Abstract
Despite lamentations to the contrary, expertise continues to play an important role in contemporary societies. In the context of pressing sustainability challenges, calls for so-called transformative change are emerging, with concomitant expectations that knowledge and expertise can contribute to such change. One area in which such expertise is being developed is in the context of alternative concepts of wellbeing to replace the dominance of gross domestic product (GDP), yielding an ongoing co-production of modes of expertise with modes of policymaking. In this paper, we explore this ongoing co-production in the context of what is called “Dutch Broad Wellbeing.” We introduce the concept of knowledge accountability to refer to what knowledge is mobilized and how it is expected to feature in policymaking. We identify two different logics of knowledge accountability, a technocratic and a political logic, and discuss how they interrelate with transformative change. We suggest that the analysis of logics of knowledge accountability not only supports understanding how knowledge features in policymaking but also underscores the need to make fit-for-purpose modes of expertise explicit. We contend that, for the benefit of both transformative change and democracy, these modes should yield expertise that politicizes.
Introduction
Here's the key point we need to grasp: GDP is not an arbitrary metric of economic performance. It's not as though it's some kind of mistake–an accounting error that just needs to be corrected. It was devised specifically in order to measure the welfare of capitalism. … Those who call for a shift towards well-being as the sole solution tend to miss this point. If we want to release our society from the grip of the growth imperative, we have to be smarter than that. ( Hickel 2020 , 204)
Despite lamentations to the contrary, expertise continues to play an important role in contemporary societies. Decision-makers draw on various forms of scientific and other knowledge, and a range of science-policy arrangements exists to develop knowledge that aims to enable more effective policymaking. Calls are mounting for “transformative change” to address current socio-ecological predicaments through the fundamental reorganization of institutions, paradigms and practices (e.g., Feola 2015; IPBES 2019; 2024; IPCC 2023). These calls often include the expectation that scientific expertise can contribute to policies enabling such change (Patterson et al. 2017; Brouwers et al. 2022; Fisher, Brondizio and Boyd 2022).
However, there is good reason to be skeptical about the potential of scientific expertise 1 to contribute to transformative change (Maas 2023; Turnhout 2024). As a large body of Science and Technology Studies (STS) scholarship has demonstrated, common assumptions about expertise, policymaking, and their relations are often not only overly simplistic, but can also result in knowledge practices that undermine fundamental public values (see e.g., Stirling 2016; Jasanoff and Simmet 2017; Chilvers and Kearnes 2020). This means that it is pertinent to consider how and under what conditions expertise can contribute to transformative change. We do this by exploring ongoing endeavors to broaden policy objectives and indicators beyond the current focus on economic growth. Particularly, we focus on attempts to replace Gross Domestic Product (GDP) as the dominant indicator to represent the state of an economy and society, with the notion of wellbeing. We present materials from such attempts in the Netherlands over the past decade to illustrate the multiple roles that expertise plays in science-policy arrangements for wellbeing, ranging from providing technical information, proposing indicators, to facilitating deliberative processes.
Our analysis is informed by STS studies that have foregrounded the performativity of expertise (e.g., Law and Urry 2004; MacKenzie, Muniesa and Siu 2007; Pallett and Chilvers 2015; Turnhout, Halffman and Tuinstra 2019). This literature has demonstrated the political consequences of expertise and highlighted how prevalent modes of organizing expertise can pre-empt political deliberation and thwart public values. This becomes especially problematic in the context of transformative change since the potential contribution of expertise to move beyond the status quo hinges on the explicit political deliberation of public values. We demonstrate this in our analysis by examining what we will conceptualize as different logics of knowledge accountability that shape how expertise is used in our case (cf. Barry, Born and Weszkalnys 2008), and by discussing their potential and limitations for furthering political deliberation and transformative change. In doing so, we also address the relation between expertise, politics and democracy, and take up calls for more explicit engagement with the notion of democracy in STS (e.g., De Vries 2007; Marres and Lezaun 2011; Brown 2015; Chilvers and Kearnes 2020; Tschersich and Kok 2022).
We begin by introducing the concept of knowledge accountability to analyze the political and performative effects of expertise, as well as the organization of expertise itself. This includes what is expected and permitted in the use and production of expertise. We use materials from the beyond-GDP debate in the Netherlands to further examine relations between expertise and politics, and discuss the implications of our analysis for the role of expertise in transformative change. We close by discussing how in its transformative role, expertise can also make a more fundamental contribution to democracy.
Logics of Knowledge Accountability
A wide and diverse body of research has examined the relations between economic concepts, expertise, and political decision-making in contemporary democracies (e.g., Hirschman and Berman 2014; Mügge 2016; Birch 2017). A key insight from this work is that modes of expertise and modes of policymaking co-produce each other and that, consequently, concepts and expertise are performative (Callon 1998; Mol 1999; Law and Urry 2004; MacKenzie, Muniesa and Siu 2007; Van Egmond and Zeiss 2010; Blok 2011; Asdal and Marres 2014). Performativity means that concepts and expertise have political effects through how they render some things visible and others invisible, some ideas reasonable and others unreasonable, and some actions feasible and others unfeasible. Moreover, this performativity extends to the organization of expertise-for-policymaking itself. Not only do different modes of expertise bring particular policies in and out of scope, different modes of policymaking also render particular forms of expertise as (ir)relevant, (un)important, or even (im)permissible (Callon 2007; MacKenzie 2009; Pallett and Chilvers 2015; Halffman and Pastoors 2019; Bacevic 2021; Turnhout 2024; Turnhout and Lynch 2024). Put differently, while different types of policy problems require different modes of expertise (Hisschemöller and Hoppe 1995; Bijker, Bal and Hendriks 2009), expectations and ideals of good expertise also fundamentally shape what good policymaking is supposed to look like. These expectations are institutionalized as concrete science-policy arrangements, including in formalized expert organizations (Weingart 1999; Halffman 2005; Borie et al. 2021).
