Collective securitisation

The theoretical starting point for this study is the literature on Collective Securitisation. Collective Securitisation offers a phase-model that revises the traditional understanding of securitisation as the process by which an actor identifies a ‘threat’, attributes a special status to that threat, and, by so doing, is able to justify ‘urgent and exceptional measures’ in response.Footnote ¹³ The Collective Securitisation model widens how and where securitisation takes place and helps to capture its effects on outcomes – especially policy outcomes. It theorises how multiple actors, themselves with individual securitising imperatives, may empower a third entity to act on their behalf in securitised ways. Collective Securitisation is thus the process of aggregating these multiple securitisations and giving them authoritative status through public articulationFootnote ¹⁴ – and so is most obviously undertaken by formal international organisations.Footnote ¹⁵

Collective Securitisation highlights the dynamic of states acting collectively within organisations to bring attention to certain concerns as security questions – and to leverage collective attention and resources accordingly. Haacke and Williams first drew attention to this dynamic, showing how securitisation can take place in regional organisations such as the African Union and the Association of Southeast Asian States.Footnote ¹⁶ Since then, studies on Collective Securitisation have identified ‘thin’ and ‘thick’ versions. The thin version refers to a state or small group of states promoting their security concerns within an international organisation, eliciting supportive responses and empowering the organisation to address the issue without granting it autonomy or agency.Footnote ¹⁷ Conversely, the thick version views the organisation as possessing distinct influence and agency, acting as a securitising actor in its own right.Footnote ¹⁸ Research on the EUFootnote ¹⁹ and UNFootnote ²⁰ often applies the thick version due to the substantial political agency and coercive authority both organisations wield in implementing security measures.Footnote ²¹ This study adopts the thick version of Collective Securitisation.

Sperling and Webber added theoretical depth and analytical precision to the Collective Securitisation approach by developing a stage-based analytical framework that theorises the various steps in the thick version of Collective Securitisation (see Figure 1).Footnote ²² They show how international organisations may act as securitising agents, with member states serving as the audience that either accepts or rejects securitising moves from actors in and around the organisation. Emphasising intersubjectivity over unilateral speech act acceptance, Sperling and Webber introduce the notion of ‘recursive interaction’, whereby the securitising actor and the audience influence and shape one another, effectively blurring the actor–audience distinction.Footnote ²³ This recursive process not only facilitates audience acceptance but also leads to the production of new policy outputs that concretise securitisation and establish a new status quo. Moreover, including and emphasising the effect of securitising moves on new policy outputs, Sperling and Webber explicitly linked securitisation processes to the nature of resulting policies, asserting that the latter should reflect and embody the characteristics of the former. Figure 1 represents these process phases, which our study builds upon.

Source: developed by Sperling and Webber, ‘NATO and the Ukraine crisis’; Sperling and Webber, ‘The European Union’; from Sperling and Webber, ‘The European Union’, p. 246.

Figure 1. The process of Collective Securitisation.

For this study, the Collective Securitisation framework provides a critical lens for analysing whether and how UN and EU adaptation policies mirrored the broader securitisation of climate change. Framing an issue as a security matter at the supranational level not only shapes the tools utilised, the resources allocated, and the strategies implemented but also significantly impacts how the issue is addressed at nationalFootnote ²⁴ and local levels.Footnote ²⁵ Understanding these international processes is vital for assessing how urgent global challenges are managed in the public interest.

Limitations of Collective Securitisation

We see the need to address two limitations in the Collective Securitisation literature, however. One is a tendency to view international organisations as closed systems. Analysis often focuses on discrete contexts, such as the World Health Organization,Footnote ²⁶ NATO,Footnote ²⁷ or the UN,Footnote ²⁸ and how Collective Securitisation unfolds within each. Singular focus offers empirical richness, but it is difficult to square with common understandings of the international institutional environment as dense and overlapping. Multiple international organisations govern across policy areas, even if memberships, remits, and levels of authority differ.Footnote ²⁹ The international policy environment is notably open, with overlapping regimes,Footnote ³⁰ epistemic communities traversing jurisdictions to influence international policy,Footnote ³¹ and intersecting policy instruments challenging governance.Footnote ³² Indeed, an entire research field explores international organisations’ interaction with and influence on one another.Footnote ³³

