Bureaucratic Reputation Theory: Micro-Level Theoretical Extensions

Abstract

Bureaucratic Reputation Theory (BRT) focuses on the role of public agencies’ reputation as an asset in socio-political dynamics. Agencies aim to manage their reputation for different audiences to have higher levels of (publicly legitimized) strategic independence, autonomy, and discretion. Considering that reputations form because of shared reputational beliefs among individuals, we study bureaucratic reputation from a dialogic perspective between agencies and the individual stakeholders in their audiences. First, we make a case that such socio-cognitive elements are relevant for a broad range of public-serving organizations, pinpointing the broader relevance of BRT beyond public agencies. Second, building on interdisciplinary insights on the formation and evolution of individual perceptions, as well as the social network interactions within and between audiences, we derive 10 micro-level theoretical propositions in three related themes: (1) distinct information sources for reputational beliefs, (2) the episodic nature of agency-audience interactions, and (3) the reputation spillovers between structurally related units.

NEED FOR MICRO-LEVEL EXTENSIONS IN BUREAUCRATIC REPUTATION THEORY

Bureaucratic Reputation Theory (BRT) posits that public agencies manage their reputations towards different audiences to be able to use these reputations as a political asset in their relations with policy makers and other stakeholders in their public governance networks (Carpenter 2002; Van Der Veer 2021; Wæraas and Maor 2014). In recent years, BRT has gained relevance, as a growing body of literature in the field of public policy and administration focuses on the reputation of public institutions and (1) how their reputation is shaped (Carpenter and Krause 2012), (2) how reputation is a crucial part of public accountability mechanisms (Bertelli and Busuioc 2021; Busuioc and Lodge 2017; Leidorf-Tidå 2022; De Boer 2023), (3) how reputation is managed by agencies through their communications and outputs (Boon and Houlberg Salomonsen 2020; Maor and Sulitzeanu-Kenan 2013, 2015; Van Der Veer 2021), and (4) how input and participation from stakeholders determine agency actions to improve and maintain their reputation (Bunea and Nørbech 2023; Fahy 2022). The degree to which these studies explicitly claim to build on and/or contribute to BRT varies. While some authors explicitly refer to and frame their reputation-related contributions within BRT, others refer to the broader relevance of studying the reputation of public service organizations. All studies build on common notions of organizational reputation of public institutions (Frandsen, Johansen, and Houlberg Salomonsen 2017), and they are similar in their argumentation on the relevance of reputation for public institutions.

Concretely, many existing contributions explicitly refer to reputation at the organizational level, with the aim to discuss macro-level dynamics between public institutions—commonly referred to as agencies—and their stakeholder groups, often referred to as an agency’s audiences (Bach et al. 2019; Boon, Salomonsen, and Verhoest 2021; Carpenter 2001). However, in many of these contributions, audiences are conceptualized as monolithic blocks, such as for example “politicians,” “citizens,” “media,” or “other agencies,” meaning that seldom attention is given to the individuals belonging to these stakeholder groups and their concrete interactions with a public agency and/or with other individuals in and outside their audience. While this organization-level approach has enabled a better understanding of the overall politico-sociological dynamics of agencies, less scholarly attention—at least in the context of public administration—has been given to the individual and behavioral aspects of reputation. Some exceptions—which also call for more micro-level elaboration of BRT—are provided by Maor (2016), Lee and Van Ryzin (2019; 2020), and Döring and Willems (2021). Such individual-level elaboration of the role of reputation can provide a better understanding of how concrete reputation-related decisions are made by various internal and external stakeholders of public agencies (Ravasi et al. 2018). A micro-level elaboration of BRT can complement and advance existing insights on agency-level strategic and coordinated considerations about interactions with aggregated audiences.

Reputation at the micro-level has been studied extensively in other fields, including management science (Barnett 2014; Bitektine et al. 2020; Highhouse, Brooks, and Gregarus 2009; Vanacker and Forbes 2016), social psychology (Baer et al. 2018; Schlachter and Pieper 2019; Suzuki et al. 2016), and behavioral economics (Fu and Li 2014; Carpenter and Myers 2010). However, in these fields, specific attention is seldom given to the bureaucratic element of BRT, meaning that the unique aspects of public institutions and citizens interacting with these institutions are ignored. Individuals act differently and consider other reputational elements relevant in interactions with organizations in their roles as citizens being impacted by policy or as service beneficiaries of public services, compared to their roles as customers of—for example—multinational companies (e.g., Bertelli 2016). In their roles as citizen or public service beneficiaries, reputational dynamics remain largely ignored, which is where insights from BRT can in turn complement the general management and social psychology literature.

Combining these two literatures, it is important to delineate the types of organizations in which individuals have roles where the public aspect of organizations matters. Therefore, we follow other BRT scholars for the development of our theoretical micro-level propositions by focusing on agencies in a non-narrow sense, meaning that we have a range of various public-serving organizations in mind—also referred to as “modern” agencies (Maor 2016). This has two main reasons. First, many organizations are potentially not strictly classifiable as a “public agency” (such as ministries or government departments) but can be considered as quasi-public organizations. These are organizations, which do not necessarily have a public legal form but are often closely linked to such ministries or government departments. For example, many organizations that are legally classified as nonprofit or for-profit organizations are often heavily or even entirely funded through government funding schemes with strict contracts and accounting/accountability measures to control public resources as well as the performance of public services (Eikenberry and Kluver 2004; Maier, Meyer, and Steinbereithner 2014; Willems 2021). Other organizations are partially or fully state-owned (Grossi, Papenfuß, and Tremblay 2015).

These legal constructions, along with the growing hybridity of various types of organizations and industries, have gained importance in recent decades, often as part of new public management restructurings (Hood 2010; Moynihan 2006). In the new public management paradigm, service contracts with or direct ownership of private for-profit or nonprofit organizations are advocated as public governance mechanisms that can increase steerability and efficiency, in combination with higher public accountability (Busuioc and Lodge 2017; Zimmer and Rathgeb Smith 2021). As a result, such public-serving organizations can be considered quasi-public organizations, and they are crucial for the delivery and safeguarding of various public goods and public services.