In this paper we build on this insight to analyze not so much what politics are conducted by science-policy arrangements but rather how they conduct politics (Marres 2013). Crucially, we contend that different arrangements are not equally conducive to transformative change. As discussed by MacKenzie (2009) in the context of carbon markets and expertise, there is a common expectation that a strict boundary exists between science and politics to ensure that action is seen as based upon “sound science.” However, because institutionalized science-policy arrangements embed norms of what counts as “sound science” and what its “proper role” in policy should be, science-policy arrangements that are based on this expectation of a strict boundary in effect constrain the scope of possible action and pre-empt discussion about potentially desirable alternatives (Hajer and Oomen 2025). This dynamic of closing down has been criticized for its tendency to frame political decisions in terms of necessity rather than choice (Mouffe 2000; Stirling 2008; Swyngedouw 2011); a tendency reinforced by persistent norms of ostensible neutrality, relevance and separation in science-policy arrangements (Stirling 2016; Lahsen and Turnhout 2021; Maas 2023). Although arguably such norms have long been debunked in and around STS (e.g., Sarewitz 2004; Beck 2011; Owens 2015; Stirling 2016; Douglas and Branch 2024), they continue to hold sway in many science-policy arrangements (e.g., Kunseler 2016; Karhunmaa 2020; Maas, Pauwelussen and Turnhout 2022). 2
These exclusionary practices of science-policy arrangements are also highlighted by the field of ignorance studies, which posits that the production of expertise inevitably also includes the production of ignorance (Gross and McGoey 2015). This makes ignorance not simply a “precursor” or “impediment” to knowledge but a source of power in and of itself (McGoey 2012). The inevitability of ignorance makes it meaningless to oppose ignorance itself. At the same time, it highlights that it is naïve to assume knowledge practices can eliminate ignorance. Instead, ignorance studies spur critical reflexivity toward what knowledge production is done, what remains undone, and who benefits from and is harmed by the knowledges and ensuing ignorances produced (Frickel et al. 2010; McGoey 2020; Bacevic 2021; Turnhout and Lahsen 2022).
To critically interrogate these dynamics between knowledge and ignorance in expertise for policymaking, we propose the notion of knowledge accountability, which we define as the way in which decision-makers are expected to explain and justify how and what knowledge influenced their decision. In our use of the term accountability, we follow Bovens (2007, 107), who defines accountability as “a relationship between an actor and a forum in which the actor has an obligation to explain and to justify his or her conduct.” Thus, knowledge accountability is concerned with justifications about the role accorded to knowledge and expertise in these relationships.
A focus on knowledge accountability extends analysis of science-policy arrangements by asking how and what rather than merely whether knowledge influenced decision-making. Rather than focusing primarily on experts and their organizations, which risks essentializing particular understandings of concepts like legitimacy, effectiveness or authority (Mahony 2020; Bandola-Gill 2021), knowledge accountability considers these concepts as themselves performed by a wide range of policy and societal actors. A crucial corollary of knowledge accountability, then, is that it highlights how expectations of “what expertise can do” structure the ways in which decision-makers can take responsibility for what knowledge/ignorance is produced. In other words, to understand the role and shape that expertise can take, we need to acknowledge that expectations of what expertise can do and ought to do structure what kinds of expertise are considered relevant, and thus what expertise is requested or dismissed. Since such expectations can be diverse, we expect multiple logics of knowledge accountability to be possible that differ in terms of how they view what good expertise is and what good expertise should do (following Barry, Born and Weszkalnys 2008). Logics of knowledge accountability thus help to articulate various expectations about how expert organizations, decision-makers, and other actors may produce, use and engage with expertise. Specifically, we examine what logics of knowledge accountability can be identified, and to critically discuss their respective appropriateness for particular purposes, in our case transformative change.
From GDP to Dutch Broad Wellbeing
Discussions around GDP form an interesting backdrop for such insights on the politics of science-policy arrangements. Originally developed as a measure of national income, GDP has become largely equated with societal wellbeing: a high GDP is taken to mean a society is doing well (Hoekstra 2019). This equation of GDP with wellbeing has led to a “growth paradigm” where continued GDP growth is considered possible, necessary, and desirable (Schmelzer 2015). GDP has become locked into science-policy arrangements where policymakers focus on GDP growth and experts project policy effects on GDP (often using quantitative modelling) and provide policy advice on optimizing GDP growth.
At the same time, it has become abundantly clear that GDP has limitations to the extent that it is unsuitable to represent societal wellbeing (Stiglitz, Sen and Fitoussi 2009). For example, it disregards non-monetary and household labor, and it excludes the environmental and social costs of economic growth (Schmelzer 2022). Moreover, growing inequality has shown that GDP growth does not benefit everybody equally. In response to these critiques, initiatives that seek to account for “beyond GDP” in policymaking processes abound around the globe (Truijens and Georgieva 2021; Turnhout et al. 2021; Berger 2022). Such alternatives include non-Western concepts like Ecuador's Buen Vivir or Bhutan's Gross National Happiness, as well as concepts developed on the fringes of modern economics, e.g., “steady-state economics” (pioneered by Herman Daly), “donut economics” (Raworth 2017), and “degrowth” (Hickel 2020). Closer to mainstream economics, a major impetus to the development of alternative concepts of societal wellbeing was given by the Commission Stiglitz, Sen and Fitoussi (2009), instated by then-French President Nicolas Sarkozy. The shift from GDP to wellbeing is considered to have significant transformative potential in contributing to alternative paradigms of progress that are not based on the unlimited exploitation of natural resources and ever-growing production and consumption (e.g., United Nations 2023; IPBES 2024). Nonetheless, despite recognition of its limitations, progress in moving beyond GDP remains limited (Biermann et al. 2022; Masood 2022).