Collective Securitisation research insufficiently addresses how international organisations interact within securitisation processes. Insights from norm diffusion and policy learning literature can help clarify when, where, and why these interactions occur. Norm diffusion research, although focused mainly on diffusion across countries, highlights several dynamics that may prompt a process of diffusion. One is the onset of an internal policy failure or an external shock that triggers a search for solutions.Footnote ³⁴ Another reflects altered pay-offs and a concern for reputation. The adoption of a policy by a growing number of governments changes the pay-off for those left behind. What once seemed like a costly endeavour – the adoption of new policies and attendant norms – is less so once it appears that like-minded authorities are doing the same.Footnote ³⁵ Finally, learning from the experience of others may trigger diffusion as new information of successful adoption becomes available.Footnote ³⁶

The literature discussed above offers valuable insights into when interaction between the Collective Securitisation process in one IO may initiate or interact with a similar process in another. Such interactions, we argue, can be mapped onto key stages of Sperling and Webber’s model, highlighting potential points of overlap or influence. Specifically, three possible points of interaction emerge. First, these interactions can occur at the ‘securitising move’ stage (see Figure 1), where an IO reframes an issue, introducing new discourses that may inspire or prompt another IO to adopt similar securitising language. The emulation of discourses and framing from one IO to another often reflects dynamics of diffusion or institutional learning, with securitising moves in one venue setting the stage for comparable actions in another.

Second, interactions may rise in the ‘audience response’ stage. Securitisation theory identifies audience acceptance as a necessary condition for successful securitisation. In the context of Collective Securitisation, the audience comprises member states of an IO, whose representatives, including diplomats and national policymakers, often participate in multiple IOs.Footnote ³⁷ Approval or disapproval of a securitising move in one IO may travel with these representatives, influencing the audience response in another IO and thereby facilitating (or hindering) a parallel securitisation process.

Third, interaction may take place at the ‘new policy outputs’ stage. Following successful securitisation, IOs often produce new policies or instruments that reflect the securitised framing. Insights from the transnational policy learning literature show that IOs frequently reference one another’s positions to enhance legitimacy or adapt proven solutions.Footnote ³⁸ Policy tools and arguments developed in one IO are often ‘borrowed’ by another IO via emulation and attention from networks of like-minded actors, thus reinforcing connections between their respective securitisation processes.Footnote ³⁹ In this way, at this stage, the discourses and policy outputs of one IO may inspire or directly shape the policies of another.Footnote ⁴⁰

In sum, the literature enables us to pinpoint when interaction between Collective Securitisation processes might occur: at the stages of securitising moves, audience response, and/or new policy outputs. These stages reflect potential pathways for emulation, diffusion, and cross-organisational influence within the framework of Collective Securitisation (see Table 1 for a summary).

Table 1. IGO interactions during Collective Securitisation processes.

Source: Created by the authors.

The second limitation of Collective Securitisation is a lack of specification in what constitutes securitisation. To be fair, Sperling and Webber’s version of Collective Securitisation relaxes the Copenhagen School’s conventional argument that an ‘existential construction’ of a new threat is required. Sperling and Webber follow recent theorising in Critical Security Studies showing that a range of referent objects may be securitised and that securitised outcomes need not be ‘urgent and exceptional’ nor dramatic in character.Footnote ⁴¹ Outcomes evolve incrementally over time and reflect a kind of routine politics focused on reducing harm through a wide range of measures. Such security concerns include a wide array of control strategies and ‘risk management’ approaches attached to governments’ efforts to address a span of current public policy problems.Footnote ⁴² This open approach to measuring and defining securitisation offers advantages in capturing a wide range of acts, practices, and outcomes. However, it also runs the risk of categorising any new practices as securitisation without providing clear criteria for distinguishing these acts, practices, and policy outcomes as genuinely ‘securitised’.Footnote ⁴³

To address the second limitation in Collective Securitisation research, we advocate for the inclusion of measures that capture both traditional threat-based approaches and risk-based approaches to securitisation, the latter associated with the concept of ‘riskification’. Drawing on the securitisation continuum proposed by Diez et al., we propose distinct empirical indicators for both threatification (threat-oriented securitisation) and riskification (risk-oriented securitisation) to more comprehensively measure securitised outcomes.Footnote ⁴⁴ The riskification literature systematically updates the traditional threat-based model of securitisation – characterised by urgency and existential threat – by recognising how public policy problems are increasingly framed and managed through the lens of risk and precautionFootnote ⁴⁵ This approach is defined by policy frameworks rooted in calculative rationality, where decisions are guided by probabilities, risk assessments, and cost–benefit analyses.Footnote ⁴⁶ Policymakers often deploy tools designed to identify, measure, address, and predict risks.Footnote ⁴⁷ The riskification literature is crucial as it provides a nuanced understanding of how contemporary public policy challenges are increasingly approached through proactive and precautionary measures, offering a complementary framework to traditional threat-based models of securitisation. To operationalise the distinction between threat and risk logics in climate change and adaptation policy discussions, we identify three key analytical indicators that help differentiate between risk-oriented and threat-oriented securitisation.