Second, these public-serving organizations—despite having different legal forms—have very similar characteristics and managerial challenges in terms of (1) how they are perceived by citizens and other audiences, (2) how they justify their role (and public funding) in society, and (3) how that impacts decisions and consequences of accountability towards various audiences, including policy makers, politicians, and citizens (Bertelli and Busuioc 2021; Busuioc and Lodge 2017; Maggetti and Papadopoulos 2023). Therefore, from a reputation perspective, these public-serving organizations have in common that their goals are often (1) multidimensional (Boon, Salomonsen, and Verhoest 2021; Carpenter and Krause 2012), (2) hard to quantify (Figlio and Kenny 2009; Lecy, Schmitz, and Swedlund 2012; Sowa, Selden, and Sandfort 2004), and (3) subjective, based on different stakeholder perspectives (Boon, Salomonsen, and Verhoest 2021; Carpenter and Krause 2012; Maor 2016). These specific characteristics of public-serving organizations are causing similar challenges in maintaining their reputations and interacting with citizens and other audiences in the public sphere, regardless of the concrete legal form under which an organization or agency is operating. Hence, in this article, we refer to agencies and public-serving agencies interchangeably. This includes organizations in the public sector that are mainly funded through taxes, but also private organizations that have a public goal, such as public-owned companies (e.g., several national train companies and municipal electricity providers in main-land Europe) or (subsidized) nonprofits that focus on a general social need such as education, healthcare, culture, or other societal issues (Willems, Jegers, and Faulk 2016).

This article starts from summarizing some recurrent lines of thought—mainly at the macro-level—across contributions that refer to BRT. The ambition of this macro-level literature overview is to provide some foundations to build on and to clarify the concepts used for the elaboration of theoretical extensions on a micro-level—the focus of this article. We then develop concrete micro-level extensions in three themes by deriving and formulating 10 theoretical propositions. For each of the themes, we also provide suggestions for further theoretical elaboration and empirical validation.

MACRO-LEVEL FOUNDATIONS OF BUREAUCRATIC REPUTATION THEORY

When arguing for the origin of BRT, many point to Carpenter’s (2001)  The Forging of Bureaucracy Autonomy. A central theme in this book, as well as in the work closely building on it, is the conceptualization of multiple audiences. An audience in the context of BRT is defined as a stakeholder group that includes actors with similar needs and preferences in relation to the focal agency, such as relying on or being impacted by the same tasks or activities of the focal agency. As a result, within a stakeholder group, actors would consider particular goals, activities, and/or tasks of the focal agency more relevant for their own needs and preferences. This provides the basic elements for individuals within an audience to evaluate the reputation of the agency (Boon et al. 2019; Boon, Salomonsen, and Verhoest 2021). As a result, for various audiences, different (dimensions of) agency goals and activities have different relevance for the formation of the agency’s reputation (Verhoest et al. 2023), leading to different audiences considering different elements to evaluate an agency’s reputation. An agency can thus have different audience-dependent reputations, and adjust its communication to each of these audiences (Boon and Houlberg Salomonsen 2020; Lee 2022; Van Der Veer 2021). In BRT, this is referred to as the inherent multi-dimensionality or multifaceted nature of bureaucratic reputation (Carpenter and Krause 2012). Consequently, an agency focuses its strategic actions not only on one overall reputation, but considers different dimensions of its goals and activities, and how they are relevant for different audiences.

While the multiple dimensions of agency goals and activities determine the audience-specific evaluation criteria of an agency’s reputation, the extent that actual agency decisions, outcomes, and achievements match with their audiences’ preferences and needs determines the formation of a good or bad bureaucratic reputation (Aleksovska, Schillemans, and Grimmelikhuijsen 2022; Bach 2022; Van Der Veer 2021). Central to BRT is the premise “that agencies can build and maintain a unique reputation which serves as a political asset” (Boon et al. 2019, 428). Agencies can actively manage their actions, as well as their communication about their actions, to form and maintain good reputations among various audiences. The resulting reputation can in turn be leveraged in interactions with political decision makers and legislators (Maor 2016; Frandsen, Johansen, and Houlberg Salomonsen 2017; Lee and Whitford 2013). Moreover, a good or bad reputation can also determine an agency’s choices on how to achieve its organizational goals, while maintaining or improving its reputation (Doering et al. 2021; Maor and Sulitzeanu-Kenan 2013; Maor, Gilad, and Bloom 2013). Consequently, reputation as a political asset is related to higher levels of (publicly legitimized) discretion, autonomy, and/or independence in public decision making (Wæraas and Maor 2014; Carpenter and Krause 2012; Gilad 2009; Lee and Whitford 2013,Gilad, Maor, and Bloom 2015).

Over the last three decades, managerial changes in the public sector labeled as New Public Management have focused on introducing more efficient and effective public services (Hood 2010). One of the baseline consequences is that the organization of public services has increasingly been structured through networks of agencies where higher levels of agency delineation and independence are considered a solution towards greater efficiency (Hood 2010). Concretely, more autonomy for separate agencies—also referred to as agencification (Moynihan 2006)—can lead to a better use of resources, which in turn leads to more efficiency of these agencies (Lindlbauer, Winter, and Schreyögg 2016). However, to be effective in terms of achieving public goals, agencies often need to collaborate and interact with other agencies as well as with various stakeholders. They can do so as parts of public governance networks, in which actions of agencies are coordinated in such a way that their accumulated outputs match well for achieving overall public (service) goals, such as good education, public health, overall citizens’ safety and wellbeing, or economic prosperity. This logic of differentiation into more homogeneous units (i.e., agencies in the public context) to increase efficiency, in combination with integration (i.e., public governance networks in the public context) to increase overall effectiveness, is rooted in the seminal work of Lawrence and Lorsch (1967). For the study of BRT, the consequence of this increased agency delineation and organizational independence is that these public-serving organizations are also more identifiable as separate units (Moynihan 2006; Verhoest 2013), which has resulted in developing their own reputations vis-a-vis citizens, government, other agencies, policy makers, nonprofit organizations, and other important stakeholder groups in the public sphere. As such, bureaucratic reputations are an important asset of agencies in the public governance networks to which they belong.