In the Netherlands, efforts to find an alternative to GDP center on the notion of “Broad Wellbeing.” 3 Beside growing references to Broad Wellbeing by political parties (e.g., Claassen and Cools 2023), several partially related initiatives have started to develop this idea. These include the Monitor of Wellbeing—an initiative by the national statistics agency that aims to report on the wellbeing of Dutch citizens “here and now,” “later,” and “elsewhere” (Statistics Netherlands 2022); the development of indicators for the governments’ budgetary cycle at the request of parliament (Tweede Kamer 2016; CPB, PBL and SCP 2022); its inclusion in an inter-administrative policy program (PBL 2022); its integration into the central government's regulatory impact assessment tool (Tweede Kamer 2022); as well as various experiments run by individual ministries integrating wellbeing into their way of working. And, as in the global debate, also in the Netherlands the expectation exists that Broad Wellbeing can contribute to transformative change. For instance, Kim Putters, the chair of the influential Dutch Social-Economic Council (SER), argues the concept holds the promise of realizing a more sustainable, just and inclusive economy and society (Putters 2024, 14).
These initiatives contain multiple and different interpretations of Broad Wellbeing, and come with different expectations of expertise. As such, it is a concept still unstable and in-the-making. This means that, despite the risks of lock-ins, there is potential for transforming science-policy arrangements. In this respect, the involvement of the three so-called “planning bureaus”—CPB, SCP and PBL 4 —in several of the initiatives is particularly interesting. These planning bureaus are independent governmental expert organizations for policy assessment, tasked with economic, social, and environmental policy analysis. They combine statutory tasks to report on given issues, research undertaken on their own initiative, and ad hoc tasks requested by government ministries or other organizations. Significantly, they are often considered to have a unique position in the Dutch science-policy landscape for their de facto role as political stabilizers (Halffman 2009), structuring what can be thought, said and done in political discourse (cf. Hajer and Versteeg 2005).
The unique position of the planning bureaus is observable in the analysis of election manifestoes performed by CPB and PBL, for instance, which are seen to “discipline” politicians to refrain from making policy proposals considered unrealistic or unconventional (Pesch, Huitema and Hisschemöller 2012; Bolhuis 2018). This disciplining by CPB and PBL has potentially problematic implications. As these analyses are largely model-based, policies that cannot be readily expressed in quantitative or monetary terms are less likely to be incorporated. Telling examples include government expenses for education or healthcare, which offer little bonus for a political party's “report grade” in the analysis (see e.g., Baazil 2024). Political parties also anticipate these models and their shortcomings, so their proposed policies are portrayed in a positive way. This role of the planning bureaus has been a regular point of public scrutiny, not least since former prominent member of parliament Pieter Omtzigt (2022) addressed these issues. Yet, while such scrutiny has prompted debate, existing lock-ins prove difficult to overcome and limit progress in transforming dominant research repertoires (see e.g., Hajer 2012; Kunseler 2017; Verwoerd 2022). The question is whether and how Broad Wellbeing can reconfigure existing science-policy arrangements and the position of the planning bureaus in these arrangements. In other words, heeding the lessons of knowledge accountability in the context of alternatives to GDP means interrogating how science-policy arrangements produce knowledge/ignorance, as well how this knowledge/ignorance ends up being employed to justify adopting or ruling out actions. 5
Logics of Knowledge Accountability in Dutch Broad Wellbeing
The unstable character of Broad Wellbeing presents an opportunity to explore its ongoing dynamics of knowledge accountability. We consider Broad Wellbeing both as an empirical illustration of the conceptual merits of knowledge accountability, and a suitable context to raise a wider discussion on the respective appropriateness of different logics of knowledge accountability in pursuit of transformative change (cf. Flyvbjerg 2006; Hitchings and Latham 2021). For this combined purpose, we present materials from 23 interviews with policymakers and researchers involved in different Broad Wellbeing initiatives, 6 in addition to examining relevant reports and policy documents (interviews are lettered and dated, documents referenced where directly relevant). Moreover, the first author of this article (Maas) is a full-time researcher at PBL, and while not directly involved with PBL's work on developing Broad Wellbeing approaches, he was simultaneously researching a PBL report comparing the concept to the Sustainable Development Goals (SDGs) (Maas and Lucas 2023).
Our analysis of Dutch initiatives led us to identify two distinct approaches to Broad Wellbeing. In the first approach, it is primarily considered an instrument for optimization, while in the second approach it is primarily considered an instrument for discussion. While the different initiatives show a prevalence for one of these two approaches, respondents also shared reflections that pointed to the other. Below, we present these two approaches and discuss what logics of knowledge accountability these approaches reflect. Recognizing that in-between forms and hybrid outcomes inevitably exist, we see these logics as forming a spectrum that allows for critical discussion of the implications of contrasting logics for the organization of expertise, science-policy arrangements, and ultimately, the transformative potential of different Broad Wellbeing approaches.
