1. Introduction

Local participatory processes that enable citizens and civil society organizations to directly express their demands to municipal authorities are justified normatively as a way of enhancing democratic legitimacy in political decision making. However, municipal authorities can undermine the legitimacy of participatory processes in many ways, for example, by limiting the agenda or selectively adopting proposals (Font et al. 2018; Newman et al. 2004). One set of practices that can affect legitimacy, but has received surprisingly little attention, is the manner in which experts who participate in the process apply their policy expertise in judging which of the proposals can be justified on legal, economic, technical or other grounds. In normative terms a significant difference exists, for example, between those participatory processes where such judgements are made in a way that is open to participants and where such judgements are made behind closed doors, after the initiative has ended and where there is no public scrutiny.

The application of specialized knowledge resonates with concerns within the extensive theoretical discussions of the democratization of expertise (Bäckstrand 2003; Brown 2009, 2014; Collins 2014; Edelenbos et al. 2011; Fischer 1993; Forester 1999, 2009; May et al. 2016; Nez 2011; Petts 1997; Sintomer 2008; Yearley 2000). However, within that body of work there is a remarkable lack of attention to the specific ways in which different participatory processes at a local level deal with expert knowledge in practice: are the assessments on the feasibility of citizen proposals incorporated in a transparent or obscure manner? A fundamental aspect of democratic legitimacy in these processes is the extent to which they allow the participants to oversee and, if necessary, challenge what we call ‘expert assessments’ so that the final proposals emerging from the process still reflect their genuine will.

Participatory processes vary greatly in their aims, scope, format, sponsors, and institutionalization, among other characteristics (Fung 2003, 2006; Nabatchi 2012; Smith 2009). Thus, for example, participatory budgeting generally integrates expert assessments in specific stages during the process where bureaucrats, consultants, citizens, and associations exchange information, arguments, and criticisms (Baiocchi & Ganuza 2014). Other participatory mechanisms follow different logics, from cases where expert considerations play a decisive role to cases where they are absent or ignored.

To be clear, our interest in this paper is not in expertise in general, but in a particular form of expert knowledge: the inputs of officers and/or external consultants who are empowered to evaluate the feasibility of proposals that emerge from local participatory processes aimed at influencing public policies. These actors often hold a privileged position of epistemic authority in local participatory processes, based on their professional training and credibility. Their judgements can determine whether a proposal is selected, abandoned or substantially reframed in the final set of recommendations to municipal authorities. A crucial distinction thus emerges as to whether these assessments are carefully incorporated as part of the participatory design or whether there is no opportunity for participants to consider and challenge them.

The literature on participatory processes has typically focused on design concerns such as the type of participants or the rules for decision making. Less attention has been devoted to the timing, style, and specific impact of expert assessments. Thus, two research questions arise: Are the expert assessments of local officers and/or consultants on the proposals raised in local participatory processes incorporated in a transparent or obscure manner? Is it possible to develop a typology of processes regarding their management of expertise? We argue that the diverse modes of incorporation of specialized knowledge have relevant consequences for the democratic legitimacy of participatory processes, that is, whether or not the participants have oversight and, if necessary, are able to challenge the impact of these interventions on the final outcomes of the process. To address this question, our study draws on a mix of qualitative and quantitative data from a set of 38 local participatory experiences that were implemented in the Spanish regions of Andalusia, Catalonia, and Madrid between 2007 and 2011. The diversity of this sample – permanent and temporary, expensive and relatively costless, complex and simple initiatives – allows us to analyze how the type of participatory design conditions the way in which this form of expert knowledge is incorporated. Although the role of experts varies in accordance with the objectives of each process, the degree of transparency of their inputs on the proposals that emerge has significant impact on their democratic quality.

The paper begins by drawing insights from the theoretical debate on the democratization of expertise and the role of institutional design in participatory processes, clarifying the type of expert knowledge that is the object of our study. The next section describes the methods through which the Spanish cases were selected and analyzed. Third, we address our research questions, distinguishing three broad types of participatory process according to how expert assessments are integrated. We offer illustrative cases from our sample. Finally, we discuss the implications of the different ways of dealing with expertise for participatory governance.