INDIVIDUAL REPUTATIONAL BELIEFS AS MICRO-LEVEL FOUNDATION OF BUREAUCRATIC REPUTATION

Throughout this article, we use the term bureaucratic reputation referring to the original definition as “a set of symbolic beliefs about the unique or separable capacities, intentions, roles, obligations, history and mission of an organization that are embedded in a network of multiple audiences” (Carpenter 2001, 33). Building on this, we stress the assumption that bureaucratic reputation is an organization-level attribute of an agency, and it is derived from the shared beliefs of groups of individuals, or “audiences” of the agency. This is consistent with the broader literature on organizational reputation and the socio-cognitive perspective on reputation, where reputation is defined as the accumulated perceptions that stakeholders have about an organization’s capacity to create value (Rindova et al. 2005; Rindova, Pollock, and Hayward 2006).

The socioeconomic perspective explicitly focuses on the fact that individual beliefs about an organization are at the basis of reputation, but that individual beliefs of different individuals overlap with each other because of social interactions. The shared beliefs—also referred to as shared cognition—among individuals about an organization constitute the organization’s reputation, making it an emergent and social constructionist concept (Deephouse and Carter 2005; Lange, Lee, and Dai 2011; Petkova 2014; Petkova, Rindova, and Gupta 2013; Rindova et al. 2005). Moreover, these beliefs can have different components, including moral, rational, and/or emotional aspects (Pollock et al. 2019). This means that individual beliefs are at the basis of various types of behavior and interactions with an organization, including rational and well-reflected (re)actions, as well as emotional and more tacit (re)actions (Pollock et al. 2019).

We build on this socio-cognitive perspective of reputation to additionally define the concept of (individual) reputational beliefs, following our aim to formulate concrete micro-level extensions to BRT. Reputational beliefs are the perceptions individuals have about agencies with respect to the extent that the agency is performing well in achieving its goals. Hence, and in contrast to bureaucratic reputation, reputational beliefs are not an organization-level attribute of an agency, but an individual-level concept related to the individuals in an agency’s audience. However, individual beliefs can be shared within particular groups of individuals, such as because of similar interactions with agencies and/or through social interactions with each other. Hence, the shared beliefs among groups of people (which distinguish reputation as a shared construct from mere individual perceptions) is the result of social behavior. This theoretical approach acknowledges the social aspect of reputation building, meaning that perceptions are mutually adjusted between people (Lange, Lee, and Dai 2011; Rindova, Pollock, and Hayward 2006).

Figure 1 gives a visual clarification on how shared individual-level reputational beliefs constitute organization-level bureaucratic reputation. Concretely, the extent to which reputational beliefs of individuals overlap determines if an agency’s bureaucratic reputation is uniform or not. Uniformity of an agency’s reputation is high when the shared component of individual reputational beliefs within a group of individuals is large or when individual reputational beliefs are strongly similar to each other.

To understand and study the reputation of agencies, we need to adopt a dialogic perspective that includes agencies and their audiences, thus combining macro- and micro-level understandings of reputation (Bitektine et al. 2020). Yet, theorizing on the micro-level perspective on bureaucratic reputation is still scarce, although the role individuals have as an audience of public institutions is significantly different from the role they have as customers or employees of private firms. To show the role of individuals from various audiences and also how they interact with each other within and across audiences, we provide various examples throughout the article, such as focusing on citizens in general and/or on beneficiaries of particular public services along with employees working in agencies. In doing so, we want (1) to show the broad potential of micro-level extensions across various audiences, and (2) to stress the importance of cross-audience interactions for relevant micro-level extensions. In contrast, we also acknowledge that for each of these specific audiences, more specific background literature is available from various fields and disciplines, and that more audience-specific elaborations to our micro-level elaborations are possible.

In terms of new theoretical propositions, we offer three themes of micro-level reputational beliefs in the context of BRT, that is: (1) distinct information sources for the formation and updating of reputational beliefs, (2) the episodic nature of agency-audience interactions, and (3) the reputation spillovers between structurally related units. We elaborate these themes by presenting 10 concrete theoretical propositions. While we acknowledge we could have focused on less propositions but with more framing of each theme or proposition in existing literature, or in contrast, we could have provided a broader catalogue-like range of more themes and propositions for other scholars to explore. However, we decided on the selection of these three themes (and 10 propositions) because they are intertwined in terms of (1) the theoretical elements on which we rely (mainly the socio-cognitive perspective on reputation), as well as (2) the concepts on which we elaborate and on which the propositions are focused. These are summarized in the section “Concluding Thoughts” and in table 1.

MICRO-LEVEL EXTENSION 1: DISTINCT INFORMATION SOURCES FOR THE FORMATION AND UPDATING OF REPUTATIONAL BELIEFS

The first micro-level extension builds on the notion of different, but networked audiences of a focal agency. This element makes a distinction between at least two different types of sources for the formation and updating of reputational beliefs: (1) information from direct interaction with the agency, and (2) information from social network interactions with other agency’s stakeholders. Distinguishing sources of information on which audiences rely for their reputational beliefs is both theoretically and practically relevant because it enables us to understand the different dynamics in reputation formation (Etter, Ravasi, and Colleoni 2019; Valant and Weixler 2022). Herein, we elaborate that different sources of information about an agency can complement, but also contradict each other. A more profound understanding of these interacting dynamics can also prepare public managers better for anticipating—or even steering—their agency’s bureaucratic reputation.

As individuals, people act in the role of a particular audience. For example, a person can interact with an agency as a public service-using citizen, as an employee of another agency relying on the agency’s output for her own daily work, or as a politician evaluating the agency’s performance for deciding on the public funding of the agency. Then, individuals can update beliefs about an agency based on information they obtain from direct interactions with that agency.