Broad Wellbeing as Instrument for Optimization
In the first approach, Broad Wellbeing is an instrument for optimization. In this understanding, Broad Wellbeing represents a challenge of finding the optimal combination of different aspects relevant for wellbeing beyond GDP. The challenge, thus, is to develop an approach which can measure and integrate these different aspects (succinctly: economic, social and environmental themes, across the dimensions “here and now,” “later” and “elsewhere”). An example of this approach is the aforementioned indicator framework developed by planning bureaus to quantitatively model the “wellbeing effects” of the government budget (CPB, PBL and SCP 2022). This focus on indicators sets up Broad Wellbeing to function as a yardstick against which the actual or expected performance of policy can be evaluated.
Among adherents of this approach, the question of how to measure wellbeing is seen as a purely technical exercise, which can—and should—be placed within technical expert organizations (interview U, 22 July 2022; see also Hoekstra 2019). This means indicator sets are developed to form a framework for wellbeing. In this framework, the selection, number, and directionality (whether having more or less of a certain indicator is better) of the indicators are considered undisputed. Only their relative priority with respect to each other (in terms of what is the optimal balance of the different aspects represented by the indicators) is seen as a matter of political choice. In a manner reflecting dominant economic approaches to projection, this framework is envisioned to assess the effects of policy decisions, ex ante and/or ex post. For instance, one informant suggests two possible scenarios when applying this framework (interview J, 30 June 2022). In the first, the projections are “perfect, thus everyone is happy.” The second scenario includes uncertainty, in which case “one can make periodic adjustments.” Of course, this begs the question whether a third scenario is possible, in which the indicators themselves are ambiguous. Prompted with this scenario however, the informant quickly reverts to discussing methodological challenges around uncertainty, comparing a wellbeing framework to GDP projections of which we are “sufficiently certain.” According to the informant, what is needed is making sure that indicators are “methodologically sound.” By extension, another informant notes how in this wellbeing framework, some indicators may be deliberately kept out of the framework precisely because they attract too much disagreement. For instance, she highlights how a potential indicator on “livestock density” might overstep her organization's mandate, as that would be an implicit political choice regarding an issue that is highly contentious in the Netherlands (interview A, 21 January 2022). Consequently, this approach follows what we call a technocratic logic of knowledge accountability. This logic assumes that the more knowledge available to policymakers, the better, i.e., the more “integrated” their decisions will be. Here, Broad Wellbeing can provide a—literally—broader assessment of these costs and benefits, which is assumed to support optimization and lead to “better” policy.
Importantly, technocratic knowledge accountability does not necessarily mean that science can provide the right course of action. This is a significant difference with arguments for outsourcing certain decision-making powers to scientific experts (Brennan 2016; for critiques see McGoey 2019; Reiss 2019; Van Bouwel 2022). For instance, in a policy brief on Broad Wellbeing, the Economic Planning Bureau emphasizes that “weighing the consequences of policy for different domains of wellbeing remains a task for politics” (CPB 2022, 5). At the same time, technocratic knowledge accountability contains a disciplining effect that is well-documented in the literature on performativity and indicators, which highlights the invisible normative choices involved in constructing ostensibly objective indicators (Mügge 2017; Porter 2020). One informant explicitly recognizes this potential disciplining effect and argues experts should actively “call bullshit” if Broad Wellbeing is used as a “political excuse” for policy choices (interview J, 30 June 2022). Also acknowledging this disciplining effect, another informant senses a latent desire—also among policy actors—for Broad Wellbeing analyses to yield unequivocal results in terms of “positive” or “negative,” despite all the talk of leaving prioritizing to politics (interview H, 23 June 2022). This means that Broad Wellbeing as instrument for optimization contains a paradoxical combination of political freedom and constraint. On the one hand, politicians are free to weigh and decide autonomously. On the other hand, their ability to explain these decisions is constrained by the affordances of a fixed and predetermined framework, constructed by specific science-policy organizations.
This paradox also comes to the fore in one informant's statement on supposedly problematic volatile political desires: The framework is developed independently of politics, but politics uses it to prioritize and develop policy. That might seem like outsourcing at first, and instinctively undesirable. However, if politics would develop the framework, a new cabinet would just create an entirely new framework. (Interview D, 6 May 2022).
Along these lines, several informants support the setting of Broad Wellbeing goals, but they disagree on what the purpose of such goals should be. Should they span long timeframes to discourage all-too-drastic changes when a new government is formed after elections, or should they, for instance, be part of the coalition agreement that each new government makes, and thus change every four years (or faster) (interviews B, 21 January 2022; D, 6 May 2022; K, 30 June 2022; R, 19 July 2022)? Here, some informants also note that the attractiveness of Broad Wellbeing itself depends on one's position on the political spectrum, with some political parties being more enamored by it than others (interview D, 6 May 2022). Another informant argues such a dependency is unwarranted, reasoning that every political movement has some sort of vision of the future which could become visible through Broad Wellbeing. Yet, referring to the scientific basis of wellbeing, the same informant reduces this kind of disagreement to scientific (un)certainty, contending that “science does have the task to narrow down possible opinions. To point out right and wrong views of the future” (interview U, 22 July 2022). Put succinctly, technocratic knowledge accountability here functions to stimulate knowledge on the effects of diverse policy options, yet also fosters ignorance of the diverse norms and values that may underlie these options.