2. Expertise and Participatory Processes

Public policies can be understood heuristically as a match between policy goals (the problem) and policy means (the tools to achieve goals). This match is performed across a techno-political process: on one hand, the technical dimension corresponds to the search for the optimal tools that help to resolve the problem; on the other, the political dimension emphasizes the disagreement between actors on the nature of the policy problem and potential impacts of the available tools (Howlett et al. 2009). According to this traditional policy-making perspective, these two dimensions are the domain of two different sets of actors. Elected representatives manage the political work to prioritize values and understandings of reality, drawing on social demands and their own beliefs. Bureaucrats, with their specialized knowledge and skills, seek the optimal match between policy goals and means.

However, the multiplication of partnerships, participatory processes and new forms of collaboration disrupts the traditional view of decision making not only in the more obvious political dimension, but also opening up technical aspects (May et al. 2016). As think tanks, interest and advocacy groups, citizen committees, lay citizens, and the like take a more direct role in the policy process, we witness the multiplication of ‘policy experts’ and the diversification of the location of expertise (Craft & Howlett 2013; Talpin 2011).

In this context, Edelenbos et al. (2011) usefully demarcates three broad forms of knowledge: scientific, bureaucratic, and stakeholder (also called alternatively lay, practical, or non-professional) knowledge. The most common distinction is drawn between scientific and lay knowledges: the former based on objective measurements and tools defined and reviewed by communities of peers, who share common theories and techniques (Maiello et al. 2013; Petts & Brooks 2006); the latter drawing on experiential knowledge, based on people’s common sense and everyday practices (Edelenbos et al. 2011; Fischer & Gottweis 2013; Nez 2011; Petts & Brooks 2006; Sintomer 2008). The literature on the democratization of expertise has focused specific attention on challenging the boundaries that have been drawn between these two sets of knowledge, in particular showing how citizen participation can contribute to scientific inquiry by complementing or correcting its universalistic principles with contextual information (Bäckstrand 2003; Yearley 2000).

A third type of knowledge that is distinguished from scientific and lay knowledge is that of bureaucratic actors. Such bureaucratic knowledge is related to understanding the appropriateness of ideas to solving policy problems with regard to administrative standards, procedures and rules (Edelenbos et al. 2011: 675). It is derived from a mix of knowledge about policy and administrative processes and particular professional and scientific training. Again, it would be a mistake to draw a tight distinction between bureaucratic and other forms of knowledge. As Hunt and Shackley stress, ‘contrary to the linear model of science policy relationships… the civil service is not solely a recipient or user of knowledge, but is actively engaged in constructing and formulating knowledge’ (1999: 144). Bureaucrats are also facilitators of public learning and empowerment (Forester 1999, 2009), combining insights from a variety of actors and perspectives with their own knowledge of procedures and professional expertise (Feldman & Khademian 2007).

According to Hendriks, in any given policy arena different actors promote their particular ‘participatory storylines’, that is, ‘narratives on who the public are, and how they should be involved in the policy issue’ (Hendriks 2005: 15). Hence, the prevailing storyline in a given matter legitimizes both the participatory mechanism and whether key actors are characterized as consumers, corporatist stakeholders, or lay citizens, for instance. The literature on democratization of the policy process suggests that a common storyline amongst bureaucrats is one of resistance towards the integration of lay knowledge in the policy process (Güemes & Resina 2019; Hendriks & Lees-Marshment 2019). Bureaucrats often perceive lay knowledge as unreliable with regard to problem solving. Citizens are taken to be incapable of dealing with complex policy issues either because of their lack of specialization and/or their partiality. In comparison, bureaucrats often identify themselves as offering an impartial perspective that privileges their assessment of appropriateness with regards to standards and rules (Ganuza et al. 2016; Maiello et al. 2013; Petts & Brooks 2006; Yang 2005). For sceptics, the democratization of expertise is often little more than a one-way relationship or a top-down approach through which bureaucratic knowledge is transferred to the public (Petts 1997). Citizens are seen as peripheral actors who lack the relevant knowledge to influence the technical dimension of the policy-making process: they provide inputs to bureaucrats who act automously on the final output (Migchelbrink & Van de Walle 2019).

