In recent decades, state actors around the world have experimented with certain democratic innovations as participatory types of deliberative policy making. These arrangements of ‘collaborative governance’ may provide cooperation and even coproduction between public authorities and citizens (Bussu 2019). However, they are at the top-down end of a spectrum compared to other innovations with a more bottom-up orientation (Bussu & Fleuß 2023). They are initiated by state actors who typically ‘invite’ several dozen citizens to participate (Cornwall 2002). Within a predetermined design, selected citizens deliberate with decision makers on recommendations on a given issue. These efforts aim to increase both effectivity of a specific policy-making procedure and strengthen the legitimacy of the political system as a whole (Warren 2014). At the same time, state actors are reshaped in the process, especially the public administration which usually designs and conducts top-down democratic innovations.

In principle, these innovations are both deliberative and participatory. On the one hand, they emphasise public reasoning whereby actors give and accept arguments as equals (e.g., Cohen 1996). Unlike forms of deliberative elitism (Papadopoulos 2012), however, they couple this ideal with a participatory component. The deliberation is thus extended to include ‘lay’ people who are normally not involved in the policy-making process, although the degree of participation is generally limited to relatively small-scale forums.

While democratic innovations combine deliberative and participatory ideas, the literature strongly focuses on the deliberative dimension and understands it as their central feature (Hendriks 2019). In fact, the scholarship on democratic innovations, after having originally been enriched also by other schools of democracy such as participatory theory, is now dominated by Deliberative Democracy (Smith 2019).1 Providing a well-established definition within this strand, Graham Smith speaks of democratic innovations as ‘institutions that have been specifically designed to increase and deepen citizen participation in the political decision-making process’ (2009: 1; see also Elstub & Escobar 2019).

The ambiguity of this passive construction is striking because it is unspecified who actually ‘designs’ top-down democratic innovations. There exists of course important research on responsible actors such as politicians (Hendriks & Lees-Marshment 2019) and professional consultancies (Friess & Herff 2023). Apart from these actors, however, Deliberative Democracy tends to neglect one important group: the public administration (Blijleven et al. 2019; Bottin & Mazeaud 2023).

Granted, several studies within Deliberative Democracy have recognised the role of public administration for democratic innovations (e.g., Boswell & Corbett 2018; Dean 2023; Mendonça 2016). They thus echo long-established findings within the Public Administration (PA) scholarship that public administrators have a significant influence on whether and how such innovations come into being (Eckerd & Heidelberg 2020; Migchelbrink 2021). First, as (co-)organisational actors, they shape important design questions about how, through what and with whom the abstract idea of participatory deliberation is translated into practice. For example, who should be invited and what topic should the participants discuss? In addition, administrators are active as policy makers with considerable discretionary power. They not only conduct democratic innovations but are deeply involved in assessing and ‘translating’ public input for political compatibility.

Nonetheless, despite some efforts, public administration in Deliberative Democracy ‘still remains a border territory, for most a foreign allusion, which only a few intrepid explorers have visited’ (Dean 2023: 5). Calls for more research have thus been voiced to address the importance of public administration for democratic innovation (Blijleven et al. 2019; Bottin & Mazeaud 2023). But necessary is not only a quantitative more, but also a qualitative how because of two limitations. First, Deliberative Democracy rarely incorporates theoretical concepts from PA and organisational sociology which per se intensely deal with public administration. Trailblazers like Boswell et al. (2023), who apply ideas from PA about ‘robust governance’ to the embedding of national climate assemblies in political systems, are the laudable exception (see also Boswell 2016). Trans- and interdisciplinary approaches are not very common, at best.

In addition, Deliberative Democracy tends to suffer from a biased understanding of public administration and its attitude towards democratic innovations. Often, public administrators are lumped together as possessive and resistant, thus perceived to be a key barrier to meaningful public participation (e.g., Bussu et al. 2022: 16–17). However, PA repeatedly shows that this widespread negative image is at least simplistic, if not misleading (e.g., Liao & Ma 2019; Migchelbrink & Van de Walle 2022b). More generally, Deliberative Democracy knows little about the factors which underlie the attitudes of public administrators, regardless of their ‘pro’ or ‘contra’ character towards democratic innovations.

