Abstract
Britain’s strategic defence and security reviews are regularly characterised as failures. This article presents original empirical data from UK policymakers and officials about how they account for this perception. Through extensive interviews and focus groups with key participants at every level of the review process, the authors identify seven key pathologies in British defence and security policymaking. Weak central direction, a disconnect between the centre and individual departments, a disconnect between departments across Whitehall, the clustering of decision-making, a lack of alignment between budget and ambition, a lack of attention to implementation, and widespread deception, are all perceived to combine to inhibit effective policymaking. In light of these insights, the authors posit a new analogical framework, the ‘jellyfish model’, in which fluid clusters of decision-makers converge to make important choices separate from supposedly core sites of strategic direction. This captures the policy process more accurately than existing rational, linear models.
All British defence and security reviews end in failure. This is the perception in most commentary on British defence policy. For decades, successive governments are seen as failing to match the ambitious goals set out in review documents with the resources needed to achieve them (Barrons, 2024; Cornish and Dorman, 2011; Curtis, 2024). Yet, for much of this period, the UK defence budget has been high in comparison with other middle powers such as France, Germany and Japan (SIPRI, 2026). As such, there is also criticism that reviews are failing to leverage what resources they are given as efficiently and effectively as other states (Stringer, 2026: 9).
This article seeks to understand how successive reviews appear to be caught in a cycle of failure, despite different personnel, different logics, different circumstances and different outputs. It is important in practical terms as the United Kingdom is experiencing a high and increasing level of threat from hostile state and non-state actors, according to UK security and defence officials (Knighton, 2025; Metreweli, 2025). In a difficult financial climate, the UK government needs to ensure it is organising its security and defence appropriately to meet these challenges. Yet, only 6 months after the 2025 Strategic Defence Review (SDR), analysts described a ‘dangerous defence vacuum’ (Lyman, 2025), with an apparent funding gap of £28 billion and the review authors criticising the pace and quality of implementation (Economist, 2026, Martin, 2025). Despite the Prime Minister Sir Keir Starmer’s assertion that security and defence should be ‘the central organising principle of government’ (Starmer, 2025), it remains dysfunctional.
The question of why these reviews fail is also important for a theoretical understanding of defence and security governance and how states provide security. On the surface, a strategic defence review should be the epitome of the unitary state in action. The highest purpose of the state is to provide for the physical security of its citizens. Ideally, reviews are meant to ensure this by bringing together stakeholders, identifying national security threats, assessing defence commitments and capabilities and aligning resources and effort to achieve national goals. If this is not functioning in the UK case, it is important to understand why and what this means for existing models of defence policymaking.
Existing analyses of British defence policy are almost entirely atheoretical. Following each defence and security review, commentary will discuss policy outcomes and critique political choices but rarely attempt to theorise the processes under scrutiny (Boulton, 2021; Chalmers, 2016; Codner, 1998; Magill and Rees, 2022; Strachan, 2009). This work is empirically rich but grounded in the specific experience of the review in question. As such, it resists generalisable observations. Similarly, a number of excellent studies have explored defence policy and strategy from an empirical perspective without attempting to model or theorise the review process (Clarke, 2018; Curtis, 2024; Louth and Taylor, 2019). A rare exception is Cornish and Dorman’s four stage cycle of defence reviews, describing them as moving through the phases of failure, inertia, formulation and misimplementation (Cornish and Dorman, 2010).
In addition to this theoretical gap, what is missing from existing literature on British defence policy, as a number of studies attest, is analysis of the internal dynamics of how British defence and security reviews are formed, interpreted and implemented (Grattan et al., 2015; Nemeth and Dew, 2024; Uttley et al., 2019). Some work has engaged with British strategic culture, identifying the influence of long-standing attitudes and practices such as ‘muddling through’ and ‘incrementalism’ (Cornish and Dorman, 2012). Other studies have engaged with New Institutionalism to note path dependencies from prior policy assumptions, or employed organisational theory to understand the Ministry of Defence’s (MOD) role in reviews (Uttley et al., 2019). There are also analyses of the impact of specific organisational changes like the introduction of the National Security Council (Devanny, 2015). Most of this research involves external commentators interpreting public documents, rare exceptions being Thomson and Blagden’s interview research on formal and informal institutions (Thomson and Blagden, 2018) and Mark Phillips’ insider account of the 2010 review (Phillips, 2012).
