Connecting with Scholarship on the IPCC and its Social Contract

This article follows a long history of critical social science about the IPCC organization, dating back 30 years or more, and principally from scholars from science and technology studies, geography and political science (De Pryck and Hulme Reference De Pryck and Hulme2022; Hughes Reference Hughes2024; Hulme and Mahony Reference Hulme and Mahony2010; Mahony Reference Mahony2014). Most of this work centres on the IPCC as a site for the mutual co-production of global science and politics, where new global governance orders are under construction (Hulme Reference Hulme2010; Miller Reference Miller and Jasanoff2004; Shackley and Wynne Reference Shackley and Wynne1995). For these scholars, social processes of interaction and communication among IPCC scholars are significant for the production of knowledge. Yet the IPCC is charged not with the generation of new knowledge, but rather the review and synthesis of existing studies, so that its reports act as second-order knowledge, or knowledge about knowledge. This mission makes IPCC a rather special case in terms of the boundary between science and politics and the negotiation at that boundary, which scholars have analysed from several angles (Beck and Mahoney Reference Beck and Mahony2018; Hoppe et al. Reference Hoppe, Wesselink and Cairns2013; Sarewitz Reference Sarewitz2011), including the linear ‘deficit’ model of communicating science (Beck Reference Beck2011); the organizational norms and procedures (Hulme and Mahony Reference Hulme and Mahony2010); the tools and framing work of the IPCC (Petersen Reference Petersen2012); or approaches to building consensus under uncertainty (van der Sluijs Reference Van der Sluijs2012; van der Sluijs et al. Reference Van der Sluijs, Van Eijndhoven, Shackley and Wynne1998), to name a few main topics. As will be seen below, the key to understanding IPCC practices in this respect is the distinction between ‘policy-relevant’ and ‘policy-prescriptive’ knowledge.

This work is made more complicated in that ‘climate’ is a particular object of study. It is of rather recent construction and is achieved through complex means of data-gathering, setting of conventions, and modelling, as well as by bringing insights and approaches from different research areas together (Aykut Reference Aykut2016; Pearce et al. Reference Pearce, Mahony and Raman2018). Furthermore, by measuring climate by resorting to the temperature of the atmosphere, climate research addresses a phenomenon with which human beings are inevitably entangled and from which they cannot gain distance. Thus, ‘climate’ is not an object at all in the conventional sense of the term, possibly instead a ‘hyperobject’ (Morton Reference Morton2013), even though this term is arguably infelicitous.

One facet of the IPCC facing scrutiny (especially since the sixth Assessment Report) is its ‘social contract’ (Hulme and Mahony Reference Hulme and Mahony2010). Work on science’s social contract goes back (in various forms) to at least the early-twentieth century and fundamental work by, for example, Merton (Reference Merton1938) or Fleck (Reference Fleck1981), with one strand going via a concern for how science relates to social problems (e.g. Ravetz Reference Ravetz1971, Reference Ravetz1988), to a more recent concern at the end of the twentieth century with science’s role in addressing societal (and environmental) challenges. In this context, Gibbons (Reference Gibbons1999) laid out science’s social contract, while Lubchenko (Reference Lubchenco1998) related this contract especially to global environmental change. These scholars sketch a social contract for scientists to devote their energies to societies’ pressing (environmental) challenges in a way that is socially robust, transparent and participative. Indeed, such principles have taken firm root in the scientific ecosystem, as central to ethical research conduct (see an example in the NESH and NENT guidelines of the Norwegian National Research Ethics Committee’s, or ALLEA’s European Code of Conduct for Research Integrity). At the IPCC, this contract stems from the role that it was established for, to distil scientific, policy relevant and responsible evidence in support of inter-governmental negotiations.