Broad Wellbeing as Instrument for Discussion
In the second approach, Broad Wellbeing is seen as an instrument for discussion. Here, Broad Wellbeing is considered a relatively open-ended framework which can help elicit different visions, interests and perspectives in policy processes. An example of this approach is the set of guiding questions developed by the Ministry of Infrastructure and Water Management (2022) that civil servants can use in discussions and workshops. These questions involve effects on key policy priorities for the ministry (e.g., safety, climate, health) and can be used to explicate a proposal's objective and main benefits and drawbacks. This way of working is meant to ensure that a diversity of dimensions are brought to the table when considering a policy issue. In this approach, Broad Wellbeing is a tool to make policy choices transparent instead of a measurement framework to assess policy outcomes.
Informants describing Broad Wellbeing as an instrument for discussion emphasize its potential to enable transparent and ex ante justifications of policy decisions toward society and parliament, including normative choices related to trade-offs associated with those decisions (interview H, 23 June 2022). As one informant puts it, “Broad Wellbeing gives you a handle to clarify that choosing for one option, is also a choice against other options” (interview G, 23 June 2022). Informants contend that in this way, Broad Wellbeing could help to make visible what normativity is embedded in the choice for one policy instrument over another, abandoning the illusion that there is the possibility for an entirely neutral, value-free, position (interviews G, 23 June 2022; K, 30 June 2022). As one informant states: “I attach much more importance to the political dimension of the discussion than people who keep looking for some kind of analysis that is alternative to political decision-making” (interview H, 23 June 2022). Another summarizes this approach to Broad Wellbeing as informing the ongoing formulation of “what … we consider important,” as opposed to “if we calculate this accurately, it will tell us what to do” (interview Q, 14 July 2022). This approach can then be seen to follow what we call a political logic of knowledge accountability, which assumes that expertise can make explicit the normative dimensions of the choices policymakers face, and that it can enable the articulation of explanations that make this normativity visible. In this logic, knowledge accountability goes beyond merely offering insight into effects of policy decisions. It also explicitly invites deliberation about the norms and values against which effectiveness is assessed.
Broad Wellbeing as instrument for discussion puts less emphasis on capturing wellbeing outcomes in indicators, even if indicators can still help “show how you are doing” (interview O, 11 July 2022) and thereby play a role in articulating problems and putting these on political and societal agendas. To another informant, the complexity of societal transformations implies that a tight coupling between wellbeing and indicators is to be avoided. She argues that relying on highly detailed indicators and modelled versions of real-life complexity is risky, because a lot of nuance remains invisible and thereby easily skipped over in policy processes. Instead, using Broad Wellbeing as a tool to facilitate a transdisciplinary discussion on a concrete issue presents an opportunity to identify different options and an assessment of what their most important effects might be (interview G, 23 June 2022). Moreover, as another informant argues, the idea that this assessment has to be highly detailed may well be misguided for such a discussion, as information on the direction of a relationship and perhaps a rough idea of its strength may well suffice (interview K, 30 June 2022). This shows an appreciation of the expertise held by policymakers themselves and the ever-present need to act under conditions of uncertainty, with expert organizations focusing on how interrelations affect normative choices available to decision-makers. Here, political knowledge accountability functions to maintain a level of ignorance about the effects of policy options, while promoting knowledge of the norms and values that might make these effects (un)acceptable, as well as requiring decision-makers to explicitly recognize the complexity they face.
Accounting for Transformative Change
By distinguishing two general approaches to Broad Wellbeing in Dutch policy circles, we described two distinct logics of knowledge accountability in the previous section. Crucially, while both logics expect decisions to be made by politicians, they differ in whether they conceive the relationship between decision-maker and knowledge as uni- or bidirectional. Specifically, a technocratic logic expects knowledge to be used to explain the effects of policy options but considers such knowledge to be prior and external to the decision-making process. A decision-maker can be held accountable for how and what knowledge they use but bears no responsibility for how that knowledge was produced and by whom. A political logic expects knowledge to be used to facilitate political dialogue about the normative desirability of policy options. In this, the relationship between decision-making and knowledge is intertwined and bidirectional, and places a (shared) responsibility for knowledge production with decision-makers, i.e., for decision-makers to be reflexive of both decision and knowledge. We stress our description of these logics functions as a heuristic generalization, with in-between logics and hybridity more rule than exception. Nonetheless, this distinction allows for critical discussion of their implications for modes of expertise and governance. After all, as mentioned before, logics of knowledge accountability are not just descriptions of how expertise and ignorance feature in explanations about decision-making; they are also normative prescriptions about what expertise is worth pursuing and what expertise is not and, by extension, how politics and governance are to be conducted. We explore this point further by discussing what understandings of decision-making, politics and the role of knowledge are contained in the two logics.
Technocratic knowledge accountability operates from a decision-making logic of consequentiality, which sees policymakers as facing well-structured problems that can be addressed by different policy choices, whose costs can be assessed by expert organizations (Hisschemöller and Hoppe 1995; Dewulf et al. 2020; Maas, Pauwelussen and Turnhout 2022). In this view, the underlying societal goals and values of these problem framings are not disputed (Bijker, Bal and Hendriks 2009). Yet, because politicians may have different views on how to value costs and benefits relative to each other, the question of how to prioritize between societal goals is up to them. This means that issues of definition and operationalization can be delegated to the realm of expertise. Beyond that realm, questions like what counts as cost or benefit—is a cost actually a loss, is a benefit truly beneficial, and for whom?—are not up for discussion.