Participatory processes are thus of particular interest as they are a political space in which a variety of expertise is potentially present (Brown 2008, 2014; Nez 2011; Sintomer 2008). In principle, the democratic legitimacy of such processes rests on the manner in which knowledge co-production takes place: the nature of the exchanges of facts, interpretations, assumptions, and causal relations from different knowledge domains (Edelenbos et al. 2011: 675–677). Here, the role of local officers and the policy professionals or external consultants that organize participatory processes becomes critical. These actors often share backgrounds, interests and prejudices (Edelenbos et al. 2011: 683) and, when designing participatory spaces, they have an implicit control over how scientific, lay, and bureaucratic knowledge is brought to bear (Bherer et al. 2017).

Among the diverse knowledges that usually converge in a participatory process, our specific interest in this paper is to pay attention to the application of expert assessments by local officers and/or external consultants, which can lead them to select, abandon, or reframe the participants’ proposals. Their professional training, bureaucratic knowledge and expertise potentially gives such actors a privileged position in shaping the final recommendations that emerge. In this sense, the consideration of ‘expertise’ in our research focuses on the very practical question of how a particular form of epistemic authority is exercised in a participatory process. In this inductive approach, we are interested in the function of the actor or actors with responsibility for, first, design and conduct the process and, second, assess the feasibility of proposals generated. Since a broad range of participatory spaces exist at the local level, these functions can be performed alternately or simultaneously by different figures. Thus, in our research, such expertise can be exercised by the local officer (bureaucrat, technician, etc.) who organizes the process, facilitates its meetings and synthetizes the final proposals, as well as, for example, the municipal architect who evaluates the feasibility of a sports center proposed in the participatory budget. Likewise, when the local authority outsources the process (commissioning a strategic plan from a consulting firm, for example), the participatory practitioners hired to carry it out exercise similar expertise. Hence, our aim is to understand how the presence and timing of these expert considerations on proposals affects the transparency of and democratic control over the process. We are concerned with the way expertise is integrated into the architecture of a participatory process.

This approach resonates with an influential perspective in research on participatory processes that attributes an explanatory role to institutional design (Bua & Escobar 2018; Delli Carpini et al. 2004; Fung 2003, 2006; Gastil & Levine 2005; Landwehr & Holzinger 2010; Nabatchi 2012; Nabatchi & Leighninger 2015; Smith 2009), whether in terms of how design characteristics affect the realization of democratic and institutional goods or particular outcomes. What is striking about this literature is that it has relatively little to say about our particular research question. Much of the work on process design has concerned itself with analysis of the effects of participants selection, the form of communication and the extent of influence on decision-making (Fung 2003, 2006; Johnson & Gastil 2015; Nabatchi 2012; Smith 2009). A paucity of attention is apparent in this literature on either the ways in which the design of participatory processes impact on when and how expert assessments are brought to bear, and the effect it can have on the democratic quality of these experiences. Fung (2003, 2006) and Smith (2009), two of the more influential authors on the democratic implications of design choices in participatory institutions, are rather limited in their consideration of these issues.

In distinguishing between the different types of participatory process – educative forums, participatory advisory panels and participatory problem-solving – Fung (2003: 340–341) implies that the diverse goals pursued by different participatory processes has an effect on the role played by expert assessments. But it is not a line of argument that he develops. In his later work on the ‘democracy cube’ (Fung 2006), experts and professionals are one potential group of participants but the way in which they interact with lay citizens remains underdeveloped, in particular how this impacts on the dimension of ‘communication and decision’. He claims that decisions in participatory processes are not always determined through deliberation or aggregation of citizen preferences, ‘but rather through the technical expertise of officials whose training and professional specialization suits them to solving particular problems’ (Fung 2006: 69). While the democracy cube recognises the significance of the integration of technical expertise to the design of participatory processes, it gives us little sense of the different ways that such expertise can be integrated in practice.