This article seeks to make a theoretical contribution by foregrounding organisational embeddedness in the study of democratic innovations conducted by public administration. My principal effort is not to pass normative judgements about administrators, but rather help to understand how the organisational environment influences their attitudes, in addition to other factors (section 2). To theorise this particular embeddedness, I draw on concepts by German systems theorist Niklas Luhmann (section 3). His organisational sociology has not been adopted in the literature on democratic innovations but, in my eyes, can tell us much about the views of public administrators in this context.

My empirical example cites the German Auswärtiges Amt to illustrate the conceptual framework shedding light on national diplomats’ attitudes towards participatory experiments in national foreign and security policy (section 4). My broader objective is to contribute to a Deliberative Democracy scholarship on democratic innovations which is more comfortable with incorporating insights from organisational research, thereby gaining a more nuanced understanding of the role of public administration.

Public administrators are oriented towards certain attitudes when fulfilling their functions. Very broadly, attitudes can be understood as behavioural dispositions (Fazio 1986; Fishbein & Ajzen 1975). They influence the understanding held by an actor of an object by structuring that actor’s perception. Indeed, PA has repeatedly found that their attitudes towards citizen participation2 shape the reaction patterns of public administrators when they come into contact with the question of whether and how these democratic innovations can, and should, be conducted and implemented (Ianniello et al. 2019; Lia & Schachter 2018; Yang & Pandey 2011). For Yang and Callahan (2007: 259) it is even the key variable: ‘The most important factor in citizen involvement decisions is the attitude public managers hold towards the value of participation’.

Concerning the factors which shape public administrators’ attitudes vis-à-vis citizen participation, PA usually distinguishes three categories: individual, process-related, and socio-historical conditions (Migchelbrink & Van de Walle 2022a).

Individual factors include psychological traits, experience, and demographic characteristics. For example, greater trust of public administrators in citizens in general (Campbell & Im 2016; Yang 2005) and specifically in their skills (Yang & Pandey 2011) correlates with affirmative attitudes towards citizen participation. Previous experience with citizen participation is also associated with more favourable stance (Yang 2005). Finally, demographic characteristics such as gender, age and education level have been identified as important factors (Liao & Schachter 2018; Yang & Callahan 2007).

Process-related design features such as the timing of citizen participation in the policy-making process or the combination of different participatory methods shape the attitudes of public administrators (Yang & Pandey 2011). Moreover, the perceived input legitimacy of citizen participation in terms of its size and demographics is correlated with a more positive positioning (Migchelbrink & Van de Walle 2020).

Lastly, the socio-historical context refers to the relevance of political-administrative structures, legal provisions, and external pressures. For example, population size, median income levels and the political culture of the constituency influence public administration’s stance towards citizen participation (Liao & Ma 2019; Liao & Zhang 2012; Neshkova & Guo 2018). Moreover, when political leaders support and/or demand citizen participation, public administrators have a more receptive view (Liao & Zhang 2012; Yang & Callahan 2007; Yang & Pandey 2011).

In sum, PA shows that the attitudes of public administrators towards citizen participation are influenced by a range of factors. However, the level that lies ‘between’ individual and systemic conditions has hardly been systemically explored: the organisation as the work context in which public administrators act. While PA generally recognises the importance of organisational factors, a coherent theoretical framework is lacking as to why and how organisational embeddedness plays a significant role (Åström 2020; Yang & Pandey 2011).

Why does the organisational context matter for public administrators’ attitudes towards democratic innovations? First, as public administrators they are inevitably members of an administrative organisation. Unlike ‘normal’ citizens who do not directly represent anyone in participatory processes, administrators understand themselves, and are understood, to act in the capacity and role as members of the public administration. However, to understand how this membership influences their attitudes, it is necessary to abstract from any specific public administration and look more generally at what an organisation is and how it functions.

To do so, I draw on Niklas Luhmanns’ organisational sociology (2018) situated in his broader systems theory (1995). However, before offering my conceptual reasoning, I should point out that systems theory may seem abstract and demanding. Its language and form can be irritating, also because it decidedly breaks away from conventional ideas. In return, however, systems theory rewards with a powerful approach to reconstruct the attitudes of administrators in an organisational environment.