Conversely, political knowledge accountability reflects an understanding of decision-making as following a logic of meaning, in which policymakers need to make sense of problems that are treated as irreducibly complex, ambiguous and contested (Bijker, Bal and Hendriks 2009; Dewulf et al. 2020; Montana 2020). This logic does not delegate matters of definition and operationalization to expertise. Rather, it treats these as value-laden and as subject to political deliberation and decision-making. In addition to prioritizing costs and benefits of different policy choices relative to each other, this logic explicitly opens up discussion about what counts as a cost or benefit. Political knowledge accountability, then, expects decision-making to explicitly reflect on and make choices between different policy options and their politics (Stirling 2008; Kenis, Bono and Mathijs 2016; Graeber and Wengrow 2021).
Thus, the two logics of knowledge accountability pose different questions and require different things from expertise. Technocratic knowledge accountability asks: “are we doing things right?” and expects expertise to offer clear answers based on policy guidance on what “things” to focus on and what “right” means. Taking the aforementioned example of livestock density, this logic would expect the indicator to be unambiguous in its direction so that decision-makers can straightforwardly use it for prioritization with respect to other indicators. In contrast, political knowledge accountability asks: “are we doing the right things?” It expects expertise to support the mapping of diverse alternative “things” that can be considered, and to support the deliberation, and contestation, of why these alternatives might be the “right” thing or not, and what their consequences might be. Taking livestock density as an example again, this logic would focus on the multiple potential meanings of both “high” and “low” livestock density. It would ask experts to pluralize such potential meanings and decision-makers to articulate which they would value under what conditions. It follows that technocratic knowledge accountability works well if the ambition is to improve performance on current development trajectories. However, because it takes existing prevalent problem framings as the starting point, the policies that expertise can inform are those that are relevant within these definitions. We therefore deem it has little to offer to fundamentally challenge and change current trajectories since it will not require, and may even actively discourage, expertise to consider what alternative problem framings exist and what policy options could be relevant for these alternative framings. In contrast, political knowledge accountability requires expertise to continuously question problem framings and policy trajectories and can, we argue, therefore support fundamentally shifting or transforming them.
Notwithstanding the importance and usefulness of technocratic knowledge accountability for undisputed and well-structured problems (cf. Bijker, Bal and Hendriks 2009), transformative change requires the challenging and shifting of problem framings, goals and values (Avelino 2017; Blythe et al. 2018; Scoones et al. 2020). Here, the question of “are we doing the right things?” foregrounds that what ought to be sustained is an inherently political issue (Patterson et al. 2017; Scoones et al. 2020; Visseren-Hamakers et al. 2021; Yunita et al. 2022). This means that in political knowledge accountability, expertise is expected to politicize issues. As Castree, Bellamy and Osaka (2020, 74) highlight, politicizing can help to “reveal the real but very different decision-spaces that are defined by competing, though sometimes complimentary, political worldviews.” To political knowledge accountability, it is worth knowing explicitly what different views exist on what is to be transformed, to what end, how, and at what and whose costs. Political knowledge accountability thereby helps to reveal the “hidden moralities” of existing and potential alternative choices (cf. Prettner et al. 2021). Such knowledge, however, is seen as uncomfortable and irrelevant according to the logic of technocratic knowledge accountability (cf. Rayner 2012). This de facto choice not to explicitly politicize issues serves a kind of strategic ignorance in which alternatives are kept out of view, and in which vested interests can remain dominant by maintaining a state of unknowability (McGoey 2020). Recognizing that this form of strategic ignorance is untenable and unjust in the face of contemporary socio-ecological challenges and the need for transformative change raises the question of how political knowledge accountability can be advanced.
Politicizing Broad Wellbeing
So far we have established the limitations of technocratic knowledge accountability in fostering the production of expertise that can enable transformative change, and points to the potential of political knowledge accountability to overcome these limitations. Indeed, we contend that if expert organizations want to position themselves more actively as contributing to transformative change, they should emphasize political over technocratic knowledge accountability. To further explore this potential, we return to our case study of Dutch Broad Wellbeing. There, we find that the involved expert organizations unevenly enact the different logics of knowledge accountability. Specifically, we see a prevalence of technocratic knowledge accountability, as well as that at present there are no unequivocal indications that Broad Wellbeing will lead to transformative change. At the same time, individual interviewees did recognize the potential affordances of political knowledge accountability. This means that there still exists an opportunity to remake Broad Wellbeing and its associated science-policy arrangements. After all, expert organizations like the Dutch planning bureaus do have a choice; they can decide how and whether to produce knowledge/ignorance about explicitly political options. The question this raises is: how can expert organizations enact expertise for Broad Wellbeing in such a way that it supports political knowledge accountability?
To some degree, our earlier characterization of Broad Wellbeing as an instrument for discussion already contains several hints; expertise should support discussion in ways that make explicit underlying values about problems and options. At the same time, our interviews showed that it is much harder for people to imagine what this means concretely; what should experts do or not do to politicize Broad Wellbeing? How will they interact with policy and how will they secure authority? This inability to imagine alternative science-policy arrangements underscores the extent to which technocratic logics of knowledge accountability have become naturalized and unquestionable (Maas, Pauwelussen and Turnhout 2022). One of our informants wondered whether her discussion-oriented perspective on Broad Wellbeing would “still be concrete, still be seen as scientific” (interview K, 30 June 2022). This hesitation shows that something significant is seen to be at stake. A politicizing approach will change the science-policy-society contract, including norms about what is considered good science, and how science and society should relate. This is likely to face opposition from those experts and policymakers who have built their careers and reputation on these norms (Turnhout and Lahsen 2022).