But Glavovic et al. (Reference Glavovic, Smith and White2022), following the sixth Assessment Report, proclaimed that since the IPCC was incapable of sparking meaningful governance action, then this contract is broken. The organization has lost its compass, its identity and its authority, insofar as it no longer fulfils its mission in society. This article represents one of the more recent, and explosive, critiques of the IPCC, although scholars have long, for over 20 years, questioned whether the IPCC fulfils its aims (see, for example, Sarewitz Reference Sarewitz2004; van der Sluijs et al. Reference Van der Sluijs, Van Eijndhoven, Shackley and Wynne1998), including by authors with experience working within the IPCC organization (Swart et al. Reference Swart, Bernstein, Ha-Duong and Petersen2009). Others responded to Glavovic et al. (Reference Glavovic, Smith and White2022) by arguing that the contract is not broken, but under re-negotiation, and that the IPCC needs to change direction towards better incorporating the social sciences and humanities, with implications for the IPCC’s identity and authority. Indeed, Gibbons’ (Reference Gibbons1999) claimed that science’s social contract is ‘a dynamic process in which the authority of science [is] legitimated again and again’. It is a stretch to proclaim that the Glavovic et al. (Reference Glavovic, Smith and White2022) article, and ensuing furore, is evidence for a widespread identity crisis in the IPCC. But as we show below, the interviewees that participated in our limited study also raised topics of shifting identities, implying that these are concerns held by at least some (central and senior) IPCC authors, and that they are worthy of further discussion.

So, considering the IPCC’s evolving social contract, what has this meant for the identity and authority of the IPCC for climate governance? The scientific authority of the IPCC is a major part of its self-identity and how its members perceive its role in climate governance, so we consider these two concepts as interlinked (Hajer Reference Hajer2012; Mahony Reference Mahony2014; Miller Reference Miller and Jasanoff2004). We were interested in the elements that IPCC authors draw on to define the IPCC and their work, and its perceived authority for guiding international relations and global governance. Some scholars have tentatively set out elements that are relevant to identity, but rarely in discussion with IPCC authors themselves. These elements include the IPCC’s claims to universality (Hulme Reference Hulme2010); political neutrality ensured by a demarcation model (Hoppe et al. Reference Hoppe, Wesselink and Cairns2013); a linear communication model (Beck Reference Beck2011; Dudman and de Wit Reference Dudman and de Wit2021); its framing of the science (Pearce et al. Reference Pearce, Mahony and Raman2018); transparency in addressing mistakes (Hajer Reference Hajer2012); the quality of the information as fit for the function of decision-making (Bremer et al. Reference Bremer, Wardekker, Jensen and van der Sluijs2021); or the degree of diversity and participation in the organization, relative to disciplines (Bjurstöm and Polk Reference Bjurström and Polk2011), inclusiveness and modes of communication (Caretta and Maharaj Reference Caretta and Maharaj2024), or knowledge systems (Bremer et al. Reference Bremer, Klenk, Bastian and Kwan-Lafond2024).

In framing this study then, we focused on the dynamic process of renegotiating the IPCC’s social contract relative to the co-evolution of some key elements that underpin its identity and authority for governance. This focus was emergent, mutually guided by conceptual viewpoints in the literature (above) in dialogue with topics that arose in our interviews. That is, we did not progress from a strict a priori conceptual framework but sought to highlight elements under broad headings of identity and authority that interviewees themselves identified, and which we came to distil as seven elements (discussed under three headings). In this way, for instance, the identity crisis that we take as a point of departure is less a theoretical standpoint than one that became apparent in the interviews.

Method

This pilot study was initiated by scholars in a taskforce combining two European academies – Academia Europaea and the Young Academy of Europe – which is investigating models of science-policy for environmental sustainability. The research was designed as a set of semi-structured interviews, which was determined to be the method best fitted to eliciting self-reflections in a confidential setting, towards fulfilling the research aim. In total, nine IPCC contributors were interviewed between October 2024 and January 2025, with all interviews held virtually (given the geographical spread of interviewees), and in English, with two interviewers (the authors) in dialogue with one interviewee. The interview protocol was loosely informed by topics in the literature, and interviewees were sent the questions at least a week in advance with information about the pilot study. At the start of the interview, interviewees were asked to give verbal informed consent. Interviews were automatically transcribed, and transcripts sent to interviewees to review.

The taskforce identified potential interviewees from members of the two Academies, and via a snowballing technique, through interviewees’ networks and recommendations, with the aim of securing a diverse set of perspectives on the IPCC. The areas of scholarly expertise were the following: climate and environmental physics; earth sciences and paleoclimatology; biology and earth sciences; geography; oceanography; economics and environmental science; energy economics; technology policy; science and technology studies and energy policy. While this range is rather broad, in the end the sample was skewed towards scientists in Working Group I on the physical science basis (WGI), with a smaller number from WGs II and III on adaptation and mitigation respectively, and towards more senior male scientists from the Global North. In more detail, five of the nine interviewees participated in WGI and two each in WGs II and III. Four of the interviewees contributed to all six assessment reports; three to reports 4, 5 and 6, and two to the fifth and sixth reports. Six interviewees are male and three female; six interviewees are based in Europe and one each in Latin America, Africa, and Oceania. We consider this set of interviews as a pilot study to be extended in a second step with further interviews, but the interviewees’ range of expertise and variety of experience with IPCC lend themselves to the elaboration of a rather consolidated picture.

