Institutional logics make it possible to consider interactions across social, field, organizational, and individual levels (Thornton et al., 2012). At the level of society, scholars have identified seven distinct institutional orders and their associated logics [i.e., family, community, religion, state, market, profession, and corporation; Thornton et al., 2012; also discussed by Friedland (1991) and Thornton (2004)]. These logics encompass broader cultural, social, and economic systems that influence ways of thinking, organizing, and behaving in society overall (Besharov and Smith, 2014). Whereas field-level institutional logics refer to the prevailing sets of beliefs, values, norms, and practices that shape organizations’ behavior and decision-making within a specific field or industry. These logics provide a shared understanding of what is considered appropriate and legitimate in the field, guiding the actions and interactions of organizations, professionals, and other stakeholders (Besharov and Smith, 2014).

To gain insights into the different institutional logics to which individuals are exposed in (H)EIs, we draw on the work of Cai and Mountford (2022), whose systematic literature review finds that field-specific logics of higher education can be clustered in four main groups: professional logics, market logics, corporate logics, and state logics. The academic, or professional, logics consider the profession as a relational network and emphasize status in a profession through high-quality teaching and research; market logics view (H)EIs as a transitional entity aiming to increase profit; corporation logics consider (H)EIs as hierarchical entities with an emphasis on status; and state logics emphasize the important role of (H)EIs in contributing to the common good (Cai and Mountford, 2022).

The logics that drive individual behavior are to a large extent determined by the availability and accessibility of logics as these direct the individuals’ attention (Thornton et al., 2012). This makes it important to consider that these field-level logics will probably not be equally central to the organization and, and consequently to the individuals working within it. Based on Durand and Jourdan (2012), we refer to the logics that play a bigger role in informing an organization’s objectives, vision, and practices as the ‘dominant logics’ and any additional logics as ‘minority logics.’ In the educational field, market, corporate, and professional logics are generally reported as the dominant logics (Louw, 2019), with the state logics being considered a minority logic. Table 1 provides an overview of the educational field-level logics used in this study.

To understand how engaged practitioners experience institutional complexity, and the nature of this complexity, individual motivations and behavior can be associated with the above-mentioned logics (Greenwood et al., 2011). For instance, engaged practitioners’ might experience institutional complexity because their own intentionality is shaped by the state logics of common good, aiming for redistribution (Vargiu, 2014; Shea et al., 2023). These practitioners could experience increased institutional complexity due to the dominant field-level market logic of increasing efficiency (Louw, 2019; Shea et al., 2023). The presence of different institutional logics elicits contrasting responses from individuals, including ignoring, rejecting, and adopting them (Pache and Santos, 2013; Gautier et al., 2023).

To gain insights into possible responses of engaged practitioners, we build on the work of Pache and Santos (2013), who describe five patterns of individual responses in the face of competing institutional logics, namely: Ignorance, Compliance, Defiance, Compartmentalization or Combination. Ignorance refers to the lack of response because the individual is unaware of the logics’ influence. Compliance refers to ‘an individual’s full adoption of the values, norms, and practices prescribed by a given logic’ (p.12). Defiance on the other hand, refers to an individuals’ explicit rejection of a given (dominant) logic. Compartmentalization refers to the purposeful segmentation of two competing logics by an individual. Finally, combination refers to the individuals’ attempt to blend values, norms and practices associated with the competing logics.

Pache and Santos (2013) also note that when individuals are exposed to a single logic, they might resort to ignorance, compliance, or defiance. When individuals are exposed to institutional complexity as a result of tensions between competing logics, they might resort to more complex responses such as compartmentalization or combination. Pache and Santos (2013) also outline the importance of exposure and adherence to a particular logic in shaping certain responses. For instance, when individuals are less familiar with a particular logic and identify with another logic, they will comply with the logic they identify with. However, when an individual identifies with 2 competing logics, they will most likely try to combine these logics.

In order gain insights into how engaged practitioners experience institutional complexity and how they respond to this complexity, we included three (H)EIs within the region of Amsterdam, the Netherlands.