In addition, new observations and information about an agency can also come indirectly from the social networks to which an individual belongs (Maor 2016). Through social exchange and communication with other individuals on various aspects of an agency, individuals can also update their beliefs about the agency in an indirect way. Indirectly implies that an individual might not have (had) direct interactions with an agency but might rely on information from other individuals when forming and updating reputational beliefs about the agency. That is, before a first interaction with an agency (e.g., before applying at a new agency as an employee, or before a first-time interaction with the customs or asylum office upon arrival in a new country), individuals might already have received information through other stakeholders. On the one hand, information can come from within-audience social interactions, where individuals rely on information from social interactions with individuals with similar needs and preferences in relation to the agency, and who might have had earlier experiences with the agency. On the other hand, information can come from cross-audience social interactions, where individuals rely on information from individuals from other audiences. For example, an individual might rely on information from third-party sources such as newspaper articles written by journalists (e.g., Boon et al. 2019), (nonprofit) rating platforms that operate as a watchdog for the agency’s performance and practices (Aleksovska, Schillemans, and Grimmelikhuijsen 2022; Szper and Prakash 2011; Willems et al. 2017; Willems, Waldner, and Ronquillo 2019), consumer or patients organizations (Carpenter 2002; Joosen 2021), or ombudsman services in the financial sector (Gilad 2009). Figure 2 illustrates these different information sources (direct interactions, within-audience social interactions, and cross-audience social interactions). With our first proposition, we thus make a theoretical distinction (i.e., a dual-source perspective) between information from direct interactions with the agency, and information from social network interactions with other agency’s stakeholders.

Proposition 1: Reputational beliefs about agencies are formed and updated based on information from direct interaction with the agency, as well as from social network interactions with other stakeholders (from the same and other audiences).

In addition, it is important to note that different sources of information for forming and updating reputational beliefs about an agency can reinforce, complement, or contradict each other. For instance, an individual might experience the agency to be client-oriented in direct interaction, while it might also receive this information through social interactions with other individuals. At the same time, a nonprofit organization or rating platform might provide additional information, such as that the agency is very efficient, and social interaction with a client organization might even reveal that there have been many complaints about a lack of client orientation.

The theoretical distinction (i.e., a dual-source perspective) between information from direct interactions with the agency and information from social network interactions with other agency’s stakeholders is particularly important for public-serving organizations, because their goals are often multidimensional and different aspects of their goals are differently relevant for their stakeholder groups. Moreover, the performance (i.e., goal achievement) of an agency might be hard to assess, because the performance itself is oftentimes hard to quantify and/or the unique contribution of a single agency in a public governance network cannot be pinpointed. Due to this complexity, individual stakeholders reckon that even when direct interactions with the agency take place, it might be hard—if not impossible—for them to assess what objectively good (stakeholder-specific and overall) performance would be in terms of an agency achieving its goals. Hence, higher perceived ambiguity in terms of an agency’s goals and the extent that the achievement of these goals can be assessed, increasingly requires extra information for individuals to form and update reputational beliefs. In such cases, individuals probably rely more strongly on their social network and indirect information from within-audience and cross-audience social interactions, in addition to information from direct interactions with the agency. Therefore, we extend Proposition 1 by identifying (1) perceived goal ambiguity and (2) perceived ambiguity about an agency achieving its goals (i.e., ambiguity on agency performance) as discriminating factors for the extent that stakeholders rely relatively more or less on one source type in the formation of their reputational beliefs.

Proposition 2: Higher levels of perceived ambiguity about an agency’s goals (e.g., in terms of multi-dimensionality) relate to stronger reliance on information from social network interactions (with individuals from same and other audiences)—relative to information from direct interactions with an agency—to form and update reputational beliefs about the agency.

Proposition 3: Higher levels of perceived ambiguity about an agency’s goal achievement (e.g., in terms of complexity to assess agency performance) relate to stronger reliance on information from social network interactions (with individuals from same and other audiences)—relative to information from direct interactions with an agency —to form and update reputational beliefs about the agency.

Another aspect is that individuals in distinct audiences might have—due to the inherently different agency tasks they are involved in or impacted by—varying levels of direct interaction (Verhoest et al. 2023). For example, all citizens might be highly impacted by the activities of an agency responsible for public health, but most citizens might not have direct interactions themselves with that agency. In this case, citizens form an important audience of the agency, but as an audience, most individuals in this stakeholder group have no direct interaction with the agency. Similarly, all citizens are influenced by the actions of a country’s national bank or service regulator, but most citizens have no direct interactions with such agencies, and as an audience they are also not expected to have recurrent direct interactions with these agencies (Thiemann 2019). Nevertheless, the overall public opinion on the performance of such agencies is often highly relevant in terms of political autonomy and discretion (Leidorf-Tidå 2022). Hence, in such cases, reputational beliefs are strongly or even entirely dependent on information from social network interactions with stakeholders from very different audiences, such as media, other agencies, (watchdog) nonprofit organizations, ombudsman services, or politicians (Boon et al. 2019; Etienne 2015; Gilad 2009). Hence, Proposition 4 extends the previous propositions by introducing the frequencies of direct interactions as a determining factor for the relative importance of information sources to update individual reputational beliefs.

Proposition 4: When direct interactions with the agency are scarce, reputational beliefs about agencies are relatively more strongly formed and updated based on information from social network interactions with other stakeholders from other audiences—(1) relative to information from direct interactions with the agency, and (2) relative to information from social network interactions with other stakeholders from the same audience.

Potential Avenues for Further Research Related to this Micro-Extension

These propositions can be further elaborated in terms of the factors that determine the relative importance of information from the same and other audiences. For example, one individual might not have (recurrent) interactions with an agency, while other individuals from the same audience (i.e., with similar needs and preferences towards the agency) might have information from direct interactions with the agency. The individual might thus rely in such a case on information from social interactions with individuals from the same audience as well as from other audiences. However, because individuals within an audience are similar in needs and preferences, they are likely also similar in terms of frequency and type of interactions with the agency. The relevant information from individuals from the same audience might thus be constrained, in particular when it is a low- or indirect-interaction audience. Moreover, there might also be other reasons for (additionally) relying on information from other audiences, as these other audiences might be (perceived as) more specialized in evaluating the activities of an agency, in particular when it regards complex matters. For example, specialized journalists or researchers might be given relatively more weight due to their expert status, compared to less specialized stakeholders. Consequently, further research can also explore these propositions in terms of the interacting dynamics (e.g., complementation or contradiction) between information from direct interaction with the agency versus social network interactions with other stakeholders.