As such, this article makes two important contributions to understanding. First, it provides extensive first-hand empirical data on the internal workings of security and defence review processes, as understood by those within the system. This thick description (Geertz, 1973) offers original insights into practices that have hitherto been unknown to the wider public. Second, it uses this original empirical data to propose a descriptive analogical model, ‘the jellyfish’ that captures the strategic process as conveyed by practitioners, particularly its noncognitive, irrational elements. In the process, it makes a new theoretical contribution to defence and strategic studies.
According to our analysis, policymakers perceive strategic reviews as failing due to the way defence and security policy is made in the United Kingdom. As the following data and analysis will show, a series of recurring pathologies inhibit effective policymaking. When taken together, they suggest a lack of coherence and rationality in the policy process. This is, we argue, problematic for many existing analytical frameworks, which assume a degree of cohesion and linearity in decision-making and rationally motivated behaviour among participants.
In place of these models, we advance a neural understanding of policymaking. Using the metaphor of a jellyfish, the authors posit that British defence policy is surprisingly decentralised, with quasi autonomous elements converging to make important choices, often separate from supposedly core sites of strategic direction. Although review documents draw together the disparate elements and posit an ‘integrated’ whole, the way decisions are reached is far more redolent of an organism with no central nervous system, such as a jellyfish. This has implications both for the theoretical understanding of policymaking generally, and for future defence and security policy in the United Kingdom in particular.
Methods
Defining a policy as a failure is an interpretive act (Bovens and t’Hart, 1996; Oppermann and Spencer, 2016). While empirical metrics may indicate that a review has not been carried out (delays in procurement, failure to achieve a desired outcome), or not been achieved efficiently (cost overruns, duplication of effort), whether these amount to failure is a subjective judgement based on how the review overall is interpreted. Defence and security reviews are complex processes with multiple facets. Interpreting them as a failure depends on how their intent is defined, which aspects are seen as vital and which peripheral, what scope for agency is identified, what policy levers are available and deemed appropriate, how far desired aims are perceived as achieved, what intervening factors apparently shaped outputs and how outcomes are understood.
Given that weighing all these matters requires interpretation, our research approach is informed by Interpretivism. This tradition emerged from hermeneutics (Gadamer, 1996: 164–169, 505–541), the philosophy of social science (Bevir and Blakely, 2018), social constructivism (Hopf, 2002; McCourt, 2016) and historical method (Carr, 2008; Collingwood, 1972). Interpretivists emphasise the importance of meaning; how it is constructed, interpreted and operationalised in textual and linguistic expression. They see reality as constructed socially through texts (speeches, documents, interviews, etc.), which articulate the beliefs of participants in particular practices (Beech and Bevir, 2024).
While external commentary has highlighted sites of failure in reviews over decades (Jordan, 2019; Kaldor, 2025), an important missing component of much existing interpretation is the internal views of participants. By contrast, in strategy literature in other areas, it is the perceptions of participants in the strategy process which are used as the yardstick of success or failure (Noble and Mokwa, 1999). For that reason, we focused our attention on ascertaining the beliefs of individuals within the UK defence and security policy machinery who had direct experience of the conduct of these reviews. The key contribution of this article is that it presents the first ever substantive analysis of how these processes are interpreted, enacted and evaluated according to the first-hand accounts of those responsible for formulating them.
The data for this research emerged from a study, funded by Research England, looking into the conduct of UK strategic defence reviews from 1998 to 2023. 1 Rather than focus on any one review, our aim was to understand long-standing patterns of beliefs and perceptions of failure which might point to endemic issues in the policy process. Data collection was carried out between March and December 2024 and comprised N = 60 semi-structured interviews with officials, Ministers and armed forces personnel. Participants ranged in seniority from secretaries of state, senior foreign policy advisors and permanent secretaries to programme leads and lower-level officials. The authors engaged in purposive sampling (Guest et al., 2006: 61), selecting interviewees on the basis of institutional involvement in reviews (as identified in departmental reports, organograms, evidence to parliamentary committees and informed commentary). They then utilised snowball sampling until they reached saturation – where themes were repeated and new insights were limited (Parker et al., 2019).
To balance individualised accounts in interviews, the authors sought a collective sense of the formation and execution of the reviews via three focus groups conducted with senior former and serving officials in April, May and June 2024. 2 Each group comprised N = 10 participants identified in interview data as playing pivotal roles in review processes. They were asked open questions about the context, formation and execution of the reviews and allowed to engage in free discussion in response.