To address this problem of imagining alternative science-policy arrangements, we proceed with a kind of “speculative design,” in the spirit of Haraway's (2016, 12) call to “stay with the trouble” and her emphasis, referencing Marilyn Strathern, on the importance of “what thoughts think thoughts.” We do this because speculating on how political logics of knowledge accountability could be fostered in Broad Wellbeing can help imagine this as a realistic possibility (cf. Oomen, Hoffman and Hajer 2022). This speculative design can start from the different treatment of diverse modes of expertise we have described for the logics of knowledge accountability in the two approaches we described. Broad Wellbeing as instrument for optimization seeks to integrate knowledge from different scientific disciplines into a single framework, which fixes what is wellbeing, to be able to measure progress toward it. Conversely, Broad Wellbeing as an instrument for discussion seeks to open up social and political deliberation about what wellbeing could be and should be. In this latter interpretation, it would function as a policymaking tool through which to realize what Aarts (2018) refers to as dialogue; a way to deal with conflict in a productive sense and create a sufficiently wide range of options without requiring the elimination of diverging norms.
Importantly though, politicizing Broad Wellbeing requires more than dialogue about alternative worlds. It contains a substantive commitment to pluralism, to ensure that such dialogue is always ongoing, to recognize difference and foster contestation, and to avoid “definitive definitions” of what counts as wellbeing. As such, it operates on the fine line between opening up to plural policy options and realizing that it is “simultaneously necessary, inevitable, and desirable” to make a choice between them (Stirling 2008, 284). In such a dialogue, participants would be encouraged to speculate about and articulate the possible worlds that Broad Wellbeing could support realizing and what could be done to achieve these. Expert organizations can contribute to such a dialogue by analyzing these as speculative or experimental facts (cf. Marres 2018). In that case, the role of expertise is not to adjudicate these possible worlds against presumedly impartial truths, but rather to examine under which conditions these futures could be real; what might it mean to actualize this claim of what counts as wellbeing? Such speculative analyses offer an opportunity to think through possible alternative worlds and their (un)desirability (Puig de la Bellacasa 2011), in full recognition that any alternative world will have its own ignorance. Here, politicizing Broad Wellbeing helps us to consider what values and (un)certainties we are willing to cast our lot with (cf. Latour 2018).
A modest example of what it could mean to politicize Broad Wellbeing expertise is found in a recent study on mobility policy by PBL Netherlands Environmental Assessment Agency (PBL 2021). Discussing what mobility policy could look like from the perspective of Broad Wellbeing, the study shows how a choice between different theories of distributive justice leads to different kinds of policy measures appearing as the “logical choice.” The study can be seen as part of wider attempts to reframe policy debates on mobility in the Netherlands—which can be one-sidedly focused on automobility—to include other modalities (e.g., public transport, bicycles) and accessibility (e.g., the disparate accessibility of public amenities by car or public transport). 7 The kind of expertise provided by this kind of studies can facilitate dialogue by opening up new options, questioning dominant options, empowering marginalized options, and navigating conflicts between options (Chambers et al. 2021, 2022). This also resembles what Castree, Bellamy and Osaka (2020, 74) dub “political expertise,” namely “to present the most elaborated versions of various worldviews, while also noting areas of complementarity, ambiguity, contradiction or confusion within them.” Adding such political expertise to a policymaking process using Broad Wellbeing will help the actors involved in the process to understand their respective positions, including their own, echoing the decision-making logic of meaning (Aarts 2018; Dewulf et al. 2020).
Undoubtedly, a politicizing approach to Broad Wellbeing means that the entanglement of expert organizations with decision-making processes becomes much more visible. Consequently, we anticipate that this approach will be opposed for violating presumed ideals of independence and neutrality. It may well be that in politicizing Broad Wellbeing, the contribution made by expert organizations is no longer recognizable as “scientific” by conventional standards. Here, the example of the mobility study is a report that still largely aligns with conventional expectations of what expertise looks like. A politicizing approach may further deviate from such expectations through, for example, the concepts used (e.g., using justice may seem problematically “normative” to some) or the forms in which expertise is “presented” (e.g., a collaborative workshop rather than a presentation). Importantly, we propose that expertise politicizes issues and supports pluralism, which differs not just from expertise that prescribes what options should be adopted, but differs from expertise that strictly separates facts from values. Moreover, taking the role of expertise to support politicization and pluralism points to how political knowledge accountability may contribute to realizing a particular understanding of democratic politics (Marres 2013). While we need to avoid taking particular understandings of democracy for granted (cf. Chilvers and Kearnes 2020), we follow Brown's (2025) argument that it is nonetheless valuable to “clarify, specify and expand” the potential contribution of expertise to democracy and its relevance to transformative change through engagement with political theory.
Specifically, we think political knowledge accountability can equip democracies to tap into the productive potential of pluralism (cf. Dijstelbloem 2007; Cuppen et al. 2019). This potential contribution resonates with qualities ascribed to democratic governance by both pragmatist and agonistic political thought. In John Dewey's pragmatism, democracy is co-constituted by publics and issues that require collective care, which through “issue articulation” institutionalize in particular (state) arrangements (Dewey [1927] 2016; see also Marres 2005; 2007). Importantly, such arrangements require continuous remaking, as they tend to hinder change and are unable to respond to emerging issues (Dewey [1927] 2016). Similarly, in her work on agonistic politics, Chantal Mouffe (2000, 2005) highlights that in societies that inevitably contain fundamentally different worldviews, consensus is always provisional. At most, consensus can offer “temporary respites in an ongoing confrontation” stemming from power relations and exclusion (Mouffe 2000, 102). In her view, the purpose of democratic politics is to ensure that confrontations of such pluralism are subjected to political contestation rather than to violence, without needing to subsume dissenting voices. Contestation is both foundational to politics as well as productive to democracy as an enabler of change (Marres 2012).