MICRO-LEVEL EXTENSION 2: EPISODIC AGENCY-AUDIENCE INTERACTIONS AND VARYING HETEROGENEITY OF INTERACTIONS

Lifestyles in modern societies have evolved substantially to a reflexive way of living (Hustinx et al. 2022; Hustinx and Lammertyn 2003), meaning that individuals are continuously confronted by a plethora of choices. This plethora of choices implies that individuals engage in a continued updating of beliefs, values, decisions, and behaviors to continuously adjust what they do to what they believe, and what they believe to what they do. For example, customers can continuously choose from various versions of similar products and services to best fit their personal needs and preferences. For employees, flexibility in career choices has increased with more international mobility, adjusted work-life balance programs, and life-long learning opportunities. For citizens, a larger range of citizen-participation channels has been created, directly (e.g., citizen-participation platforms) and indirectly (e.g., through the short news media coverage on which public decision-makers immediately react). As a consequence, an overabundance of options exists to socially contribute and show beliefs, from slacktivism in social media, over project-based engagement in one-time campaigns, to direct participation in decision making and in the co-production of public services (Irvin and Stansbury 2004; Jacobs and Kaufmann 2021; Schmidthuber et al. 2019; Schmidthuber, Stütz, and Hilgers 2019; Yackee 2015).

As the possibilities for interactions have steadily increased over recent decades, interactions between agencies and their various audiences have become increasing volatile or “episodic.”¹ For example, the modern citizen-bureaucracy relationship can be considered as a continued sequence of many micro decisions triggered by the agency, by the citizen in a direct interaction, and/or by an external contextual change (e.g., availability of new information in the social network, or a third-party providing additional information about an agency). These micro decisions by individuals are concretely about adjusting their concrete behavior based on existing reputational beliefs about that agency, altered by the permanent updating of those same reputational beliefs about the agency due to direct interactions and new information. Consequently, an episodic decision event can be defined as a situation in which a choice has to be made by an agency’s stakeholder. Such an event is called “episodic” because the relationship with a stakeholder can be told as a story with multiple subsequent episodes, where each episode builds on various elements of previous episodes. These episodic decisions are, for example, about the extent to which citizens collaborate, participate, co-produce, provide feedback, and/or engage towards other stakeholders in the agency-interaction process. For agency employees, episodic decision events concern adjusting work efforts, such as starting, continuing, or quitting to work for the agency (Boon, Wynen, and Kleizen 2021; Bustos 2022). Additionally, employees can also update their beliefs on the employer attractiveness of the agency and their willingness to recommend the agency to others (Gross, Ingerfurth, and Willems 2021). For agency leaders, episodic decision events include how and when to react to critique about (their agency’s) performance (Andersen et al. 2021; Maor 2016). Integrating this episodic perspective into BRT allows to better account for the frequency of recurrent interactions between agencies and stakeholders, and how reputational beliefs are influential for micro-level decisions, while in turn new information results in the updating of reputational beliefs.

Hence, the episodic approach can advance BRT in at least two ways. First, more explicit attention is given to context-specific decisions and how reputational beliefs matter for those one-by-one decisions. This opens up opportunities to study in a generalizable way a broad variety of potential effects of bureaucratic reputation, in addition to the (self-)perceived discretion, independence, and autonomy at the organizational level (Lee 2022; Lyon and Cameron 2004). For example, for (future) employees of an agency, its bureaucratic reputation can be an important factor to decide to start, continue, or stop working for the agency, and/or to recommend it to other (future) employees (Bustos 2022; Gross, Ingerfurth, and Willems 2021; Kolltveit, Karlsen, and Askim 2019). For regulators, watch-dog organizations, and journalists, the bureaucratic reputation of an agency can trigger decisions on using agencies as benchmarks and/or to initiate further scrutinization about the agency (Salomonsen, Boye, and Boon 2021; Carpenter 2002). For beneficiaries and clients of an agency, the agency’s overall bureaucratic reputation as well as their personal reputational beliefs can influence whether they would opt again for the agency’s services (in case a choice between agencies is possible). In sum, an extension of BRT to episodic micro-level decisions allows for identifying and testing a broader variety of reputation-related decisions and behaviors of various stakeholders, which in turn enhance the predictability of stakeholders and help agencies to develop and improve their activities, processes, and outcomes.

Second, with an episodic perspective, a more dynamic approach can be taken with respect to how bureaucratic reputations emerge and change through episodic sequences of actions and decisions from agency stakeholders (Carpenter and Krause 2012). For example, in recent years, a substantial body of literature in the context of behavioral public administration has focused on the decision level of citizen interactions with agencies (Lee 2022; Matthias Döring 2021; De Boer 2020). This fast-growing body of literature has been instrumental in fostering better understanding of how citizens as well as employees of agencies dynamically behave and interact under various contextual conditions (Bhanot and Linos 2020). However, the insights from these studies—which are often single decision studies—could be further validated and substantially elaborated by more explicitly building on the understanding that such decisions are part of episodic—thus long, dynamic, and iterative—processes between individuals and agencies (Bertelli and Riccucci 2022; Willems, Faulk, and Boenigk 2021). The starting point for this micro-level elaboration is the formal acknowledgement that there is a mutual influential process between reputational beliefs on the one hand, and the concrete decisions at multiple instances by stakeholders on the other hand.

In sum, for the first proposition in the second micro-level elaboration, we focus on the relationship between new relevant information, individual reputational beliefs, and concrete individual decisions regarding a focal agency as an origin of episodic decision-making.

Proposition 5: An episodic event with new information about an agency results in the updating of individual reputational beliefs, while in turn, reputational beliefs influence micro-level decisions of stakeholders about the focal agency. These can be decisions in the context of direct interactions with the focal agency or in social network interactions with other agency stakeholders.

This sequence at the origin of episodic decision-making is a basic element in a continuous and repetitive process that clarifies how individuals continuously update beliefs, which have themselves impact on the decisions stakeholders make in the context of the public agency. These decisions, in turn, can lead to access to new information, again causing updates in reputational beliefs. The basic notions of volatile interaction sequences between stakeholders and agencies, as well as episodic decision events are further clarified in figure 3.