When it came to data collection and analysis, the project team adopted an inductive approach (Hopf, 2002), avoiding hypotheses in the early stages, following Pratt et al.’s (2020: 7) direction that ‘if researchers look only for what they are seeking, they are not doing good inductive or abductive qualitative work’. Rather, through a developing process of naïve coding, constant comparison, thematic ordering and conceptualisation, based on standard grounded theory (Schuler et al., 2023), the researchers identified themes in the data which might account for the perception of policy failure. We then go on to utilise these insights for subsequent abductive theorising (Bellis et al., 2024; Bourgeois and Brodwin, 1984; Noble and Mokwa, 1999), linking these findings with existing models and theories about defence and security policy.
All participants have been anonymised in an agreed form and the study was conducted in accordance with the research ethics policy of King’s College, London and their guidance on research using human subjects. The reviews considered are listed here.
UK Strategic Defence Reviews 1998-2023
In our research, participants in the study identified seven pathologies as contributing to the failure of strategic security and defence reviews. These were weak central direction, a disconnect between the centre and individual departments, a disconnect between departments across Whitehall, the clustering of decision-making, a lack of alignment between budget and ambition, a lack of implementation, and deception/dishonesty. In combination, they both describe the dysfunction of UK defence and security policy and point to a new model for understanding state behaviour which may have generalisable applications to other contexts. We will set out each of these in turn before reflecting on what they mean for our understanding, and modelling, of defence and security policy.
Weak central direction
One of the primary rationales for strategic defence is to reassert central direction and control over defence structures. As a former National Security Adviser put it, they are meant to articulate: how we organise ourselves . . . we have a mixture of interests, defence and security, economics and trade, influence and soft power. You have to be able to use all of the levers against all of the objectives. That’s what these reviews are fundamentally about. How do you assemble all of that for a country the size of the UK? (Interviewee 1, 2024).
They are supposed to offer a moment to consciously pause, reflect and ‘ask what is the overall shape, what is the overall orientation of this?’ (Interviewee 1, 2024). Out of that process, review authors attempted to construct a unifying narrative. As one put it, ‘one thing we were very clear about, we weren’t going to put together a mishmash, we were going to have a coherent story . . . [although] some of the bits had to be shoehorned into the story to fit’ (Interviewee 2, 2024).
In the period between reviews, many commentators note the risk of organisations experiencing strategic drift, with budgets being extended, commitments expanding and understanding of the ‘threat environment’ suffering from inertia (Cornish and Dorman, 2011; O’Rourke, 1972: 52–53). Yet, interviewees expressed the frequent lament that review processes did not establish a clear strategic baseline for their deliberations. For instance, of the 2010 review – produced in 5 months by a newly formed coalition government – one respondent argued, ‘Did Peter [Ricketts] or did anybody else, (David) Cameron, new Prime Minister, set out a really clear, articulate vision for how we wanted this thing to end up? It honestly didn’t feel that way to me’ (Interviewee 3, 2024). Even with the Integrated Review of 2021, which is often viewed more favourably in terms of its framing of grand strategy, respondents suggested a lack of a strong central steer on strategy (FG3, 2024).
This lack of central direction carried over into the review process itself. Once reviews were announced, the Prime Minister would often step back and the review would be undertaken with little or no guidance from them about priorities. 3 As a senior member of the IR21 review team noted ‘I never felt there was a point in the process where the centre was the authority’ (Interviewee 4, 2024). The exception for them was ‘when we were in the final throes of drafting it’ and choices had to be approved (Interviewee 4, 2024). At the departmental level, one Foreign Office official indicated, ‘I don’t remember getting any particular direction from Foreign Office Ministers, from the Foreign Secretary or the Ministerial team, or indeed very much from the Board at that time – the PUS [Permanent Under Secretary] and the DGs [Director Generals]’ (Interviewee 5, 2024).
Some effort was made to introduce greater coherence to later reviews, via a ‘fusion’ approach, but as we will see below, this encountered significant resistance. The public nature of the final document was also perceived to be a hindrance to strong central articulation of priorities. Review texts are politically sensitive and read by allies, competitors and enemies alike. As a consequence, participants noted a high degree of circumspection from the centre about choices and the reasoning behind them. A number explicitly called for a parallel secret review process and document which could be more forthright. In its absence, an MOD official argued ‘there was never a classified national security strategy that went into the detail that you needed to understand how all the bits fit together’ (FG3, 2024). The 2025 SDR tried to give strong central direction through a process innovation (an external review team); yet, it provided no recommended force structure, no specific capability decisions and experienced the same budget problems of previous reviews (Economist, 2026). As such, while it did provide (external) central direction in the process, there was a near-absence of directed output.