In the Netherlands there are three main educational routes after secondary education:

From now on, we refer to MBO, HBO, and WO institutes. The Dutch context is a good basis for understanding how institutional logics at the organizational level play out at the individual level for multiple reasons. First, the Netherlands has a rich history of engaged educational practices, such as the Science Shops movement in the 1970s (Dixon, 1988; Urias et al., 2020). Second, in the Netherlands, the expansion of core missions and visions are apparent across the whole spectrum of post-secondary education institutes from MBO (Simons et al., 2000; Geurts and Meijers, 2009) to HBO and WO (Meijs et al., 2019). Lastly, this expansion is informed and influenced by policy initiatives such as the City Deals on Education, which have fostered engaged educational projects in which cities (local governments), MBO, HBO, WO, civil society and / or the private sector collaborate to find solutions to pressing local social and urban challenges (see for example: Mees et al., 2019; Sibbing et al., 2021). Within the Dutch context, the city of Amsterdam is one of about 10 cities in the Netherlands in which WO, HBO and MBO are present.

In this study we included one WO: the ‘Vrije Universiteit Amsterdam’ (VU), one HBO: the ‘Hogeschool van Amsterdam’ (HvA), and one MBO, the ‘Regionaal Opleidingscentrum’ (ROC) institute. Each institution has a mission statement that includes some kind of emphasis on the importance of embedding engaged education through engaged knowledge production and active citizenship. Among other terms ‘active citizen’ in the MBO, ‘enables innovation in the professional sector and community’ in the HBO and ‘fully engaged with society’ in the WO. For the entire mission statement for each institute see Table 2. We deemed it necessary to include institutions of the three post-secondary routes because each, in its own way, aims to contribute to social, economic, and cultural development in the region of Amsterdam. Furthermore, they all operate in the same landscape—as illustrated by the City Deals on Education—and are all exposed to the dominant field logics of market and corporation (van Houten, 2018).

Within each included institute, we used a purposeful sampling strategy (Merriam, 2015) to identify four to five engaged practitioners that work across diverse domains (in the case of MBO) or faculties (in case of HBO and WO). To safeguard anonymity, we refer to all engaged practitioners with fictional names starting with an M for the MBO institute, an H for the HBO, and an W for the WO institute. Respectively: Mila, Max, Mark, Melanie, and Maya (MBO); Hellen, Hope, Hanna, and Henk in (HBO); and Will, Wade, Walter, and Wesley (WO; see also Table 2). All practitioners included in the study were involved with an engaged education project. To qualify as an engaged education project, the project had to be part of a credit-bearing course that aims to contribute to a social issue in collaboration with an external social partner (based on: Bringle and Hatcher, 1995). Examples of issues addressed are: the direct support of elderly people by students in a health facility, the assistance of students in in financial matters such as tax returns, or creating an action plan for an external commissioner to increase individual sustainable consumer behavior. A description of all the included engaged education projects (anonymized and in random order) can be found in appendix A.

In total, 13 in-depth semi-structured interviews were conducted with practitioners that design, coordinate, and sometimes teach engaged education projects within different institutional domains or faculties: five with practitioners from MBO, four with practitioners from HBO, and four with practitioners from WO (see Table 2). The interviews took place between January and May 2020. The interviews lasted for approximately an hour, and were, because of the COVID-19 pandemic, mainly held online through Zoom. All interviews were conducted by the first author. Interviews were held in Dutch and quotes were translated into English.

The interview guide builds on a systematic literature review of the process of embedding engaged education within (H)EIs. This review found that this process can be segmented into startup, scale up, and sustaining phases. The review concludes that the individual practitioners play an important role in embedding engaged education but also acknowledges the importance of organizational practices such as awards, promotion criteria, funding allocation, assessment, and formalization within educational structures (just to name a few).