In Smith’s analytical framework for democratic innovations (2009), the role of expertise is recognised as a potential element in realising the democratic good of considered judgement. The effect expert knowledge can have on different designs – assembly-based, randomly selected mini-publics and direct legislation – is explored to some extent, but like Fung, insights on the relationship between participatory design and expertise are not developed systematically. Fung, Smith, and other authors are liable to the criticism that they tend to focus on exemplary participatory processes (Spada & Ryan 2017) where the democratic consequences of expert contributions have been carefully considered. This does not capture the way that more common and everyday local participatory processes incorporate specialized inputs.

To conclude, previous studies have recognized the significance of the exercise of expertise in participatory processes and how this can raise significant questions about the democratic legitimacy of citizen engagement in the policy process. What we are missing is any systematic understanding of how the design or architecture of participatory processes affects the way in which public officials or external consultants are able to exercise expertise and thus influence the proposals that emerge.

4. How Expertise Is Incorporated

We use participatory process as the unit of analysis to describe the prevailing dynamics of the integration of expert knowledge. From our sample, we can distinguish three broad approaches to integration. The first is where no expert considerations are introduced: in nine participatory processes participants suggested proposals without any specific assessment of the feasibility of their ideas. A second approach corresponds to 15 participatory processes where expertise was introduced in a transparent way during the process. As a result, participants were active in the weighting of these considerations. For the other 14 cases, expert assessments were incorporated in a way that was not accountable to citizens and associations, often once the engagement process had ended. In sum, most processes introduce expert knowledge in one way or another (29 out of 38) and almost half of them did so without direct oversight by participants, thus leaving room for discretion and opportunities to ‘cherry-pick’ or modify proposals.

In what ways are these strategies related to other design characteristics of the participatory processes? While recognizing that we are dealing with a small number of cases, Table 3 presents four process characteristics that have significant associations with these three broad approaches to expert integration. First, the type of participatory process. Participatory budgeting appears more often (47%) amongst accountable processes, whereas strategic planning is strikingly prevalent among those processes where the incorporation of expertise is not accountable to participants (86%). Second, the level of information provided to participants during the process also has a clear association: where participants are able to hold local officers or external consultants accountable, they will only rarely face this challenge with a low level of information⁶ (13%), whereas this situation appears in 78% of the processes where expert considerations were absent. Third, non-accountable processes tend to be supported by external funding from other supra-municipal administration (69%),⁷ a characteristic which appears rarely among processes where there are no expert assessments (11%). On the other hand, accountable incorporation of expertise appears almost equally among both externally and non-externally financed processes. Finally, the type of participants also exposes relevant differences. All the participatory processes characterized by non-accountable incorporation of expertise within our sample combine the presence of associations⁸ and individual citizens. On the other hand, designs with accountable expertise are more mixed in their composition: citizens as the only participants is more common than in other approaches (33%); but a mix of citizens and associations is the most common (53%). Processes without expert assessments tend to include only associations (44%) or citizens and associations (44%). Thus, who participates becomes an important design factor that appears to condition how expertise is incorporated: the challenge of combining individual and associative participation (different degrees of involvement, expertise, or perceived legitimacy, for example) seems to create the conditions for external expert involvement, but, in many cases, without any clear democratic scrutiny.

These associations between variables suggest that there are relevant relationships between aspects of process design and the different strategies for dealing with expert knowledge. To shed more light on these issues, the qualitative description of cases that follows provides a richer sense of the dynamics at play in the three broad strategies identified within our dataset of participatory processes.

5. Discussion

This paper has addressed the different ways in which participatory mechanisms deal with expert assessments on citizen proposals and whether this alters the democratic quality of the process. Our analysis offers an initial characterization of three broad types. The first is a set of quite simple participatory processes where there is no explicit incorporation of expertise. Such designs are quite common and tend to have low costs and no formal links to political decision making. Precisely because of their limited impact, external funding, and complexity, such designs do not receive much attention from the academic community. However, these processes are a relevant part of the reality of participatory experiences in the South of Europe, even if they are much less visible in research dominated by exemplary cases or more sophisticated experiences (Font et al. 2014; Spada & Ryan 2017). The absence of expert considerations does not necessarily undermine the quality of these processes, as in some cases the citizens’ lay knowledge is sufficient to make good judgements. Thus, an advisory panel such as the Catalan Council of Wise Women prioritizes the participants’ experiential knowledge over the application of expert criteria. That could partially explain why 78% of the processes that did not incorporate expert assessments also provided participants with low levels of information, assuming that they could rely on their lay knowledge.