Lack of direction from the centre is vitally important. Given that it is a primary rationale for having a review in the first place, the repeated references to weak central guidance offer a strong explanation for why most reviews are perceived as failures. Furthermore, it tells us something important about the overall process of making defence and security policy. Rather than a rational delineation of defence and security priorities by a core ‘brain’ located at the centre of government, the picture offered by participants is of a far more fragmented process whereby the review is constructed iteratively by multiple actors, the centre being one. The weakness of the centre’s input into review outputs is compounded by a perceived disconnect between the centre and other departments.
Disconnection between the centre and departments
A consistent theme of discussion as to why defence and security reviews fail is the perceived disconnection between the centre and the departments feeding into each review. Three aspects of this disconnect stand out: a lack of clarity in the concepts employed by the centre, the belief that the centre is out of touch with everyday practice, and institutional pushback against efforts by the centre to integrate policymaking.
In terms of concepts, one official suggested of the 2015 National Security Strategy/Strategic Defence and Security Review (NSS/SDSR) that one thing that I think surprised me, maybe my own naivete going into this, I kind of imagined there would be a long period of, if you like, workshopping of ideas and concepts before we got to the writing stage. In fact, the review was the writing. (Interviewee 5, 2024)
In other words, it was left to review authors to do the conceptual work, rather than Ministers. This lack of open discussion meant that the central direction provided by reviews was frequently couched in language that was not explained or understood by its recipients. As one official noted of the concept of ‘persistent engagement’ in the 2023 MOD defence review: despite the fact we published our Defence Command paper with it referenced something like 16 times . . . we’d never actually find what it was. Same with “strategic advantage”. We’d used the term 20 odd times in the IR and Defence Command Paper and we had no idea what it actually meant in practice. (Interviewee 6, 2024)
Thus, concepts were frequently defined after they had already been employed.
One interviewee blamed this on the study investigators: ‘I blame academics. Actually, I blame you for wanting to attach labels and categorisations to things which policymakers then react to and go: “Yeah, absolutely. Right. Well, I need to invent one in that case”’ (Interviewee 7, 2024). There is evidence to support this assertion. One senior policymaker said of an equipment programme ‘we’re going into this defence review. I’ve got to think of a way that we can use it. So again, I came up with an army concept called (deleted for anonymity) . . . The thing that captured the imagination the day the defence review was published was (deleted for anonymity), much to the irritation of everybody else’ (Interviewee 8, 2024). 4 Thus, a political incentive is identified to coin new phrases and concepts, but due to their novelty and ad hoc formulation, they are ambiguous and can create confusion among officials who then have to define and implement them (Rauta and Monaghan, 2021: 489).
Indeed, there was extensive evidence in the data of general confusion about the rationale for decisions and overarching logic of the review process across this time period. One former senior MOD official in focus group 1 noted that when he was tasked with undertaking a helicopter study as part of a review: ‘I had been given no warning that the task was to be given to me, I was not even aware that a support helicopter study had been created, nor why I had been chosen to do it’ (FG1, 2024). Focus group activity was useful for interrogating the underlying logic of such decisions. In the case of the helicopter study, other participants indicated personnel were sometimes allocated to roles if they were not perceived to have ‘skin in the game’ and so could evaluate choices dispassionately (FG1, 2024). Yet, even where central direction might have had a rationale, the lack of communication and disconnection between the centre and the other parts of the system undermined its coherence in practice.
Even senior figures questioned how far choices from the centre connected with actual policy outputs. One interviewee argued, I do feel strongly about the need for the Ministry of Defence to have more strategic control over its TLBs, top level budgets, come the Service HQs, because I don’t think you will get effective change, or implementation of change, unless that is the case. And I think they’d lost the ability to be able to do that. I don’t think they really know what’s going on at grassroots level any longer. (Interviewee 8, 2024)
In other words, in centralising decision-making and removing the agency of the individual services, this interviewee suggests that a gap had emerged between the centre and the ‘grassroots’, resulting in a lack of effective change. Yet, over the period of study, budgetary control had been centralised, delegated and centralised again many times, while the perception of poor central coordination and policy failure persisted.