Combining this longitudinal and decision-level perspective with the dual-source perspective elaborated in the previous section enables the theoretical elaboration between reputational beliefs at individual level and bureaucratic reputation at organizational level. Being confronted with new information, either through direct interactions with an agency or through social network interactions can thus lead to the individual updating of reputational beliefs. In doing so, an individual’s reputational beliefs can become more or less similar to the reputational beliefs of other individuals. When more similarity occurs, bureaucratic reputation becomes more uniform (i.e., individual reputational beliefs become more similar across individuals, see figure 1). In such cases, agencies (and their internal decision-makers) can perceive their own bureaucratic reputation as less “noisy,” meaning that individuals do not tend to deviate much from the shared reputational beliefs about the agency.

In addition, the level of audience homogeneity, in contrast to audience heterogeneity (Van Der Veer 2021), can potentially explain varying levels of similarity in individual reputational beliefs. As discussed in the macro-level foundations of BRT, audiences are defined by unifying the needs and preferences of individuals within a particular agency’s audience. When various individuals update their beliefs about an agency in a similar way—for example, because of similar direct experiences with the agency—those individuals have shared beliefs, in turn constituting the social constructionist element of an agency’s bureaucratic reputation (Petkova 2014; Rao 1994) (and also figure 1). Agencies can vary substantially in how standardized their interactions are with particular audiences, which results in variation in the information stakeholders receive from direct interactions (how similar or different that information is). Some public service processes can have very standardized processes for a particular audience. For example, when filing their taxes, citizens’ interactions with the tax agency can be considered more homogeneous compared to the multitude of different types of interactions a citizen can have with, for example, the police force, including various aspects such as passport controls, assistance in traffic, interrogations during investigations, and evacuations in cases of a crisis. When an agency’s processes are thus highly homogeneous across individuals, citizens likely create and update reputational beliefs in a very similar way, and bureaucratic reputations become relatively uniform. In contrast, a non-standardized multitude of different kinds of interactions with an agency likely reduce the similarity between reputational beliefs of individuals across agency audiences, but also within the same audience. In sum, we identify homogeneity in terms of direct interactions between stakeholders and an agency as a factor that determines similarity—or overlap—in individual reputational beliefs, which then also means that the aggregated bureaucratic reputation is more uniform within a particular audience.

Proposition 6: High levels of homogeneity in stakeholder interactions with an agency increase the similarity of reputational beliefs among individuals, in turn making the agency’s bureaucratic reputation more uniform.

Similarly for information from social interactions, homogeneity in social network interactions can also be assumed to increase the similarity in individuals’ reputational beliefs. Homogeneity in social network interactions can be understood as interactions with other individuals who have similar attributes, for example in terms of needs, preferences, background, and demographic variables. The broad theoretical and empirical literature on social network theory suggests a positive relationship between higher homogeneity in social network interactions and shared (reputational) beliefs (Desmidt, Meyfroodt, and George 2019; Elgin 2015; Su, Borah, and Xiao 2022). In homogenous social networks, social network interactions are instrumental to increasing the similarity of reputational beliefs within groups of individuals. This relationship can be expected based on the following mechanism that relates to the logic built on for the formulation of Proposition 4. Individual stakeholders can receive information from various other stakeholders that are either in the same audience or in other audiences. Stakeholders in the same audience have similar needs and preferences (i.e., homogeneous social network interactions). Therefore, the obtained information from such social interaction is likely also more relevant to update reputational beliefs in a relevant way for one’s own context. Moreover, individuals can assess this similarity of other stakeholders and consider that as an additional signal about the relevance of the information from similar stakeholders (Riche, Aubin, and Moyson 2021). Hence, updates to reputational beliefs might result in more similarity of beliefs with other stakeholders, because the information of similar others is more relevant for updating reputational beliefs, compared to information from less similar stakeholders (i.e., in heterogeneous social networks and/or from other audiences) (Ambuehl and Li 2018). In sum, updates to reputational beliefs might result in sharing more similar beliefs with other stakeholders, not only because of the relevance of information itself, but also because the information is given more weight as it comes from a source that is perceived as more relevant for one’s own situation.

Proposition 7: High levels of homogeneity in stakeholders’ social network interactions increase the similarity of reputational beliefs among individuals, in turn making the agency’s bureaucratic reputation more uniform.

In sum, Propositions 6 and 7 focus on the uniformity of an agency’s bureaucratic reputation resulting from how homogeneous interactions with audiences are, as well as how homogeneous the networks within and between audiences are. A more uniform bureaucratic reputation is helpful for agency managers in their reputation-related decision-making, because an understanding about the (relatively large) shared component of what bureaucratic reputation is (i.e., shared individual beliefs) covers more accurately what most individuals believe about the agency. As such, actions that impact the shared component of reputational beliefs are likely more accurate in anticipating and even predicting the behavior of most individual stakeholders.

Potential Avenues for Further Research Related to this Micro-Extension

As this micro-level extension focuses on individual decision events of stakeholders, these propositions can also be a stepping stone to further explore the particular roles of the concepts risk and reputational threat in BRT (Rimkutė 2018; Van Der Veer 2021). Across various contributions on BRT, risk—in terms of consequences from actions and information that can harm an agency’s bureaucratic reputation—is a recurrent concept to clarify and predict agencies’ actions (Wæraas and Byrkjeflot 2012), as well as actions of structurally related units (e.g., bureau chiefs in Carpenter (2001), or agency heads in Maor (2016)). The assessment and mitigation of risk thus play a role in the strategic actions of agencies that can affect their bureaucratic reputation (Maor and Sulitzeanu-Kenan 2013, 2015; Rimkutė 2018). These risks for agencies come concretely from potential reactions from individual stakeholders related to (1) existing negative reputational beliefs that can in turn result in stakeholder decisions with negative consequences for the agency and/or (2) newly available information from or about the agency, resulting in stakeholders negatively updating their reputational beliefs about the agency (Busuioc and Lodge 2017).

Evidence from various disciplines suggests that negative and positive updates to reputational beliefs are asymmetric, meaning that the opposite logic of how positive updates work does not match with how negative updates to reputational believes work, and vice versa (Boon, Wynen, and Verhoest 2023; Levine 2021; Suzuki et al. 2016). In behavioral sciences, and concretely in behavioral public administration, a multitude of examples have been reported in which different reactions are hypothesized and tested as a result of opposite types of information or consequences (Boon, Wynen, and Verhoest 2023; Kafle 2023; Shinohara 2023; Wæraas and Byrkjeflot 2012). In addition, Willems, Faulk, and Boenigk (2021) demonstrate that the mission valence of an agency—perceived by the agency’s audience—can aggravate the negative reaction to negative information about an agency, but also additionally augment the recovering of reputational beliefs as a result of subsequent positive signals about the agency. Further research on the asymmetry in terms of how positive versus negative updates to reputational beliefs happen can also answer the call in BRT literature for a more profound understanding of how reputation-related risk considerations determine agencies’ strategic actions (Salomonsen, Boye, and Boon 2021).