The first analytical indicator is a discursive shift from framing issues as existential threats with dramatic imagery, and calls for immediate action, towards emphasising inherent vulnerabilities that require ongoing management, monitoring, and assessment. The second analytical indicator is signs of a reliance on and belief in expertise and technical knowledge in addressing vulnerabilities, rather than the higher-level authorities and resources (often military, in cases of traditional threat-based securitisation). A riskification of an issue frequently elevates actors with scientific expertise or close connections to such experts, highlighting a shift in focus from immediate responses to calculated, ongoing management.

The third analytical indicator emphasises the importance of examining not only the actors and their discourses but also the tools and instruments deployed to address perceived security problems.Footnote ⁴⁸ In contrast to the military tools commonly associated with threat-based approaches, risk management instruments often include analytical models, prescriptive guidelines, and assessment matrices – each reflecting assumptions about the nature of the problem and how it should be managed.Footnote ⁴⁹ Indicators of riskification, alongside traditional threatification, can thus be identified and analysed, providing a more nuanced understanding of the specific effects generated by a securitisation process.Footnote ⁵⁰ See Table 2.

Table 2. Examples of threat-oriented versus risk-oriented securitisation.

Source: Created by the authors, building on Diez et al., Securitisation of Climate Change.

Our analysis below applies this supplemented version of Collective Securitisation that can account for interaction across organisations and measure securitisation in a more precise fashion. We study carefully where securitisation moves, audience acceptance, and new outcomes may be shaped by Collective Securitisation processes unfolding in other international organisations.

The UN and climate change adaptation

Research shows the UN has pursued a securitised approach to climate change generally,Footnote ⁵⁶ but this approach conforms mainly to a risk-oriented perspective rather than a threat-based one.Footnote ⁵⁷ To what extent do such findings carry over to the question of climate adaptation policy per se? Our analysis shows that the UN’s attention to climate adaptation, which preceded the EU’s in terms of depth and scope, was subject to a process of Collective Securitisation that played out over decades rather than years. Furthermore, we find that UN processes resulted in a specific kind of securitisation – namely, riskification – which reflected a specific set of discourses carried by certain actors and prioritising risk management tools rather than security management practices.

The starting point for applying the Collective Securitisation framework to the UN’s approach to climate change adaptation requires an understanding of the status quo prior to evidence of securitised outcomes. Before the 1980s, the UN’s engagement with climate change was growing, but its interest in adaptation as a response was marked by neglect, as climate change had yet to achieve public salience and other pressing global issues dominated its agenda.Footnote ⁵⁸ Political attention grew slowly through the 1980s, as climate change mitigation, associated with greenhouse gas emissions, climbed the international policy agenda.Footnote ⁵⁹ Even though attention to mitigation grew, adaptation was addressed only sporadically and within a small group of technical experts associated with the UN.Footnote ⁶⁰

For instance, the UNEP, together with the WMO and International Council of Scientific Unions (ICSU), organised an ‘International Conference on the Assessment of the Role of Carbon Dioxide and of Other Greenhouse Gases in Climate Variations and Associated Impacts’ in 1985 with scientists from 29 countries, including European states, participating. Uniquely for that time, the conference called for improved analysis of the impacts of climate change, including immediate assessments of the ‘widest possible range of social responses aimed at preventing or adapting to climate change’.Footnote ⁶¹ The 1987 Brundtland Report on sustainable development also served as a pivotal moment in the UN’s development of its adaptation policy. The report’s conclusions stated that several strategies must accompany climate stabilisation, including plans for adaptation to climate change ‘in the event that mitigation policies cannot be implemented rapidly’.Footnote ⁶² The report features language related to the security effects of climate change, arguing that the reduction of global warming was, then, an ‘essential task to reduce the risks of conflict’.Footnote ⁶³ Yet such language in documents from that period was rare. These years resemble a moment in time when no clear securitisation had yet taken place.

