the control coefficient is positively correlated with the score of the Guttman scale, the dialogue coefficient reveals a negative sign.

In sum, the study’s results indicate that quality managers have different understandings of quality management that are inspired by their daily practices. Some understand quality management as a soft instrument inspiring negotiations and communication about quality in teaching and learning (Stensaker and Leiber 2015, 332), while others consider it to be an instrument for the execution of standards and norms. While the former corresponds with the logic of appropriateness in terms of dealing with independent professionals, the latter corresponds with the logic of consequentiality that presumably wants to force them into rationalised steering instruments.

Furthermore, the first model shows two other relevant coefficients that are significant at the p < .10 level: quality managers who perceive quality management as an instrument for the provision of evidence to higher education institution management tend to perceive it also as a normative instrument with which to sanction and control others (H5). It seems reasonable to infer that, if quality managers assign credibility to information on teaching and learning gathered by many different instruments, they would try to use this information for further action too. However, this result, considered alongside the coefficient on a quality management’s own interests, indicates a possible tendency for ‘agency drift’ (H1). Hence, quality managers may tend to follow at least partly their self-interests, and this allows and justifies controls and sanctions from a normative point of view.

Both results could have serious implications for the future implementation of quality management. That of the control variable indicates that quality managers who already use quality management as a tool of control would probably make the next step. Consistent normative attitudes towards control and sanctioning may be considered as an opener towards a restrictive quality management in the sense of the logic of consequentiality, which directly wants to influence academics and may further challenge their particular role as professionals. Contrastingly, the result with the dialogue variable shows the opposite and points to the logic of appropriateness. Such a non-restrictive quality management functions as an instrument of dialogue, which leaves room for manoeuvre for academics and may accept their particular role as professionals. Beyond this, the results may also give first hints on future developments of quality management presumably resulting in different institutional logics (Thornton and Ocasio 1999) at least partly influenced by their normative notions, meaning that institutionalisation processes in quality management are still underway.

Model II in Table 3 includes the control measures; namely, the dummy variables for existing resistance in the higher education institution, type of higher education institution and quality managers’ occupational background. After including the control variables, the significance of the coefficients for quality managers’ self-interests (H1) and quality management as a tool for the provision of evidence to higher education institutions’ management (H5) is lost. In contrast, the effects for quality management as an instrument of control (H6) or as an instrument of dialogue (H4) remain significant. Beyond this, one of the control variables (occupational background) becomes significant, indicating that occupational background matters (H9).

The inclusion of the control variables not only reveals that the results are robust, but also provides further evidence for the ideas resulting from the first model. As shown in Table 3, the socialisation of quality managers relates to their normative understandings towards sanctioning and controlling. The coefficient of this control variable even outweighs quality managers’ perceptions of quality management practices. Hence, those who come from a higher education institution’s administration level, which is characterised by routine tasks and bureaucratisation, normatively prefer quality management that perpetuates these experiences. Consequently, they may transmit these routines onto the implementation of quality management in teaching and learning. This could force some academics into further formalisation and bureaucratisation (logic of consequences) of teaching and learning, while others may experience constant or even increasing levels of academic freedom (logic of appropriateness).

So far, these conclusions seemingly relate to institutional logics. However, these results may also have implications for rational myths because it remains to be determined whether higher education institutions with stronger pressure on academics correspond with higher levels of decoupling, while more inclusive and participative quality management might reveal lower levels or even no decoupling of formal structure and action. At a first glance, this result seems to be counterintuitive, but it gives hints on the strategic use of coupling or decoupling in terms of reactions and counter-reactions of various actors within higher education institutions.

Furthermore, these results show that the selection of staff may have serious implications for the implementation of quality management in terms of what should be done (Houston and Studman 2001, 484). That is, the recruitment of quality managers may become an important source for a general understanding of the implementation of quality management: on the one hand, it might be implemented as a tool of administration and control, and, on the other, it could be implemented as a creative task for internal policy development.

Conclusion

The present article addresses one of the main issues of contemporary quality management within higher education by questioning how quality managers perceive the functions of quality management from a normative point of view, and then explains the main determinants of these assessments. For this purpose, the logics of appropriateness and of consequences were together applied as a relevant theoretical perspective (March and Olsen 2004). To further investigate the research questions, I constructed a Guttman scale measuring quality managers’ normative perceptions of functions of quality management (i.e. reporting, publishing, steering, controlling and sanctioning) as the dependent variable. The results showed that quality managers demonstrate consistent answering behaviour regarding the notions related to these functions, such that they preferred softer functions like ‘reporting’ and ‘publishing’ if they also favoured harder instruments like ‘steering’, and, conversely, they did not consider combinations like ‘sanctioning’ and ‘controlling’ without ‘reporting’ or ‘publishing’.

In sum, my research provides evidence that quality managers’ normative perceptions of the functions of quality management are primarily driven by the current practices of quality management, but also by the quality managers’ individual backgrounds. This has far-reaching implications for the interactions of quality managers and academics, and thus for the effectiveness of quality management systems in higher education institutions.

Beyond the influence on the direct interactions between quality managers and other actors within higher education institutions, the different logics of action and the normative notions may influence the institutionalisation of quality management and show thus interesting connections with neo-institutionalist theories. Considering that quality management is still a rather young phenomenon, the results seemingly point towards the development of different institutional logics (Thornton and Ocasio 1999) of quality management. The logics of action could transform into something that belongs to the legitimacy-generating organisational structures of higher education institutions, and, although the logics differ, both intend to follow the same purpose, which is to ensure quality in teaching and learning. Undoubtedly, they will have very different implications for the actors within higher education institutions.

Beyond this, conceptual connections with theories like decoupling (Meyer and Rowan 1977) also appear relevant for further research. The different logics of action put diverse pressures on various groups within higher education institutions. Hence, they may cause forms of decoupling between daily actions and formal structure, implying that quality management may become just an organisational and formal exercise, which rather perpetuates rational myths instead of providing procedures that influence the development of teaching and learning.

In general, the further investigation of quality management in higher education institutions not only incorporates interesting empirical questions, it makes theoretical combinations necessary in order to understand the intended and unintended effects and outcomes, as well as their implications. That is, particularly in this field, research on higher education institutions may also inspire organisational research and its theoretical development.

Acknowledgements

I am grateful to Ina Radtke and Moritz Ansmann for their comments on earlier versions of this article. Furthermore, I want to thank the two anonymous reviewers for their very helpful comments.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work originates from a research project supported by the German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung) [grant number 49-01PY13003A].