This points to a normative synergy between political knowledge accountability, transformative change, and these understandings of democratic politics. Our “speculative design” of Broad Wellbeing thus functions not just to show it could be, but also to show it ought to be. After all, we may consider efforts toward transformative change as efforts at issue articulation. A great deal of pluralism exists around these efforts, meaning transformative change is unbounded in scope and direction (Escobar 2020; Stirling et al. 2023), depending not just on different perspectives of what needs to change but also how and to what ends (Pascual et al. 2021). This absence of both factual and moral certainty should imply a fundamental indeterminacy to transformative change. Both Dewey's and Mouffe's views of democracy call for a similar commitment to indeterminacy, and they also offer a way to interpret this indeterminacy as an ongoing process of issue articulation in the face of considerable pluralism. Political knowledge accountability offers a means to resist urges for expertise to tame transformative change by oppressing pluralism, to instead support a continuous remaking of governance in pursuit of transformative change (Braun 2015; Stirling 2016; Bulkeley 2023; Muiderman et al. 2023).
Reworking Relations Between Science, Policy and Society
We close by reflecting on how we might go about fostering political knowledge accountability and what alternative practices it may be able to engender within expert organizations. In these reflections, modesty befits us. We have not aimed to provide a blueprint for such alternative practices, nor can we claim to have significantly re-made these practices from the inside. 8 Moreover, we realize our analysis is grounded in the context of a particular Western European country. The speculative design we set out in the previous section is one way to stimulate thinking on alternatives, which, in and of itself provides a kind of experimental fact. With it, we hope to contribute to a wider discussion on how to improve relations between science, policy and society. Of course, we cannot assure political knowledge accountability will deliver the transformative potential we attribute to it. Yet, we are astonished by the many expert practices that stubbornly insist on a technocratic logic that is evidently impotent.
We recognize that it would be naïve to expect changes to arise endogenously in expert organizations. For one, because it entails a change in priorities that affects people's careers (Turnhout and Lahsen 2022). But also because relations between science, policy and society are a collective responsibility, and by direct extension, so is enacting political knowledge accountability. Change, thus, not only depends on “the experts.” It also depends on policymakers and politicians explaining their decisions in normative terms, rather than deferring their responsibility to some transcendent scientific authority (Stirling 2010). The same goes for the various interlocutors referring to expertise in public debates, including the media and activists, who have a responsibility to interrogate decision-making and the role of expertise in a way that promotes discussion on what is desirable rather than what is necessary. In this collective responsibility, not all worries and feats about a changing authority of science are created equally, since a changed authority of science might just be the revolution we need to break through the dominance of established actors (Esguerra and Van der Hel 2021).
At the same time, it is pertinent to point out that fostering alternative practices will require two important changes in the institutional epistemologies of expert organizations (Borie et al. 2021). First, it is crucial that expert organizations cease to feign ignorance of questions that are still commonly perceived as “non-scientific.” Stengers (2018, 103) writes that expert organizations should cease thinking they are “trespassing on matters to be decided in terms of political or ethical values” when dealing with “non-scientific” or “non-objective” questions. Second, it requires expert organizations to resist the urge to be relevant, at least for as long as this relevance continues to be defined in terms of a fluid connection between a shared and singular dominant frame for both knowledge and action (Turnhout 2024). Dealing with strategic ignorance requires a critical reflexivity to recognize the questions whose answers may currently be politically inconsequential but nevertheless may hold transformative potential. These questions and the political consequences that may derive from answering them should lead our expectations of what expertise can contribute to decision-making. Ultimately, we think this is where the added value of expertise in navigating contemporary challenges lies.
Even so, we caution that any resulting “properly politicized transformative change” is not guaranteed to be in accordance with the (progressive) politics scholars in STS may hold personally, emphasizing the need to defend democracy even when it does not lead to “the right” outcomes (i.e., the outcomes we personally prefer). This is also a call to STS itself. We think STS can and should play an active and politicizing role in reconfiguring relations between science, policy and society. In line with popular notions of care and “making and doing” (Puig de la Bellacasa 2011; Downey and Zuiderent-Jerak 2016), we remind our readers to question “what kind of world we want to participate in building” (Law and Urry 2004; Gibson-Graham 2008). STS engages with, cares for, even meddles with, its technoscientific objects of study not merely to understand them, but also to change those objects. After all, Hickel's call to be “smarter” in this paper's epigraph concerns not just changing how we account for wellbeing but also changing to what and whose ends we do this accounting.
Footnotes
Acknowledgements
We are grateful to our informants for sharing their thoughts and insights with us. We also thank Abe Hendriks and two anonymous reviewers for thoughtful comments on earlier versions of this article.
Ethical Considerations
This study was exempted from ethical approval by the Wageningen School of Social Sciences, based on the fact that informants are highly educated government officials, and that the research objective was disclosed prior to the interviews.
Consent to Participate
All informants were sent an information sheet with the invitation to participate in this study. The sheet outlined the objectives of the study (both academic research and a PBL study on Broad Wellbeing and the Sustainable Development Goals), data management and anonymity of literal responses.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