The situation changed in the late 1980s as scientific evidence accumulated, confirming that mitigation targets would be missed, and as belief spread among actor networks that effective mitigation was impossible.Footnote ⁶⁴ This rapidly growing realisation, reflected in our observation of a widening network of actors and a steep increase in press coverage,Footnote ⁶⁵ can be interpreted as a precipitating event that directed newfound attention to climate change adaptation, specifically. We find that discourses shifted from adaptation as a secondary concern towards an ‘urgent issue’ that must be addressed by ‘unprepared populations’ as the failure of climate mitigation grew apparent. This was the central theme emerging from a text analysis of events and reports held in the late 1980s, for instance, including widely attended events in 1987 such as the UNEP, WMO, and ICSU workshops on ‘Developing Policies for Responding to Climate Change’ (in Bellagio, Italy, and Villach, Austria) and, in 1988, at the UNEP and WMO Toronto conference on ‘The Changing Atmosphere: Implications for Global Security’.Footnote ⁶⁶

Not only was attention and analysis growing, but a clear securitising move can be detected. This followed on from a framing of climate change as an ‘existential risk’, yet one that, with anticipation and preparation, could be managed with proper analysis and attention. Along with growing networks of experts, policymakers, and a (initially slow but emerging) group of civil society representatives, focus was placed on ‘decision-making under uncertainties’Footnote ⁶⁷ identifying ‘possible’ and the ‘most probable impacts’,Footnote ⁶⁸ assessing risks, managing ‘anticipatory adaptation’,Footnote ⁶⁹ ‘minimizing adverse effects’, and avoiding ‘forced adaptation’.Footnote ⁷⁰ The risk framing of anticipation is particularly strong in official adaptation documents during this period (see Figure 3), corresponding with a belief that the world faced a plethora of problems linked to climate change, across societies, but that risk management tools could help avoid the worst.Footnote ⁷¹ In short, this marked the beginning of a risk-oriented discourse that would position adaptation as a central response to the escalating challenges associated with climate change.

Source: Created by the authors.

Figure 3. 10–year rolling average number of securitisation references in texts by the United Nations (1972–2023), coded in line with coding scheme in Annex 1.

Through the 1990s, the UN’s securitising move quickened as risk-related discourses grew and experts focused on developing technical tools as an apparent solution to the adaptation challenge. The first IPCC report, published in 1990, kicks off the decade, while the second report in 1992 codifies the risk-oriented discourses emerging from previous reports and conferences. This seminal report treats climate adaptation not as a normal public policy problem, nor as an existential issue requiring extraordinary action, but rather as an issue best managed through risk thinking. It encourages ‘immediate responses’ to the ‘inevitable danger’ of climate change and asks member states to focus on adaptation, including likely impacts, based on ongoing risk assessment and risk analysis, and by continuing research on risk impact.Footnote ⁷² The second World Climate Conference in 1990 was timed to coincide with that inaugural IPCC report and brought together the key actors grouping around the adaptation question. Actors arrived from UN bodies, including the UNEP, the FAO, WMO, and notably the EU, and reappeared in both conference proceedings and subsequent years’ texts in the form of recurring names and titles (see below).

The UN’s framing of climate change as a risk-laden issue requiring adaptation-focused strategies evolved significantly in the early 1990s, characterised by a shift towards tool-based methodologies and the institutionalisation of risk-oriented frameworks through the efforts of the IPCC and the UNFCCC. The IPCC advanced its work by establishing an expert group and issuing its ‘Preliminary Guidelines for Assessing Impacts of Climate Change’,Footnote ⁷³ the first significant step towards a tool-focused approach to adaptation rather than policy strategies. Two years later, the IPCC released its ‘Technical Guidelines for Assessing Climate Change Impacts and Adaptation’ (1994), which refined the earlier guidelines and provided detailed recommendations for conducting vulnerability analyses. These guidelines laid the foundation for what scholars later termed the ‘first generation’ of adaptation risk frameworks adopted by the UN.Footnote ⁷⁴ By 1997, the second meeting of the UNFCCC’s Conference of the Parties (CoP) requested the UN secretariat to ‘accelerat[e] the development of methodologies of adaptation technologies, in particular decision tools to evaluate alternative adaptation strategies’.Footnote ⁷⁵ A surge of guidelines and reports, many issued by the UNFCCC’s newly established secretariat, further cemented the use of risk-based language and analytical tools in adaptation efforts. Simultaneously, the network of actors addressing the issue expanded significantly, bringing in several increasingly active international organisations.

Alongside the growing use of risk-oriented language, UN bodies increasingly pushed for the adoption of specific models to identify and address adaptation challengesFootnote ⁷⁶ – a further indicator of a risk-oriented approach to securitisation (see Table 2). The UNEP, drawing on many of the same experts participating in the IPCC, and cooperating with the UNFCCC secretariat, developed in 1998 a ‘handbook’ containing guidance on the practical methods of impact and adaptation assessment.Footnote ⁷⁷ That effort would merge with work done by the IPCC in 1998, effectively consolidating the position of adaptation experts, requisite language, and a technological approach to solving the problem. These trends accelerated through the 2000s. By 2001, the IPCC meetings released their third assessment report focused primarily on adaptation. The risk language is remarkably consistent, with little to no threat-oriented language that a more classical securitisation approach would identify.