MICRO-LEVEL EXTENSION 3: AGENCY AND INDIVIDUAL STAKEHOLDER REPUTATION SPILLOVERS

The focal units of analysis in BRT have mainly been agencies (1) as a whole and (2) as stand-alone actors, meaning that within-agency distinctions and/or networks of multiple agencies have hardly been elaborated. Most studies discuss the formal and coordinated actions of these organizations as anticipation or reaction to the shared reputational beliefs of their audiences (Bach and Wegrich 2019; Boon et al. 2019; Van Der Veer 2021). However, agencies consist of individuals who are employed within these agencies, led by appointed officials. While external audiences might “view public agencies as being more unified than they actually are,” it is also acknowledged that “[n]ot everyone in a public agency is on the same page” (Carpenter and Krause 2012, 28). The agency as a whole and the group of employees or managers can be considered as structurally related units, meaning that (1) they can be the object of (different) reputational beliefs, but they are (2) related in such a way that one unit is a necessary component of, is defined by, or is managed by the other unit. Actions of managers and/or employees can thus influence the reputations of organizations, and vice versa (Capelos et al. 2016; Maggetti and Papadopoulos 2023; Maor 2016).

In addition, collaborating agencies within coordinated networks could also be considered as structurally related units, where each agency delivers a part of general public responsibilities in a public governance network (Bautista‐Beauchesne 2022; Busuioc 2015; Rimkutė 2020). Similarly, separate agencies as well as the network to which they belong can be the object of reputational beliefs. As an extension of figure 2, figure 4 provides an additional visual clarification about structurally related units in the context of BRT. Acknowledging and disentangling this aggregation of individual and coordinated behaviors enables the formulation of additional micro-level extensions that concern reputational beliefs on structurally related units (Andersen et al. 2021; Bustos 2021; Kolltveit, Karlsen, and Askim 2019). Herein we elaborate the different effect sizes of reputational spillover effects between structurally related units, in particular when information from structurally related units is (in)congruent with particular stakeholder expectations.

Disentangling the different focal units of bureaucratic reputation allows for answering the call to further elaborate the differences of internal versus external audiences (Bustos 2021). A detailed example of such disentangling is provided by Maor (2016), where he makes a distinction between an agency and the agency’s head. While actions and reputations of these two units are structurally related, an explicit distinction allows understanding of how individual actions and reputations might spillover in terms of organizational consequence, arguing that “[a]gency heads who enjoy a good reputation will tend to increase transparency especially regarding their active, bolder policy activities when compared to agency heads with a bad or indistinct reputation.” (Maor 2016, 86), or how the reputations of such different units might influence each other (e.g., “[…] the level of agency reputation will converge towards the level of the agency head’s reputation” (Maor 2016, 86)). For example, Dantes Cabral and colleagues (2022) find that symbolic representation in terms of an agency’s leader being relatively more representative for vulnerable citizens living in a Brazilian favela influences procedural justice expectations of local police (Dantas Cabral, Peci, and Van Ryzin 2022). Similarly, Döring and Willems (2021) hypothesize and test the effect of the overall bureaucratic reputation of the public sector on the perceived level of professionalism of public employees, and confirm indeed a weak but significant spillover effect (Döring and Willems 2021). For agencies structurally related through funding—where one agency funds the other—Willems and colleagues (2019) find that there is a reputational spillover effect of the funding organization on the funded organization (Willems, Waldner, and Vogel 2019). Hence, stakeholders distinguish different structurally related units in their interactions with public agencies, where one unit is partnering with, a part of, led by, or managing the other unit. As they are related units, reputational beliefs about units can potentially interact with reputational beliefs about another unit. We summarize this in the following proposition.

Proposition 8: Reputational beliefs about structurally related units within and/or related to public agencies can mutually influence each other (i.e., reputational spillover effects).

While Proposition 8 outlines the possibility of reputational spillover effects between structurally related units, the following propositions focus on potential factors that explain the possible size of such spillover effects. For some of these related units, it might be relatively less straightforward for stakeholders to distinguish between them. For example, “police officers”—as one type of unit—are quite iconic for what the “overall police force”—as another, related unit—does, how it operates, and to what extent it achieves its goals. Hence, there might be strong spillover effects, for example in terms of expectations about the overall police force, which then influences what people think about how single police officers should act, and in how far they can be considered as competent or professional (Dinhof and Willems 2023). In contrast, some units might have more distinct and uniquely defining characteristics. For example, doctors work in hospitals and provide a crucial part of what hospitals do. Hence, interactions with doctors can thus also influence what patients think about the hospital as a whole, and vice versa. Nevertheless, doctors also have quite distinct and uniquely defining characteristics, compared to other professional groups in the healthcare sector. Due to more clearly distinguishing characteristics from other related units, spillover effects might be less strong, compared to when less such uniquely characterizing features are visible or known to stakeholders. We summarize this:

Proposition 9: When audiences are aware of uniquely defining attributes of structurally related units, reputational spillovers are lower, compared to when the agency’s audience perceives the structural relatedness of these units as high.

By disentangling the actions and reputations of structurally related units, another important micro-level elaboration can be made in terms of (1) how bureaucratic reputation is shaped based on audiences’ expectations about overall agency goal achievement and (2) how (observed) individual employees’ and leaders’ behaviors are contributing to or conflicting with such organizational goal achievement. For example, when audiences obtain information on employees’ behavior that is congruent with the expected goals of the agency, audiences might update their reputational beliefs about both the agency and its employees in such a way that both structurally related units are perceived as congruent. However, when information on employee behavior would not conform with the expectations about the agency’s goals, reputational beliefs about both the agency and the employees will be updated, but potentially not in a congruent way. For this, we can consider the following example: One citizen might have strict expectations about an agency to apply high standards in documenting and controlling various administrative steps in a public service encounter (e.g., for granting citizenship, or providing a passport). When this citizen obtains information about a first public servant thoughtfully considering and controlling every step in the administrative process, the citizen might update reputational beliefs (1) towards good agency reputation in terms of achieving expected agency goals, and (2) that the concrete observed employee’s behavior is supportive towards these goals (congruent).