At the start of each interview, respondents were asked to describe the design of, and the intentionality behind their engaged education project. So, what was the intention or urgency of the practitioner in starting this project? Subsequently, the respondents were asked to elaborate on their considerations in relation to organizational support related to policies and practices such as recognition, development programs, and allocated time. Lastly, respondents reflected on the possibilities for sustaining their project and the challenges they thought they might encounter in doing so. Therefore, reflecting together with the practitioners on their considerations in relation to the factors identified in this review for each phase, enabled us to outline their individual intentionality as well as their perception of the existing or desired organizational support or tensions when embedding engaged education. These considerations could then be associated with institutional logics (see next section) to allow for the consideration of how institutional complexity is experienced and responded to. All interviews were recorded and transcribed verbatim.

All the transcripts were analyzed in ATLAS.ti 22. First, through deductive reasoning, we associated the field-level logics as identified by Cai and Mountford (2022; see Table 1) to the practitioners’ intentions and considerations relating to their perceived organizational support and tensions when embedding their engaged education project. For instance, when practitioners noted that their intentionality in the engaged education project was to attract more students and thus boost numbers, it was coded as: ‘adhering to market logic’. Alternatively, when a practitioner would note tensions in relation to work productivity or ranking when embedding their engaged education project, this was coded as: ‘conflicted by corporate logics’. This allowed us to associate field-level logics to individual intentions and considerations.

In parallel to deductive reasoning, we employed inductive reasoning to identify patterns in logics related to intentionality, conflicting logics and ways in which practitioners would deal with institutional complexity. By comparing and contrasting the experiences of those 13 individuals, we could identify one additional field-level logic that is not mentioned by Cai and Mountford (2022). This is described in the first part of the result section (see also Table 3 in the result section).

Finally, to better grasp how those individuals responded to the multiple logics that are available to them, as well as to better embed our findings in the literature, the last stage of our analysis consisted of an abductive reasoning approach which required a constant movement back and forth between theory and empirical data (Mantere and Ketokivi, 2013). Our initial empirical findings directed our search for additional literature that could offer theoretically-sound explanations to the emergent patterns that we observed by comparing and contrasting the responses of the 13 individuals in our study. Through this process of iteration between data and theory, we identified the study of Pache and Santos (2013; see also section 2), who developed a model that predicts which response organizational members are likely to activate as they face two competing logics.

We found that most of the individuals in our research activated responses (as well the associated roles attached to those responses) as modeled by Pache and Santos (2013). For instance, when a practitioner described purposefully segmenting their engaged teaching and research roles to deal with the bureaucratic organizational roles and structures, they were classified as taking the role of ‘compartmentalizer’. When a practitioner described how they combine their practices related to different logics, for instance by involving community in acquisition thereby adhering to increasing revues but also purposely contributing to community good, they were classified as taking the role of ‘combiner’.

By comparing and contrasting the responses of the individuals in our research, we found however, that the model of Pache and Santos (2013) could not predict all the patters that we identified in our empirical analysis. This led to further iteration with literature (e.g., Perkmann et al., 2019) that gave us a sufficient basis to propose explanations for responses and roles that we observed in our study and that were not yet represented in the literature in a way that corresponds to our study context. We describe those roles and responses in detail in the first part of our results section.

Participation in the study was on a voluntary basis. All parties signed an informed consent form before participating in the interviews. Participants were told that they could withdraw from the study at any stage if they wished to do so. Audio recordings and transcripts were stored on a safe drive that is password protected and can only be accessed by the researchers. After transcription, the audio recordings were destroyed. All data transcripts were anonymized. In accordance with the regulations of the VU Amsterdam the anonymized transcripts are stored for 5 years.

In adherence to the ethical standards established at VU Amsterdam, our research aligns with the prescribed conditions for human subjects, which encompass:

Moreover, our research adhered to the following guidelines:

The study also adhered to the Code of Ethics for the Social and Behavioral Sciences of VU Amsterdam (2016). As a research undertaking classified as ‘standard’ and compliant with the established guidelines, further scrutiny by the Research Ethics Review Committee was deemed unnecessary.’