Secondly, we find a set of participatory processes that incorporate expert assessments on proposals in a way that enables a degree of democratic control by participants. The most common type of such accountable processes is participatory budgeting, in which local officers and other policy practitioners participate in one or more steps of the process, but with oversight and the final decision remaining in the hands of participants. Our focus here is not on the systemic role of bureaucracy, but on its impact on particular sensitive stages of participatory processes. Local officers may well exert significant power in shaping the broader structure and reception of participatory processes but, in what we term ‘accountable process’, a moment exists within which participants are able to scrutinize and challenge expert interventions on proposals. This finding chimes with evidence on the translation of participatory budgeting from Brazil to Europe. While in many cases translation watered down elements of democratic control, forms of ‘empowered’ participatory budgets that retained elements of the original Brazilian design are more prevalent in Spain compared to other parts of Europe (Baiocchi & Ganuza 2014; Sintomer 2008).

The third approach is one in which experts typically play a critical but unobserved role in producing the final recommendations of the process. They filter, order, and modify citizen proposals – in other words, the process is open to ‘cherry-picking’ through the application of expert criteria. This occurs for many strategic planning processes in our study, where citizen engagement is often framed as a way of sourcing multiple ideas to be filtered at a later stage by local officers or external consultants hired by the local administration. While it was a consistent finding in our study, it is not necessarily the case for strategic planning in compared research, which makes necessary to further analyze the ‘participatory storylines’ taking place within this mechanism.

A key distinction that emerges from our study is whether or not the participatory design incorporates formal stages for holding experts to account, especially when there are several venues of participation in the same participatory process, what Spada and Allegretti (2017) call ‘multiple channels of engagement’. This is the case in participatory processes which are specifically designed to target different segments of the population. Here, different participatory settings are planned in the same decision-making process in order to reach different objectives. Although the diversity of participatory arrangements is often seen in the literature on democratic innovations as a positive way to enlarge citizen engagement (Bherer & Breux 2012), focusing on expert incorporation reveals the difficult challenge of dealing with multiple proposals coming from very different kinds of participatory settings (Spada & Allegretti 2017).

From the analysis of processes such as the Participation Plan in the big Andalusian city, it becomes difficult to ascertain whether an obscure style of expert assessment is a strategic attempt to reduce the direct influence of citizens or a pragmatic solution to combine the outputs from several participatory sources and methods. The sceptic would find evidence to bolster the broader critique of bureaucratic actors in the literature on the democratization of expertise that we discussed earlier in the paper. On this reading, many participatory processes are seen as tokenistic forms of co-option, with elitist manoeuvres undermining the will of participants in the interests of the administration (Arnstein 1969: 217; Bua & Escobar 2018: 132; Newman et al. 2004). In a significant number of participatory processes, ones that usually have external funding (69% of cases in our sample) and deal with more complex and costly issues, it is difficult to attribute the emerging proposals to the will of those participating. That participants do not seem to object or protest against this lack of transparency may be evidence of fairly low expectations or emotional commitment to a process in which they are peripheral actors, or an acceptance that it is legitimate for decisions to be made elsewhere.

A more charitable reading is that the larger the scale of the participatory exercise, the less direct control participants are able to exercise (Dahl 1998: 110): the final intervention of bureaucrats or other policy practitioners is an efficient way of dealing with a large number of inputs. For example, in the Plan for Gender Equality described earlier, citizens’ preferences are derived from a range of sources, including questionnaires, workshops, interviews, and discussion groups. Such processes typically mix individual citizens and associations (in all the cases of our sample) and the public authority is faced with the complex task of synthesizing the diverse contributions of these actors. Here, the filtering of proposals is the only option for the local authority when the sensitive issue of citizen oversight has not been planned in advance. The counter-example from participatory budgeting shows how a process that engages large numbers of participants can be carefully designed to enrich citizen proposals with expert assessments in a transparent manner (Smith 2009: 176). These designs ensure the provision of higher levels of information (medium or high levels in 80% of cases in our sample) to participants. The implication is that a more knowledgeable citizenry will be more effective – both democratically and epistemically – in enacting oversight and their proposals will be in less need of unrestricted scrutiny by specialists.