Mutually reinforcing reputational spillovers can thus occur between structurally related units. In contrast, information about a second public servant being less thorough—for example, to facilitate a quick process for a particular client—might then lead to incongruent information, with more divergent updating of reputational beliefs on the agency on the one hand and its employee on the other hand; for example, leading to updating beliefs about public servants being lazy or corrupt (De Boer 2020; Willems 2020b). However, another citizen might value efficient and quick service processes as an agency goal more. For this other citizen, the behavior of the first public servant might result in congruent belief updates on the agency being rigid or bureaucratic (this time, focused on the negative connotation of the word, involving red tape and administrative burden) and the public servant being slow, while the second public servant would be considered dedicated to help clients, despite the rigid agency structures (Willems 2020b). Consequently, one can assert that stakeholder expectations and preferences on agency goals moderate spillover effects between reputational beliefs of structurally related units.

This example focuses on agencies and their employees as structurally related units on which reputational beliefs can converge or diverge. However, similar logics can be developed for related units such as agency heads, political decision-makers (Bertelli and Busuioc 2021), or other internal stakeholder groups (e.g., volunteers in publicly funding nonprofit organizations, or highly involved co-creating/participating citizens in citizen-participation projects). We summarize our theoretical reasoning in the following proposition:

Proposition 10: Differences in personal preferences and needs (in terms of expected agency goals) explain whether new information on structurally related units leads to congruent or incongruent updating of reputational beliefs about those units.

Potential Avenues for Further Research Related to this Micro-Extension

This micro-level elaboration of BRT can additionally benefit from integration with the growing body of literature on (public) employee stereotypes, as such stereotypes are a particular type of shared reputational beliefs about public servants (Bertram, Bouwman, and Tummers 2022; Dinhof and Willems 2023; N. De Boer 2020; Willems 2020b). The concepts of stereotypical groups of individuals and (bureaucratic) reputation are conceptually related—or even overlapping to some extent—as they both focus on (1) subjective perceptions of multiple stakeholders, (2) where the level of similarity in individual beliefs is relevant to understand socio-political dynamics. Moreover, both reputations and stereotypes (3) are shaped by direct and indirect social interactions, and (4) can influence concrete decisions about interaction with those units (Dinhof et al. 2023). As a result—and because organizations and their employees are structurally related units—spillover effects can further be explored in relation to reputational beliefs about various professions in the public sector and various types of agencies, or the public sector as a whole(Döring and Willems 2021). In turn, stereotypes about public servants can potentially also influence reputational aspects of public agencies, such as employer attractiveness, political (in)dependence of decision-making, or overall public service quality. In line with our framing of the propositions within the broader range of public-serving organizations—that is, also including quasi-public organizations with similar reputational challenges—such mutual spillover effects can also be studied for employees in state-owned enterprises and/or for employees in highly-subsidized nonprofit organizations.

CONCLUSION

In this article, we call for a micro-level elaboration of BRT. We have identified a first set of three themes in which we develop theoretical propositions for further elaboration and verification. For each theme, we have postulated some avenues for further research, while we also see the potential for (1) further audience-specific elaborations of each theme and proposition, and for (2) other themes in terms of micro-level extensions to BRT. We have introduced a set of concepts and distinctions that allow a more elaborate understanding about strategic and political actions of agencies as well as of (individuals belonging to) an agency’s audience. Table 1 summarizes these newly introduced concepts with a definitional explanation and examples.

By use of the introduced concepts, one aim of this article is to theorize on the socio-cognitive processes of an agency’s audiences. This includes external audiences such as citizens, other agencies, journalists, and politicians, but also internal audiences such as agency leaders and employees. We call for studying bureaucratic reputation increasingly in a dialogic approach (Bach et al. 2019; Gilad, Maor, and Bloom 2015), where reputation can be considered as a currency between an agency and its audiences, and where both the agency and the audiences are active actors who adjust their decisions and behaviors as a result of the agency’s bureaucratic reputation, but also to influence that reputation. For these concepts we introduced and/or elaborated, we have relied on a selection of combined literature, which has been instrumental to understand the main reasoning to derive each of the propositions. However, we also acknowledge that our goal has not been an exhaustive literature review on BRT in general, nor for each of these concepts or themes we introduce or elaborate in this article. Further research can thus complement our initial efforts by elaborating—for example—propositions within the context of a specific audience and/or by providing a more exhaustive and structured literature overview on these concepts and themes.

The concepts of the episodic decision event, reputational beliefs at the individual level, and the shared part of these beliefs with other individuals enable the building of the conceptual link between bureaucratic reputation as an organizational concept and the actual behaviors of the agency’s audiences. In particular, centering micro-level elaborations on episodic decision events, embedded in a series of reputation-based decisions and updates, provides an opportunity to better acknowledge the active role of audiences in the dynamic formation and updating of bureaucratic reputations. For example, in recent years, survey scales have been proposed and validated to quantify individual reputational beliefs of agencies and their structurally related units (Andreeva and Willems 2024; Lee and Van Ryzin 2019; Overman, Busuioc, and Wood 2020; Pedersen and Salomonsen 2023; Willems 2020a). These contributions are instrumental for further empirical validation of various micro-level theoretical extensions to BRT.

Finally, a more formal distinction between structurally related units enables scholars to better understand how updates to reputational beliefs are attributed to an agency itself, while updates on related units can be convergent or divergent with the updates about the agency itself. Hence, this distinction enables matching specific reputations and reputational beliefs to distinct units. Moreover, it also allows for elaboration on how reputations of related but distinct units interact with each other (e.g., reputational spillover effects). In sum, with our propositions we have put forward theoretical expectations on how reputational beliefs matter for the interactions between agencies and their stakeholders.

FUNDING

This research was funded by the Austrian Science Fund (FWF) [P36098-G].

Footnotes

REFERENCES