The review writing process itself is often conveyed as reinforcing this sense of disconnection. As one former senior MOD official noted, ‘If I’d ask somebody “how do you think it’s going?” in all, most all, of the MOD, the response would be “No idea, how would we know?”’ (FG1, 2024). Negotiations over the content of the review, and especially final budget decisions, were necessarily restricted. When it came to formulating the review documents, a senior FCO official involved in the 2015 review argued, the trick lay in getting the document to a stage, and getting the overall strategy to a stage, where the financial details were feeding into it, and it was all cohesive, when actually bits of the machine didn’t know what stage those negotiations were at. (FG2, 2024)
As one of the review’s authors, they felt they had managed to achieve a cohesive outcome which linked to wider strategy. Yet, there was a frequent lament from other focus group 2 participants that the obscure and patchy way in which decisions were made meant that it was not clear how decisions had been arrived at, who had ownership of them, and how they should be prioritised (FG2, 2024). Rather than being integrated and coherent, relations between the centre and individual departments were conveyed as sporadic and fragmentary.
However, it would be wrong to frame the lack of central direction and control as entirely a top–down error. Across our data, there are numerous allegations of individuals and teams resisting influence from the centre. As an example, interviewee 17 noted pushback at efforts to implement the fusion doctrine: ‘that got fought tooth and nail across Whitehall because the last thing that anybody wanted was an approach that we all agree on because it forces you to cohere and confront issues’ (FG3, 2024). Defining a doctrine at all was seen as problematic as it might prompt change in practice. In its absence, attempts to promote cooperation at the Cabinet Office faltered because ‘lowest common denominator compromise is the best that can be achieved, and “consent and evade” begins immediately after the meeting ends’ (Interviewee 17, 2025).
Moreover, the system actively seemed to incentivise departments to buck efforts to integrate decision-making. Of a recent review, an interviewee argued, The review didn’t feel like we were part of a cohesive whole of government effort. It was competitive against other departments, ‘there was a certain amount of money, we’d go get it’. There was a strong narrative of: ‘We can’t expose our thinking too early, we can’t let the Cabinet Office or others look in too early because they’ll unpick it’. We incentivised people to behave like that, those who led that approach were considerably rewarded. (Interviewee 6, 2024)
Given the Cabinet Office was meant to co-lead the review alongside a team located in the Prime Minister’s office, this represents a rejection of the hierarchical structure of the review process, as well as the premise of a ‘whole of government approach’ which is supposed to govern national strategy. This leads us to our next policymaking pathology: a disconnection between departments.
Disconnection between departments
Participants in defence reviews frequently highlighted a lack of cooperation between the agencies of government. As one senior official asserted, ‘Defence is made up of a number of different tribes, who don’t necessarily collaborate and work together’ (Interviewee 9, 2024). Yet, the drivers of this disconnection appeared to be senior politicians. From a Foreign and Commonwealth Office (FCO) perspective, the Foreign Secretary, Dominic Raab, apparently prevented his officials from engaging directly with those of other departments who were supporting the Integrated Review in 2020–2021. 5 At the same time, Ben Wallace, the Defence Secretary, had supposedly stopped the MOD from sharing information with the review team as part of a deliberate strategy to engage in personal bargaining with the Prime Minister over the defence settlement (Focus Group 3, 2024). Even when MOD officials did try and engage with Cabinet Office colleagues, they noted ‘there was no consultation, no reaching out’ from the Cabinet Office in return (Interviewee 9, 2024).
In that sense, efforts to provide an integrated and coherent policymaking structure to reviews, centred round the ‘fusion doctrine’, failed. Rather, individual Ministers and departments seemingly bargained for influence and budgets in a zero-sum Whitehall game.
Yet, there are qualifications to this picture. One is the suggestion by a senior leader that the various branches of the armed forces were, by and large, not engaged in ‘beggar thy neighbour’ approaches to the review. Many senior military figures had apparently spent much of their careers in joint environments such as Permanent Joint Headquarters and Joint Forces Command and so their attachment to a specific service was lessened (Interviewee 9, 2024). In that sense, the infighting was more interdepartmental than interservice.
In addition, the fact that there was tension between departments does not by itself mean that any one department would win out and be decisive in outcomes. In fact, the process of arriving at choices was far more diffuse, according to participants in the study. Many noted that decisive moments were brought about by very small numbers of individuals working in ad hoc groupings made up of more than one department – rather than the monolithic impression of big departments fighting it out via senior leaders. Thus, another feature of UK policymaking is the importance of decision-making clusters.
Decision-making clusters
A strategic defence review generally takes between 9 and 14 months depending on how one calculates the start and end dates. The review team profess to consult widely (with the 2025 review purporting to have solicited 8000 submissions from the public) (Robertson, 2024). The length and breadth of this process might suggest that decisions are made on the basis of a cross-Whitehall process encapsulating the collective interests of the unitary state.