Following Luhmann, I understand organisations as systems of recursive decision making. They are ‘made up of decisions, and capable of completing the decisions that make them up, through the decisions that make them up’ (Luhmann 2003: 32). In other words, an organisation connects its own decision to one of its own previous decisions, thereby reproducing itself. It therefore constitutes itself as a self-referential system which is operationally closed from the environment, including its own members.

Every decision sets certain conditions for future decisions. Taking a decision means absorbing contingency although the decision itself is contingent itself – after all, it could have been made differently – and thus requires further decisions. However, an organisation cannot leave the interconnection of its decisions to pure luck or arbitrariness. Instead, it establishes structures which provide internal order and orientation. These decision premises both create and restrict decisions (Luhmann 2018: 181–185). Without them no occasion for decision making would arise in the first place. At the same time, decision premises limit the scope conditions for a multitude of other decisions. Although they cannot determine future decisions, those later acts must refer to the established premises.

More generally, decision premises are structures of expectations. This is how an organisation and its members are interconnected. Members relate to their organisation in the form of being psychic systems. A psychic system is consciousness-based and reproduces itself through self-referential thoughts (Luhmann 1995: 261). Again, thought processes do not proceed randomly but are pre-structured through expectations. To become socially relevant, these expectations need to become reflexive: ‘Ego must be able to anticipate what alter anticipates of him to make his own anticipations and behaviour agree with alter’s anticipation’ (Luhmann 1995: 303). Only at this level of the ‘expectation of expectations’ (Luhmann 1995: 304–306) between different psychic systems can individual behaviour be integrated and stabilised in an organisational system of multiple members.

Although the organisation as a social system and its members as psychic systems operate differently, they are structurally coupled. The organisation certainly cannot decide what its members should think. Rather, by establishing a structure of expectations which are socially anticipated it provides for the psychic systems a stimulus which ‘if it does not impose, then at least suggests the next thought’ (Luhmann 1995: 265). When members of the same organisation interact, they already know to a considerable extent what the other side expects of them and orientate their behavioural possibilities accordingly. This structural coupling does not amount to a causal relationship but facilitates a socially condensed and therefore more likely selection of thoughts.

In this vein, attitudes of organisational members are communicated expectations, that is structured thoughts. Crucially, they are not ‘part’ of the individual psychic systems. Instead, attitudes are inter-socially constituted as they refer to reflexive expectations in an organisation. A member’s attitude towards democratic innovations is thus shaped by the expectations which exist in the organisational environment and by the member’s own perception of what attitudes towards democratic innovations are expected of them as member of the organisation. I understand these reflexive expectation structures in the form of decision-making premises as organisational embeddedness.

Decision premises can be differentiated into three types: programmes, communication channels and personnel (Luhmann 2003: 45–47; 2018: 210–272. Programmes concern what is traditionally known as ‘tasks’. They predefine expectations of ‘correct’ decision making. These rules for ‘right’ and ‘valid’ decisions come in two types: conditional programmes and purposive programmes. Conditional programmes set sequences of decision-making in a ‘if/then’ form. If, and usually only if, a certain trigger is perceived, decisions must be made in accordance with the programme’s regulation. Most routine decision making in an organisation is based on this logic. While conditional programming links conditions and effects, purposive programmes operate on means and ends. They define ends on which decisions on appropriate means are to be taken. The end itself may limit the choice of possible means but purposive programming tends to be more freewheeling. Both types of programmes can be nested within each other and sequentially ordered.

Communication channels refer to expectations within an organisation as to who communicates decisions with whom, why and about what. In an organisation not everyone talks to everyone about everything. Rather, the initiation, participation and acceptance of communications are regulated. Thus, communication channels structure who can (and ought to) communicate in which area of responsibility and who must adopt this information. The most prominent manifestation is the ‘official channel’ which distributes decision-making powers, typically as a hierarchical order. Formal decisions are recognised as organisational decisions only if they follow this communicative top-down premise.3 At the same, other channels based on informal expectations also influence how communications actually takes place and may differ significantly from the static official organisational chart.