Through our deductive coding we associated individual intentions and perceived institutional support and tensions to the state, market, corporate, and professional logics (see Table 1). However, during our analysis, in which we compared and contrasted the 13 interviews, individual intentions as well as perceived tensions went beyond the categorization of these four logics. We found for instance, that some partitioners emphasized the need for non-hierarchical structures, common boundaries, and trust in relation to their engaged education projects. Though similar to the state logic, these considerations seem to fit better with the community logic (also defined by Thornton et al., 2012), as this logic particularly stresses the importance of common boundaries, trust, and reciprocity. We therefore opted to introduce a fifth field-level logic namely the community logic and coded this as ‘adheres to community logic’ (see also Table 3).

In addition, our analysis uncovered instances where the responses of engaged practitioners extended beyond the five patterns of individual responses outlined by Pache and Santos (2013) namely: Ignorance, Compliance, Defiance, Compartmentalization or Combination. Specifically, we found that some practitioners refused to compromise toward more dominant market or corporate logics of bureaucratic roles or increasing revenues, in order to protect the minority logics of, for instance, increasing community good. This response is different from the response of Defiance as described by Pache and Santos (2013), in which the minority logic is ignored. It is also different from the response of Perkmann et al. (2019) in which the dominant logic is shielded from the influence of minority logics. We therefore opted to introduce a new response namely: Protector. Within this response, practitioners not only adhere to minority logics (of state and community) and reject more dominant logics (of market, corporate, and profession), but also protect the minority logics from the influence of the more dominant ones. For instance, when a practitioner refuses to compromise in their flexible approach, going against rigid bureaucratic roles to protect their commitment to community—this response was coded as ‘protector’.

After comparing and contrasting 13 interviews through combined deductive, inductive, and abductive reasoning, we eventually identified five field-level logics that engaged practitioners might adhere to or be conflicted by when embedding engaged education, namely: state, market, corporate, professional, and community logics. Moreover, we identified four engaged practitioner roles within our study: Compliers, Compartmentalizers, Combiners, and Protectors.

We present the rest of the results section within the four roles that emerged after comparing and contrasting 13 interviews. Within each section, we describe the intentionality, institutional tensions and support (connected to their corresponding logic), and responses (if any) for each of the 13 practitioners included in this study. Table 4 shows an overview of each role and the related response, the name of the engaged practitioner, the logic associated with their intentions, and the logic they perceive as conflicting (if any).

The current study included only four to five examples of engaged education in each (H)EI. Although we conducted purposeful sampling strategies with the aim of including a variety of domains and faculties, we cannot draw conclusions at the institutional level of differences and similarities between the institutions based on such a selective and limited sample. Our aim was, however, not to map the institutional-level structures in their entirety, but rather to gain insights into how individuals within different (H)EIs in the Netherlands navigate institutional complexity when designing and coordinating engaged education. Future studies should consider whether the individual roles of Complier, Compartmentalizer, Combiner and Protector are more widespread. The exploratory nature of our study does not allow us to draw inferences about the relative frequency of those roles or which are more or less frequent in each institute that we considered. For instance, the fact that we did not identify combiners in HBO does not mean that they do not exist or that they are less prevalent in that institute. Conversely, the fact that protectors were identified only in HBO does not mean that individuals do not take on this role in other institutes.

Moreover, it should be noted that we did not take the background and contextual factors of the engaged practitioners into account when researching their responses. A recent study by Gautier et al. (2023) conducted 14 life-story interviews and showed the importance of contextual factors and the crucial role of individuals’ adherence to a certain logic. Similar to this study they considered how field logics influence individual responses. However, for Gautier et al., individuals were not embedded in an organizational setting and therefore the findings are bounded by an extra-organizational setting. It would be interesting for future research to apply the findings van Gautier et al. (2023) to the institutional and organizational field of education. For instance, we could not draw any conclusions about the individual and organizational factors that made those responses possible. At the organizational level, further research might examine how far factors such as position, job security (e.g., permanent versus temporary contracts), career stage, organizational (sub-)cultures, among others, influence the role and associated responses that individuals take when confronted by institutional complexity.