The evolution of methodological tools for addressing climate adaptation took a major step in the early 2000s when several UN bodies published what is considered the ‘second generation’ of adaptation risk frameworks, which focused less on top-down predictions of climate change impacts (including narrow cost–benefit analyses) and more on general risk management of physical and social impacts, using a bottom-up perspective.Footnote ⁷⁸ These reports move fully towards the principle of risk-based modelling for climate adaptation, urging members to apply those models as part of ‘National Adaptation Programmes of Action’ (NAPAs).

By 2007, and following vigorous debate over the various risk frameworks available,Footnote ⁷⁹ an integrated ‘risk management’ approach that incorporates local capacities, likely vulnerabilities, and potential impacts was adopted in the IPCC’s Adaptation Report 4, which constituted the third generation of adaptation risk frameworks adopted – and a major methodological milestone in the UN’s approach to adaption.Footnote ⁸⁰ The 1990s and 2000s thus showcased a prolonged but clearly discernible set of securitising moves, including continued use of risk-based terminology and the promotion of risk management tools and practices by actors coalescing around key UN platforms.

The Collective Securitisation approach posits that securitising moves are significant only when accepted by the intended audience, typically the formal decision-making body in an organisation.Footnote ⁸¹ However, audience acceptance of a security move is usually preceded by recursive interactions between securitising actors (usually taken to be bureaucratic actors and entrepreneurial experts) and the audience. By the mid- to late 2000s, evidence shows national leaders were adopting the language and tools associated with a risk-oriented version of securitisation.Footnote ⁸² A pivotal moment occurred during the first Security Council debate on climate change, highlighting a confirmation in the discourse shift and high-level approval of the risk approach to adaptation. The UN Secretary General’s High-Level Event in September 2007 brought the Security Council and General Assembly together to debate the exigencies of climate change a ‘growing and existential risk’ requiring urgent ‘risk management’ and adaptation – in contrast to threat preparation.Footnote ⁸³

Subsequent events in 2008, including discussions at the UNFCCC, led to the creation of the ‘Compendium on Methods and Tools to Evaluate Impacts of, and Vulnerability and Adaptation to, Climate Change’, endorsing ‘risk management’ as the ‘best’ decision-making framework.Footnote ⁸⁴ By the late 2000s, a risk-oriented approach to climate adaptation was the firmly established UN norm,Footnote ⁸⁵ exemplifying risk-oriented securitisation as it advanced to the new policy outputs stage in the Collective Securitisation phase-model. The central outputs are well documented in the UNFCCC’s own history of climate change adaptation policyFootnote ⁸⁶ and are reflected in the risk-oriented language present in the Paris Agreement on Climate Change, in which adaptation was included as a fully independent objective for the first time.Footnote ⁸⁷

The EU and climate change adaptation

We now turn to the EU’s own Collective Securitisation process, again employing our revised Collective Securitisation framework to explore how the UN and EU processes influenced one another as climate change adaptation entered the agenda. Until the mid-2000s, the EU paid little attention to climate change adaptation and exhibited no clear preference for either risk- or threat-oriented approaches. However, this began to change in 2005, following the UN’s audience acceptance moment, when a risk-oriented approach to climate change adaptation was adopted at the highest level. As we show, this development appears to have prompted the start of a securitising move in the EU in 2007. In doing so, the EU initially mirrored the UN’s risk-oriented approach, reflecting how Collective Securitisation processes can shape and intersect with one another. This section traces that evolution, unpacking the EU’s securitising move and subsequent steps to shed light on the interconnected dynamics of these processes.

Applying the Collective Securitisation framework, we begin by examining the EU’s initial status quo on climate change adaptation, which was characterised by non-securitisation and the absence of risk or threat language in EU texts on the topic during the 2000s (see Figure 4). This is explained mainly by the overall neglect of climate adaptation science and policy development. An exception is the work by officials in the Commission’s Directorate-General for Research (DG Research), which has an expert-led approach to investigating policy problems.Footnote ⁸⁸ Scientists from DG Research attended the early UN symposia on the matter (and even organised some of their own) in the 1990s. Several were involved in providing input into the early UN reports on the need to shift focus towards climate adaptation.Footnote ⁸⁹ For the most part, however, there is only scattered mention of adaptation in EU policies and no clear signs of securitisation. Striking evidence of this stems from the EU’s first major position on climate change, in early 2005, when it published ‘Winning the Battle against Global Climate Change’.Footnote ⁹⁰ This text focuses mainly on mitigation, with only a passing reference to adaptation.