Yet, the descriptions offered by participants indicate a far more ad hoc and fragmented process of decision-making. For example, the 2010 review was seen by a former senior MOD official as ‘a lot of individual reviews that were kind of pulled together’ (FG2, 2024). Instead of the state (or even individual departments) acting in a unifying manner, the Defence Secretary of the time, Liam Fox, was said to be ‘actually running a group inside the Ministry of Defence looking at all the big defence decisions before they went into the centre’ (FG2, 2024). Rather than a sense of a governing rationale or strategic sensibility, many reviews seemed to be shaped by choices made by shifting clusters of decision-makers. As a result, an official argued, ‘we were not making sensible strategic choices about the custodianship of a force over time. We had a bit of a mishmash mess’ (Interviewee 9, 2024).
When negotiations happened between departments, especially with the Treasury, they were apparently undertaken by very small teams of people, which a former senior official in the UK security community labelled ‘apertures’ (FG2, 2024). These groups were the ‘bit in-between all the people who need the money and all the people who are going to decide who gets the money’, and their ability to control the flow of information meant they became ‘critical’ (FG2, 2024). Yet, they were not publicly known, tended to be described in vague terms, and changed according to the review or even the issue at hand.
One consistent factor noted was the last-minute nature of decisions. Despite the public attempt to portray reviews as an extended rational process of resource allocation, it was often suggested that review documents were authored by small writing teams to very tight deadlines – in the case of the 1998 SDR, supposedly over the course of a weekend; for the Integrated Review, a few weeks (FG1, 2024; FG3, 2024; Interviewee 4, 2024). Final budget settlements would be arranged at a late stage between small groups in the Prime Minister’s office and the Treasury, whereby ‘magically money appears’ or gets taken away (FG2, 2024). In response to efforts to reduce budgets at a late stage, officials would often promise to make efficiencies (storing up issues for later, as we explore below). This process was necessarily restricted to a tight circle of decision-makers and could not be based on wider consultation or evidence of feasibility. Even in the 2015 review, which officials saw as very consultative, a FCO official noted ‘in the final fortnight or something like that, it basically went dark. Basically, the lead official at the Cabinet Office said, “right, thanks everybody, it’s now us and Number 10”’. As a result, the official suggested ‘it felt like one integrated thing except for particularly that final bit when we were told to keep our distance’ (Interviewee 5, 2024).
The fragmentary nature of decision groups carried through to their interpretation by officials. As one respondent recalled ‘the lack of clarity meant that decisions were interpreted by different people through their own particular lens’ (Interviewee 10, 2024). In other words, rather than a unitary state making rational choices based on the calculation of interests, or departments looking to protect budgets or advance their position within the policy ecosystem, we are presented with a kaleidoscope of shifting configurations of decision-makers. The strong sense is of a policy organism with no central nervous system. Instead, different parts of the state make choices in response to immediate stimuli.
Budget and ambition not aligned
A key point of failure repeatedly identified in the data was that budget and ambition were not aligned (Chalmers, 2018; Clarke, 2018; Curtis, 2024: 144; National Audit Office (NAO), 2023: 7). This creates its own set of dysfunctions in subsequent policymaking. Given the lack of central direction, confusion over purposes, and poor coordination of decision-making noted above, any disconnect between budget and ambition has to be resolved by officials and service personnel. Yet, a former senior Royal Navy officer observed, ‘if you don’t link the money to the decisions that you’re going to go and implement, when they get passed down to the front line commands and the delivery agencies and so on, then actually what you do is you end up in this really difficult spot’ (FG2, 2024). The exemplar of this problem is the UK’s aircraft carriers. Having originally envisaged only operating one carrier, with the other in reserve, the Royal Navy was subsequently tasked with running two but not given more money (Interviewee 11, 2024).
The influence of the Treasury is often cited by commentators as a source of failure, inhibiting long-term investment and planning (Hampshire, 2026; Smith, 2022: 8). Curiously, the data from interviewees was ambiguous. Some interviewees indicated that the Treasury put pressure on budgets which stored up problems for the future. As one put it, The Treasury had a mantra when I was there, and it was always that ‘Defence was good for a billion’. And that meant, effectively, you could always get a billion pounds out of defence . . . And that’s because they couldn’t spend it, or it was badly executed or programmes weren’t running to time.
That attitude was said to have had a knock-on effect of massive, negative impact on finance, that defence then has to manage and it can’t manage it through non-multi-year settlements . . . Therefore, it’s having to do terrible things to programmes on an annual basis because of annualised accounting and bad decision making. (Interviewee 18, 2024)
Last minute interventions by the Treasury after the finances were agreed, as with the SDR in 1998, were said to have created budget black holes.