The following presentation of the findings concentrates on the interviews themselves, due to both their richness and the limitation of space. We opted for a narrative style that synthesizes topics raised by interviewees rather than making use of quotations that would considerably overlap. When there were significantly different opinions among the interviewees, this is pointed out, on occasion by referring to the kind of involvement of the interviewees. In this sense, almost the entire following section should be read as giving voice to the interviewees. Thus, it complements a document analysis of the IPCC reports as well as their reception, which to some degree exists in the available literature that we mentioned above. It is only in the third subsection, which deals with shifting issues in the IPCC work, that we felt compelled to cross-refer the interviewees’ statements to other existing analyses of the IPCC mode of operation. Our proper analysis can be found in the concluding section on the anatomy of a crisis, in which we carve out from the interviews seven elements that emerged as under negotiation for the future work of the IPCC.

Task Identification and Interpretation: Ambition and Limitations

From the start, the IPCC benefited from a rather clear mission that led to the development of an identity, in the sense of a common resource of meaning that guides action. By assessing the state of knowledge about climate change, the IPCC was to determine whether and to what extent global warming was due to human activity and whether and in what form it required urgent coordinated global action. In its First Assessment Report (AR1), in 1990, the IPCC answered the first question affirmatively and urged global climate action. AR1 played a decisive role in the creation of the United Nations Framework Convention on Climate Change (UNFCCC) and in the preparation of the United Nations Conference on Environment and Development (widely known as the ‘Earth Summit’) held in Rio de Janeiro in 1992, at which the UNFCCC was formally agreed upon.

The following Assessment Reports confirmed the initial assessment with an ever-greater degree of certainty and also specified the significance of the increase in global temperature through enhanced historical and current data and improved climate modelling. Thus, it tends to be forgotten that the early success of IPCC in framing the global climate debate and stimulating global climate action could not be taken for granted. Climate scepticism was widespread at the time, if by this term we mean that doubts were raised about all steps of the reasoning: whether global warming was indeed happening; if it was happening, whether it was due to human activity rather than planetary factors; if it was due to human activity, whether it required urgent coordinated action to mitigate climate change. When affirmative evidence consolidated on all three questions, such justified scepticism partially transformed into what came to be called ‘climate denialism’. By the latter, one refers to rejection of credible evidence for reasons of interests broadly understood, which can be the profit and revenue motives for businesses and states, but also a reluctance to consider changes to citizens’ ways of life.

For an extended period, the IPCC was subjected to fierce criticism by climate sceptics of various kinds, sometimes present within IPCC but more publicly visible when it came from the outside, and often supported by organized interest groups or governments. Such criticism was questioning the findings directly by raising doubts about the methods, or it was aiming at the degree of certainty with which the findings could be expressed, underlined by the IPCC practice of indicating intervals of confidence with which statements are made. With the evolving knowledge that confirmed the initial assessment, though, outright climate denialism has been reduced to pockets in debates around the IPCC. By the time of AR5 in 2014, which triggered the Paris Agreement of 2015, the evidence presented was so overwhelming that the interpretative space for climate denialism had been considerably reduced (leading to a shift in debate that will be addressed below).

From the interviews it became clear that most (though not all) of the scholars involved in IPCC – let us label them ‘climate scientists’ – had none of the doubts raised by sceptics. For them, it was rather the existing state of knowledge, at the outset, that led the UNEP and WMO to take the initiative to create the IPCC in the first place. Thus, it becomes clear that the IPCC indeed had a double mission: beyond surveying and synthesizing the evolving state of climate knowledge, its task was to convey this knowledge and its significance to governments. In the latter regard, the IPCC is a convenor of global climate debate among climate scientists and office-holding politicians rather than a mere producer of comprehensive ‘state of the art’ reports.