Source: Created by the authors.

Figure 4. 10–year rolling average number of securitisation references in texts by the European Union (1972–2023), coded in line with coding scheme in Annex 1.

Between 2007 and 2008, however, the EU’s approach to climate change adaptation underwent a notable transformation, with key policy texts reflecting a sharp increase in security-focused language that balanced threat-oriented and risk-oriented discourses. A series of texts, including a 2007 staff working document prepared by the Commission, titled ‘Adaptation to Climate Change in Europe’, Footnote ⁹¹ and a 2008 publication by the Commission and Council together, titled ‘Climate Change and International Security’, signalled a sharp uptick in security-laden language, including references that suggest both ‘the inescapable threat of climate change’ requiring ‘extraordinary action’ alongside the need to ‘implement long term strategies’ to understand and act upon ‘the multiplying implications’ of climate risks.Footnote ⁹²

What happened in this two-year interval? Our evidence suggests that the UN’s high-level acceptance of securitising moves in 2007 (see previous section) served as a precipitating event for the EU’s own Collective Securitisation process. This interaction demonstrates the influence of cross-institution change mechanisms outlined in the theory section. The UN’s framing of adaptation as a risk management response to a securitised climate problem signalled a shift to the EU, altering the EU’s perceived pay-off calculations and encouraging policy mimeticism. This was further encouraged by growing stakeholder networks working across organisations.

Indeed it is no coincidence that shared stakeholder networks converged on a major EU climate conference in late 2006.Footnote ⁹³ Both UN and EU officials joined with climate scientists and activists to address ‘Future Climate, Impacts and Responses’, which emphasised the need ‘to improve Europe-wide risk, impact and cost/benefit assessment for adaptation responses, as compared with no action’.Footnote ⁹⁴ In a clear sign of the mechanism of emulation, UN risk management language and concepts such as ‘the goal of building societal resilience’, ‘calculating growing climate risks’, and ‘anticipatory governance’ were subsequently incorporated into the Commission’s 2007 staff working document,Footnote ⁹⁵ among other texts, likely facilitated by academic experts drawn upon by both organisations in an example of entrepreneurial networks crossing institutional boundaries (see discussion above, circa footnote 89).

From there, the EU engaged in its own set of securitising moves. Many of these actions correspond with the promotion of risk-oriented discourses and the same kind of risk management matrices established by the UN. However, in the early years of the EU’s securitising moves, we see not only risk concepts featured, but also those associated with threat-oriented versions of securitisation. The 2008 publication mentioned above, ‘Climate Change and International Security’, discusses climate change as a ‘threat multiplier’, i.e. a ‘threat to security’ in a plethora of ways, and emphasises the need for adaptation assistance.Footnote ⁹⁶ This threat-based approach to climate change generally, and adaptation specifically, is unique to the EU; it can be explained partly by the position of a High Representative for the EU’s Common Foreign and Security Policy and, later, the role of the EU’s security-oriented External Action Service in EU policy outcomes. The EU’s contribution to a UN Secretary-General Report on ‘Climate Change and International Security’, in 2009, showcases some of the same perspectives, including the treatment of climate adaptation as a way to downgrade ‘the security threats created by climate change’.Footnote ⁹⁷

For the most part, however, the securitising moves initiated by EU actors reflected risk thinking (see Figure 4). The preparation for the 2009 climate change negotiations in Copenhagen (CoP15) resulted in a flurry of texts, balancing the EU’s extant focus on mitigation with equal treatment to adaptation. The EU’s own negotiation position, written by the Commission with member state input, called for an ‘Adaptation Framework’ to be agreed.Footnote ⁹⁸ Soon thereafter, the first adaptation-specific policy statement emerged from the Commission in the form of a 2009 White Paper on ‘Adapting to Climate Change’.Footnote ⁹⁹ A text analysis of this document shows significant overlap with the UN’s UNFCCC documents of that time, including risk-based language, proposals for specific kinds of impact and vulnerability analysis, and other risk management tools. An internal working paper in the Commission in 2009 provided the basis for the more formal ‘Impact Assessment on Adapting to Climate Change’ requirement, which demands adaptation to be measured and categorised according to risk management matrices.Footnote ¹⁰⁰