Yet, others noted that the Treasury played an important role in compelling defence policymakers to reflect on and justify their decisions (which is perhaps telling in terms of the perceived rationality of much defence behaviour). As a Treasury official argued, ‘I think that having proper challenge into the system, and the Treasury is empowered institutionally to do that, is helpful’. Rather than distorting the finances, they asserted that ‘where the Treasury is of most value is making sure you’re linking policy and money together’. For them, defence reviews had to be done in tandem with a spending review, or afterwards, but never before: The one thing I definitely would never do, is the Defence review first. Because the tendency in the system is to therefore get bottom-up proposals. Before you know it, you’ve got 3.5, 4% of GDP and the Treasury are saying, ’Well, that’s all very interesting, but we haven’t got the money. (Interviewee 13, 2024)
Indeed, the impression given is that budget and ambition often become misaligned when there is a lack of Treasury oversight and input, rather than excessive meddling on the Treasury’s part. There is an abiding sense that the defence policy community regularly drifts into overspending in an unconscious fashion, requiring painful corrective action.
This tendency to drift has been attributed to the perceived difficulty of changing course once choices were made (Thomson and Blagden, 2018: 586). Even senior officers did not feel empowered to reverse major equipment programmes, despite the opacity of the original procurement decision and its subsequent development (Interviewee 8, 2024). The impression given is that policy outputs are not emerging from the rational calculation of a particular group. Rather, they are a product of bureaucratic inertia combined with fragmented political and organisational structures in which actors are not directed or empowered to reorient policy. As a result, programmes persist despite many participants believing they are either no longer useful, or the resources required would be better spent in other ways.
This lack of coherence between budget and ambition is important because financial control is usually the clearest coordinating thread in policymaking. The need for expenditure to be authorised by responsible accounting officers would normally be the basis for policy to be supervised and rationalised. In essence, it is an integral practice of statehood. Yet, the picture presented is of budgets routinely running over and programmes extending beyond the resources allocated, leading to ‘black holes’. Neither the government, individual departments nor clusters are able to reconcile a disconnect between resources and ambition. As a result, defence ‘policy’ is operating at a remove from both bureaucratic hierarchies and ad hoc sites of decision-making. This is before we take account of patterns of misrepresentation in forecasting and budgeting, and the ways that defence may follow patterns across large infrastructure projects in other policy spaces (Flyvbjerg, 2021; Hudon and Floricel, 2023: 157).
Implementation not enforced
The largest gap between stated aims and the allocation of resources to achieve them is apparent when it comes to implementation. Given the cost, both financial and political, of conducting a review, one would expect implementation to be a priority. Yet, implementation is frequently conveyed as an afterthought (Joint Committee on National Security Strategy (JCNSS), 2021a). The 1998 SDR was seen as unusual up to that point as the first review actually to have a team designated to try and ensure decisions taken were implemented (Cornish and Dorman, 2010: 407). Even so, implementation structures since then were frequently conveyed as more notional than real and quickly dissipating.
There are four aspects of the data which reinforce this impression, namely: the time allocated, a lack of ownership, a lack of accountability, and the belief that decisions are overtaken by events so implementation is not a priority. In the first place, implementation is not afforded the same time and attention as formulating policy. A senior IR21 review team member notes ‘I seem to remember [anonymised] in the Cabinet Office, had a bit of a fight to get the very short implementation chapter included at all’ (Interviewee 4, 2024). Another interviewee responded with laughter when asked about implementation: ‘I’m chuckling because I don’t think I’ve ever come across an effective implementation plan’ (Interviewee 14, 2024).
Second, the review team is not usually tasked with carrying out its recommendations, leading to a lack of ownership and continuity. As one respondent drafting the Integrated Review (2021) noted, ‘I remember saying . . . I am losing sleep now, because I don’t know how we are going to . . . collectively implement this’ (Interviewee 4, 2024). When they raised their concerns, they were told ‘don’t worry about it. Let’s just get this over the line’ (Interviewee 4, 2024). In the event, after briefing the contents of the review, they left government and were not involved in putting it into effect. Even where implementation was afforded Ministerial attention, this was fleeting. As another interviewee noted ruefully of the 2015 review, the tracking of its 89 commitments was ‘supposedly overseen by cross-Departmental which in reality met only a handful of times in the first couple of years’ (Interviewee 12, 2024).