This insight directs attention to the way in which the boundary between science and politics was defined and practised in the IPCC. With public attention focused on the ARs, the IPCC appears as a knowledge-providing institution. From one AR to the next, the media tend to present the evolving state of knowledge, often focused on the predictions for future global warming. Arguably, though, the more significant achievement of the IPCC with each AR is to reach a consensus with all participating governments about the state of climate knowledge. This consensus becomes explicit in the Summary for Policymakers (SPM) that accompanies each AR and is approved line by line by all participants.

The creation of the SPM shows the ongoing boundary work meant to distinguish science from politics. The draft SPMs are scrutinized in detail, and contestations are frequent, in particular regarding the statements about carbon dioxide emissions and fossil fuels. It is thus evident that the SPM is seen as a political document, in particular but not only by (certain) governments. At the same time, it is a ‘summary’ of the three working group reports that constitute the bulk text of an AR, and these reports are by definition scientific, as they assess the state of published academic knowledge on climate change. By common understanding, therefore, the SPM cannot contain any statement that does not have a basis in the working group reports. Moreover, it cannot contain any statement that contradicts working group findings. This understanding provides a basis for arguing against requests to change the draft SPM, altering, adding or – more typically – deleting phrases. Participants in the working group on the ‘physical science basis’, in particular, underline that this way of conceiving the link of the SPM to the more detailed assessment thwarts politically motivated attempts to water down the SPM.

There is, though, another component of the IPCC mission that can be seen as protecting politics from science. The IPCC is held to assess the existing knowledge on climate change in a ‘policy-relevant’ but not ‘policy-prescriptive’ way. This distinction was frequently cited by interviewees demonstrating its significance as a defining element of the IPCC mission. While it was mostly stated matter-of-factly to start with, subsequent discussion often showed an ambivalent attitude to this distinction. By delimiting the range of the sayable, the distinction creates a boundary that can serve to orient the work and provide task security. Furthermore, this boundary also supports the claim to authority of the IPCC. Interviewees tended to agree that the distinction was important in enhancing the legitimacy of the IPCC work.

But the distinction was also seen as constraining. At times, interviewees felt that findings clearly indicated the actions that need to be taken, but the ‘relevant-not-prescriptive’ distinction prevented them from saying so explicitly. Some shift in evaluating this ambivalence over time can be observed. Early on, the belief could be held that the evidence of the findings would lead almost by necessity to adequate action, by sheer force of reason. However, the experience that such action persistently fails to occur can suggest to scholars that they may have to – and should be permitted to – spell out which course of action needs to be taken. Or, in other words, while it may have been seen at the outset as sufficient to provide and strengthen evidence of anthropogenic climate change, the urge to become policy-prescriptive arises given the lack of action that would be required in the light of the policy-relevant knowledge.

At least implicitly, the boundary between science and politics in the case of the IPCC is underpinned by some notion that the knowledge is already there and only needs to be presented in synthesis, whereas the course of political action still needs to be decided upon and taken. This notion, though, underestimates the extent to which knowledge needs to be collected, ordered, and organized before it can be presented in synthesis. These are social processes that would be ill-conceived if they were seen as being fully determined by a knowledge that is already ‘there’. It is to these processes within IPCC that we now turn.

Mode of Operation: Organizing Knowledge

From its beginning and until the present day, the work of the IPCC has been divided among three working groups, dealing with the ‘physical science basis’ (working group I – WGI), ‘impacts, adaptation, and vulnerability’ (working group II – WGII), and ‘mitigation of climate change’ (working group III – WGIII) respectively. The working groups operate relatively independently of each other and present separate reports, which are then brought together in a ‘synthesis report’ and, with further synthesization, in the SPM. While a division of labour appears necessary given the complexity of the issue, and the chosen sub-division seems quite plausible, any such separation of tasks raises questions and potentially creates problems of coordination.

Given the initially prevailing understanding of the task, the findings of WGI have tended to dominate the reception of the IPCC work. It was there that it was determined that global warming is happening and that it is anthropogenic. WGI is highly ‘policy-relevant’ in the sense that its findings underline the urgent need for climate action. In turn, it is in WGII and WGIII that the course of such climate action is at stake, in terms of measures to adapt human societies to climate impacts and in terms of measures to mitigate the further rise of global temperatures, respectively. With its focus on the ‘physical science’ basis, the members of WGI are mostly based in the earth and climate sciences, which itself is an area of knowledge that, if it has not precisely been brought into being by the rising climate concerns, it is at least strongly shaped by them. Among the members of WGII and WGIII, concerned with the human response, scholars from engineering disciplines and from economics have a stronger role. The presence, or lack thereof, of scholars from wider areas of the social sciences and the humanities is a topic under debate (to which we will turn below).