These moves mirrored the ‘third generation’ of adaptation tools propagated by the UN, drawing on the UNDP’s 2008–9 guidelines on ‘Designing Climate Change Adaptation Initiatives: A UNDP toolkit for practitioners’ and successive IPCC reports on climate change (combined and republished in 2010).Footnote ¹⁰¹ The IPCC’s 2012 special report advises all member states to reduce uncertainty about adaptation by applying risk management measurement tools. Titled ‘Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation’, it suggests integrating climate risks into decision-making processes with tools like decision support maps, predictive instruments, scenario planning, and scenario exercises.Footnote ¹⁰² This approach was adopted almost entirely by the European Commission’s first formal proposal for an EU-wide climate change adaptation strategy in 2013. The document parrots the verbiage used by the UNFCCC and IPCC, with references to ‘managing uncertainty’, ‘preparing for the inevitable’, and the criticality of ‘investing in … risk assessments and tools to build up capacities for adaption’.Footnote ¹⁰³ One interviewee confirmed that the EU ‘used’ many of the UN texts, especially IPCC reports, since ‘the process behind the IPCC reports is solid and sound and there is a political agreement by all the parties in the reports’.Footnote ¹⁰⁴

When a Commission proposal for legislation is formulated, generally speaking, supranational officials engage closely with national officials to (informally) test the boundaries of political acceptance. In this case, the Commission gathered key figures from the EU’s national governments (partly via its expert group on climate change) along with officials from the IPCC and the European Environmental Agency, in 2012 and 2013, in a series of meetings and workshops to help ‘sell’ the contents of a climate adaptation strategy.Footnote ¹⁰⁵ Analytically, this process resembles one of recursive interaction, when, according to the Collective Securitisation framework, securitising moves are accepted or rejected by the relevant audience. Such interaction was on full display as the Commission and member state officials worked almost side by side to devise the EU’s strategy for climate change adaptation.

There are two notable features of this process. The first is the direct emulation of much of the UN’s risk-related discourses (as shown by text analysis, see Figure 5) along with risk management tools, described earlier. A common finding in our text analyses is the importance of applying an ‘asset-based’ needs assessment. In other words, the technological tools propagated in this process were very much part of the securitising move and the object of discussion in the process of recursive interaction. The second feature is the broadly consensual nature of this interaction, since’the usual suspects’Footnote ¹⁰⁶ engaged in this interaction, including EU member state officials, EU supranational policymakers, and experts – many of whom also engaged in the earlier UN securitisation processes and represent the kinds of networked actors theorised earlier. As the Collective Securitisation framework suggests, within international organisations it is often supranational officials as ‘securitisers’ and national representatives as ‘audience’ working intimately in a process of recursive interaction.

Source: Created by authors.

Figure 5. 10–year rolling average number of risk-based securitisation references in texts by organisation (1972–2023), coded in line with coding scheme in Annex 1.

Unsurprisingly, perhaps, the acceptance of the securitising act can be documented soon thereafter. Two formal acts connote high-level acceptance. One was the Council’s approval of the Commission’s proposal in the form of an ‘EU Strategy on Adaptation to Climate Change’ (2013),Footnote ¹⁰⁷ thereby signalling member state approval both formally and informally.Footnote ¹⁰⁸ That document contains very similar texts that take a risk-based perspective on adaptation challenges and promotes the same kinds of technical approaches for managing those challenges. Another act was the Council’s adoption of a policy on ‘Climate Diplomacy’ (2013), which argues that tackling climate adaptation challenges is ‘a necessary condition for peace and security, development, and prosperity’.Footnote ¹⁰⁹ Such language reintroduces the threat-based language that appeared years earlier (an effect of this document being produced mainly by the European External Action Service, which was temporarily involved in the debate) and of which we find some evidence in subsequent UN documents at the time – a seeming case of ‘recursive interaction’ as proposed by our framework.Footnote ¹¹⁰ The preponderance of evidence, however, points to the acceptance of a risk-oriented version of securitisation taking place in the EU regarding climate adaptation policy.