Finally, personnel decisions also contribute to relatively stable expectations on how decision-making is made. Since individual members have relatively constant personality structures, specific personalised decision premises are established regarding specific members. One way personnel premises can emerge concerns decisions about which person – an individual psychic system of background, education, and preferences – can enter an organisation. For example, recruiting a lawyer instead of a geographer creates different expectations on what and how this person will likely decide in the future. Furthermore, an organisation may be engaged in pre- and post-recruitment socialisation processes which structure expectations of the members. Personnel decisions also concern redeployment and promotion which set certain expectations. If you know your new superior, you usually have a better idea of how this person will take what decision, which in turn may shape the way you interact.

These three decision-making premises are interrelated in many ways (Luhmann 2003: 47). They can support and compensate each other. An organisation can thus refer to personnel decisions if sufficiently detailed programmes are not possible. However, compatibility and mutual dependencies are limited. For example, a new communicative hierarchy level may be introduced, but it is hardly functional if appropriately qualified personnel cannot be recruited. More generally, the dynamics between the programming, communicative, and personnel decision premises enable an enormous diversity of organisational structures of expectations. It is precisely through this structural plurality that organisational embeddedness influences the attitudes of the members.

The preceding section argued why the organisational embeddedness influences public administrators’ attitudes. In the following, I illustrate these conceptual ideas by using the Auswärtiges Amt (AA), the foreign ministry of Germany, as an empirical example. This does not serve as a case study to systematically test the theoretical framework. Rather, I seek to substantiate the plausibility and applicability of the premises behind organisational embeddedness and how it can help understand the role of public administration vis-à-vis democratic innovations.

I am involved in a research project on democratic innovations in national foreign and security. This empirical field has received limited attention due to the heavy focus of existing literature on socio-political issues at the local level (Blijleven et al. 2019: 210). The empirical account draws on qualitative interviews I have conducted with diplomats specifically on their democratic innovations, as well as more general studies on these specific public administrators.

This article has encouraged the Deliberative Democracy scholarship on top-down democratic innovations to engage more with organisational research. Concepts and insights of Public Administration and organisational sociology contribute to a better understanding of the role of public administration. Drawing on Niklas Luhmann’s systems theory, I have argued that organisational embeddedness significantly shapes public administrators’ attitudes towards experiments of participatory and deliberative policy making. This embeddedness has been conceptualised as the decision premises encompassing programmes, communication channels, and personnel. Empirically, I have illustrated this theoretical framework with the democratic innovations of the Auswärtiges Amt (AA) in Germany. That and how novel participatory formats take shape in German foreign and security policy has been considerably influenced by how diplomats orient themselves to the three organisational decision premises.

While the empirical example has demonstrated the fruitful application of the theoretical argument, it also raises further questions to refine the framework of organisational embeddedness.4 First, the relations and dynamics between the decision premises need to be further explored. At the AA, certain premises such as the role of political leadership seem more decisive than others that diplomats are encouraged to initiate democratic innovations. This suggests that different premises exert varying influence, possibly depending on how established democratic innovations are in an organisation.

Similarly, further attention is warranted on the explanatory power of organisational factors on what kind of top-down democratic innovations are offered. In this article, I have largely focused on how this embeddedness shapes diplomats’ attitudes towards introducing any participatory format in the first place. By now, however, the AA has conducted several experiments with varying degrees of quantitative scope and qualitative sophistication. How do the different decision premises influence how democratic innovations are specifically designed?

A final, yet crucial future avenue involves gaining more insights into when and how organisational embeddedness changes, and with it, established attitudes of public administrators. At the very least, the example of the AA proves that even administrative organisations are not completely routinised machines, but ‘living’ and evolving systems. Their decision premises can and do change to create incentives to establish democratic innovations even in policy areas that, like foreign and security policy, were previously dominated by resistant attitudes of public administrators.

The concept of organisational embeddedness has the potential to propel us to a better understanding whether public administration implements democratic innovations and what kind. Organisational research offers a vast repertoire of knowledge into which Deliberative Democracy can, and indeed should, tap. Hopefully, these efforts will not only be one-way, but will help develop a long overdue dialogue between the disciplines to enrich each other.