Over time, the division of labour between working groups has become consolidated, or indeed somewhat petrified. Consequently, concern has arisen about issues and areas of knowledge that are not adequately covered because they do not fall clearly under the remit of any one of the working groups. IPCC has tried to deal with such concerns by elaborating ‘special reports’ – such as on cities, land, and oceans as well as a report dedicated to the target of limiting temperature increase to 1.5°C – or setting up mechanisms for ‘handshakes’ across working groups. Nevertheless, there is considerable cost in the separation of tasks, visible when one working group adopts concepts that are not adopted by another one or one group divides the globe in different world-regions than another one, in both cases decreasing the comparability of statements.

With an exponential explosion in the number of climate research studies to assess, the IPCC has witnessed a strong increase in the number of scholars participating in its activities. Therefore, the task of assessing the state of knowledge has become much more complex. If the task has remained somewhat manageable, this is at least in part due to the elaboration of practices and a mode of reporting on which every new assessment cycle can build. As such, an ongoing process, by now extended over more than 30 years, the IPCC does more than assess an evolving state of knowledge; it inadvertently (and even quite deliberately) intervenes in the process of knowledge generation.

Taken together, the contributors to IPCC reports as authors or reviewers occupy leadership roles in the global community of climate scientists (in the broad sense of this term, which rather means a set of sub-communities in different disciplines and specialty areas). When their assessments characterize a state of knowledge, they also – implicitly and sometimes explicitly – mark areas in which more knowledge would be desirable, including to impart greater certainty than currently exists. Returning from IPCC work to their own workbench, these scholars are generators of knowledge, and they may well be inclined to address their research efforts (or those of their colleagues) to those questions which IPCC assessment marks as crucial. In this way, the IPCC gives direction to knowledge production, even quite directly through a scholarship programme (financed by the Nobel Peace Prize presented to the IPCC in 2007). Such knowledge production is particularly prevalent in connection to controversy in public and political debate. For some time, for instance, the stability of the past climate was doubted with a view to claiming that current global warming may just be a climate fluctuation that occurs without human impact. Research in paleoclimatology was instigated to address this controversy and disprove this claim.

Despite this capacity to generate knowledge, IPCC reporting is constrained by the state of published academic knowledge at a given moment. Given the amount of available research, selectivity is unavoidable, and this may open spaces for interpretation, such as hinting more strongly at some policy actions as adequate than at others. Nevertheless, the demand for presenting the full range of knowledge persists, which can lead to a conservative bias in reporting, with ‘conservative’ here meaning a tendency to underestimate the extent and impact of climate change.

A case in point is climate modelling. It is widely known that the modelling results depend both on the available and used data and on the assumptions made about the relations between them, often leading to considerable divergence between models or within a model when operated under different assumptions (see the contribution by Phoebe Koundouri et al. to this issue). But there is also broad agreement that climate modelling has been considerably improved over the past decades of climate research on both grounds, more and better data and more reliable modelling of their relations. As a consequence, the degree of confidence with which IPCC reports climate projections has increased over time. More recently, though, the argument has been made that climate observations tend to exceed considerably, and negatively, the expectations of climate models. Asked about this discrepancy, interviewees point to the fact that individual climate models indeed have those recent climate occurrences within their range. But the IPCC reports a summary view of modelling predictions and thus cuts the ‘tails’, that is the more extreme findings, from models.

Beyond the organizational constraint of only summarizing certified knowledge, one can witness here an attitude of what we may call epistemic conservatism. IPCC scientists apply a principle of prudence by prioritizing knowledge that shows a high degree of certainty, tending toward presenting more conservative estimates of change (Wynne Reference Wynne2010), anchored in past estimates (van der Sluijs et al. Reference Van der Sluijs, Van Eijndhoven, Shackley and Wynne1998), underlining findings by scholars 15–25 years ago. By implication, though, applying this attitude at a science–politics interface invites violations of the principle of prudence in action (see, for example, Shue Reference Shue2015), which may demand action precisely because of uncertainty in the face of possibly disastrous consequences.