With this acceptance came a new normal in the EU’s Collective Securitisation process, one characterised by a close alignment of UN and EU efforts. In 2013 and 2014, the UN and EU collaborated to distil existing adaptation research into actionable guidelines under various dissemination programmes, including the ‘Provia’ initiatives,Footnote ¹¹¹ addressing vulnerability and impacts in a widely adopted ‘User’s Companion’. Simultaneously, the EU’s ‘Mediation’ project devised methodologies for effective climate decision-making. These guidelines were consolidated in Provia’s 2014 ‘Support for NAP Development with the Provia Guidance’, offering risk management models for national and local officials. The EU and UN jointly supported Adaptation Futures conferences from 2010 to 2021, indicating an expansion of risk management models globally. In the latter years of the 2010s, more collaborative efforts, such as the 2017 EU/UN Partnership on Climate Change and Security,Footnote ¹¹² emerged to combat climate change and promote adaptation projects abroad, with funding contingent on adopting the EU and UN’s risk management analysis frameworks.Footnote ¹¹³

However, in 2018, threat language found its way back into EU documents as the EU’s previous adaptation strategy was reviewed and updated. The EEA’s evaluation of the old strategy, for example, states that the security angle of adaptation needs more focus, considering how ‘international climate adaptation issues’ are closely linked to ‘migration risks, risk to global value chains, and security risks’.Footnote ¹¹⁴ This type of language, reflected in several EU documents of that period, is more explicit and more directly linked to security per se than the more subtle risk discourse identified previously. The Commission’s own review of the strategy comes to a similar conclusion, arguing that ‘Europe is vulnerable to climate change impacts beyond its borders through, for example, trade, international financial flows, migration, and security’.Footnote ¹¹⁵ A tension between climate change adaptation at home, and adaptation abroad, is more pronounced for the EU and can be attributed to growing geopolitical conflict on the EU’s borders.Footnote ¹¹⁶ See Figure 6 for a depiction of the EU leading the UN in threat language, although at a much lower overall rate compared to risk language in general.Footnote ¹¹⁷

Source: Created by the authors.

Figure 6. 10–year rolling average number of threat-based securitisation references in texts by each organisation (1972–2023), coded in line with coding scheme in Annex 1.

Indeed, subsequent climate adaptation policies in the EU and UN return to a focus on earlier risk management methodologies. The EU refined its methods using guidelines from UNFCCC, IPCC, and UNDP, consolidating tools under the Climate Change and Vulnerability Assessment (CCIV) label. These CCIV tools involve assessing current and future climate conditions and potential impacts on vulnerable sectors, and analysing underlying factors influencing climate risks.Footnote ¹¹⁸ In the mid-2010s, there was a debate on international modelling tools, but by 2018, focus converged on a specific set of tools for measuring, prioritising, and addressing adaptation needs. Notably, collaboration on adaptation modelling intensified through joint efforts of the UNFCCC, EU (especially the European Environmental Agency), and the International Standards Organization (ISO) in workshops and conferences from 2018 to 2020. An interviewee described this intersection as a ‘feedback loop’ between the two IOs within which individual EU member states can raise adaptation issues at the UN and then issues ‘come back’ to the EU and vice versa.Footnote ¹¹⁹ France, Germany, and Sweden tend to be the most engaged members states, although our research shows that groups of individual bureaucratic and academic actors tend to be the conduits for these securitising dynamics.Footnote ¹²⁰

In a clear instance of cross-organisational influence, the CCIV research and emerging EU approaches for measuring adaptationFootnote ¹²¹ shaped the ISO’s 2021 ‘norm’ for climate change assessment protocols.Footnote ¹²² This ISO norm specified risk modelling parameters, emphasising a shift towards ‘asset-level modeling’ for states addressing prominent climate-related issues. Such modelling approaches were then reflected in EU and UN policy cooperation by the early 2020s. The EU’s second ‘Strategy on Adaptation to Climate Change’ (2021) exemplifies this, focusing on improving the policy cycle through enhanced knowledge and data, support for policy development, and accelerated adaptation action.Footnote ¹²³ The strategy underscores the importance of using ‘state-of-the-art adaptation modeling and risk assessment tools’, aligning with the 2013 strategy’s emphasis on asset-level modelling. This risk-oriented approach became the foundation for subsequent ‘action plans’ and work programmes. In contrast to earlier threat-focused language, the 2021 EU strategy adopted more risk-oriented discourses, viewing adaptation efforts as a series of proactive measures rather than a one-time emergency response.Footnote ¹²⁴

In short, our data reveals clear intersections of UN and EU Collective Securitisation processes, with the EU’s process staggered temporally, kicking off roughly halfway through the UN’s process in 2007. That intersection seems to be led through several of our hypothesised dynamics, including the UN’s audience acceptance serving as the EU’s precipitating event and the diffusion and use of certain tools and the cross-institutional interaction of key actors. See Figure 7 for a representation of how the respective Collective Securitisation processes overlap.

Source: Created by the authors.

Figure 7. Collective Securitisation of climate change adaptation in the UN and EU.