What we find then is that the role and oversight of the application of expertise is a key dimension in institutional design and one that can have a profound effect on the democratic qualities of participatory processes. Our study shows that the transparency and accountability of expert interventions varies significantly – from those where the application of such expertise appears less relevant to those where design choices have meaningful effect on the capacity of participants to scrutinize and oversee expert interventions on their proposals. What is clear is that, resonating with Fung’s framework, this a design choice. A number of cases of participatory budgeting in our sample show that this can be done in a transparent manner. Compare this with the opacity of most of the strategic planning cases. The question we are left with for those where expert intervention is opaque is whether these are design choices that represent a lack of imagination or a strategic decision to avoid accountability.

6. Conclusions

The democratization of expertise has generated substantial normative attention but there is still a lack of empirical specificity when it comes to actual participatory processes. This paper argues that more careful attention is required to the manner in which participatory processes deal with the expert knowledge that impacts on the realization of the will of participants. It deserves as much attention as more standard design concerns such as the democratic impact of the mode of recruitment of participants, presence of facilitation or the form of decision-making, to mention a few characteristics. It is not enough simply to state whether or not expert assessments occur, rather the extent of democratic oversight of these interventions is a key dimension for understanding the democratic quality of participatory processes.

As is often the case in an exploratory research, we faced several challenges. Our categorisation of participatory forms in relation to expert intervention is potentially limited by the characteristics of the 38 cases analyzed. For instance, our sample did not include randomly selected mini-publics (Smith 2009). Such a design would fit within the ‘accountable’ category in the sense that experts are able to present to, and be questioned by, the selected citizens, but then the recommendations are left to the participants. While this participatory design is widely discussed in academic circles (Brown 2014; Fung 2007; Gastil & Richards 2013; Johnson & Gastil 2015), particularly amongst deliberative democrats, they are rarely organized in the Spanish context in comparison to other forms of participatory process (Alarcón & Font 2014). Our findings concerning participatory budgeting and strategic plans may thus be generalizable beyond the Spanish case, but the impact of expert interventions in other participatory designs and contexts requires further empirical research. Here, the study of Roberts et al. (2020) on the role of expert witnesses in citizen’s juries constitutes a relevant contribution.

We also decided to limit our analysis to a specific understanding of expert knowledge: the expertise of local officers and/or external consultants in assessing the feasibility of the proposals generated during a local participatory process. There are other types of participatory processes that are not included in our analysis in which having specialized knowledge (scientific, bureaucratic, or practical) is an explicit condition for being an active member of the process. This is often the case with advisory councils, in which most of their members are selected on account of their professional experience or because they represent important stakeholders that have relevant knowledge. Their sectorial scope – immigration, youth, health, economy, women – and the specialized profile of their members implies that there is no need for further external assessments.

Even with this caveat, our findings resonate with previous research that considers institutional design as a key explanatory factor in understanding the outcomes of participatory mechanisms. But our analysis suggests that there is work to be done, either by incorporating expertise within existing analytical and explanatory frameworks, or by developing new frameworks that deal more effectively with this design characteristic. Certainly we can see how the manner of integration of expertise could enrich Fung’s (2003, 2006) distinctions between the objectives of processes or Smith’s (2009) account of how democratic goods are realized. Furthermore, the way expert knowledge is incorporated can be seen as a relevant proxy to evaluate the manner in which citizen proposals are connected to formal decision making. Preliminary quantitative analyses of our sample show that incorporating expert considerations – accountable or not – into participatory processes seems to facilitate the acceptance of their final set of proposals on the part of local authorities. These are promising lines of inquiry for a more nunced account of how expertise is managed in participatory governance.

Competing Interests

The authors have no competing interests to declare.