Third, interviewees identify a lack of benchmarking and checking undermining accountability for putting the review into practice. Of the 2015 review, a Treasury official suggested that implementation criteria were meant to be set out in a settlement letter from the Treasury to each department, linking the budget to ‘deliverables’ (Interviewee 13, 2024). Yet, according to them, ‘that sort of petered out, frankly. So as an example, Defence didn’t deliver a 30% reduction in its civilian workforce. No one really did anything about that. It was just sort of observed as an issue’ (Interviewee 13, 2024). Where benchmarks existed, they are often represented in a cursory fashion:
I don’t know if you’ve looked at that implementation process, but it was just a list of 100 things which . . . we horribly gamed. The baseline was so low for a lot of those things that you could say you’d met them, even though actually, the substantive tasks probably hadn’t been completed. So as an implementation process, that was really dreadful (Interviewee 14, 2024).
Another respondent recalled of the Integrated Review (2021): I wrote and established a score card until I got told to stop doing that . . . week-by-week we said ‘have we done all these things?’ And people said ‘no, we haven’t done those things’. ‘Why haven’t we done these things?’ Usually for absolutely terrible reasons. Not because we couldn’t but because we chose not to. (FG3 Interviewee 17, 2024)
This makes the lack of implementation seem intentional (‘chose not to’) but it is often conveyed as because events or requirements have moved on. 6 Indeed, some interviewees rejected the idea that defence and security reviews should be understood as formal policy plans, instead seeing them as about establishing a tone, sharing ‘vibes’ or setting out a general approach rather than a specific plan of action (FG3, 2024; Interviewee 4, 2024). As one former Minister put it, ‘None of these documents tell you how to do anything anyhow . . . it’s just sort of getting on with it . . . there’s a “vibe”’ (Interviewee 11, 2024). Others perceived overly centralised control and accountability as contrary to good policy: ‘You want to be setting the system free and energising it to get on with it . . . That, to my mind, is a better spirit than holding people’s feet to the fire in a very micromanaging, tight way’ (Interviewee 14, 2024). In evidence to Parliament, a National Security Adviser suggested it was more difficult to implement strategy in foreign affairs due to the disruptive input of hostile states (Joint Committee on National Security Strategy (JCNSS), 2021b).
Where implementation was articulated, there was regular criticism of a lack of prioritisation. One focus group participant noted that Lord Sedwill, former National Security Adviser, acknowledged in evidence to a parliamentary committee that he: should have started with a handful of NSIGs (National Security Implementation Groups), I don’t know how many NSIGs were created under the Fusion Doctrine in 2018, there were quite a lot. He felt that actually he should have focused on a few and put his efforts into making those work before rolling it out more broadly. (FG3 Interviewee 4, 2024)
In practice, an MOD official in focus group 3 suggested, even though implementation groups proliferated, ‘there were hundreds of them by the end of it’, these were not brought together in a coherent fashion:
It’s actually proving how your ways and means achieve your ends. And we never did that work so the NSIGs were sort of left to float a bit . . . and it wasn’t clear how the whole came together to achieve your national security (FG3, 2024).
This picture of disinterest or disorder in the implementation process is important (Pressman and Wildavsky, 1984). Defence and security reviews are supposed to engage in a conscious, rational assessment of threats to the state and mobilise resources to meet them. Yet, interviewees suggest that implementation was either ignored, illusory or abandoned soon after the review was concluded. This undermines the sense of these reviews as a meaningful process. As the next section indicates, this is reinforced by the widespread evidence of a lack of good faith among participants, at times veering into deception and even lying.
Deception, alchemy or lies?
Put baldly, deception is endemic in UK defence. 7 Multiple interviewees and commentators identified this as a factor in defence and security policymaking. For obvious reasons, this is difficult to explicate. Respondents tend to presage acknowledgement of deception with the caveat ‘this is not for minuting’ or ‘this is off the record’. To label something as ‘deceptive’ implies a value judgement, which may mischaracterise the ethics of the activity. After all, people can deceive others for noble reasons (Mearsheimer, 2011). Nevertheless, it is arguably a significant factor in the dysfunction of defence policy and so needs further analysis.
There are four types of deception which are identifiable in the data and the broader discourse on UK defence, these are deception of others (usually hostile actors), self-deception, alchemy, and dishonesty. Deception is an inherent aspect of security and defence since concealing one’s true capabilities and intent could be vital to maintaining an advantage over rivals or enemies (Curtis, 2024). Thus, the United Kingdom does not reveal the precise number of nuclear weapons it possesses and when nuclear weapons delivery tests fail, as the last two Trident tests in 2016 and 2024 have, the UK government employs the euphemistic description of failure as an ‘anomaly’ before arguing ‘The test reaffirmed the effectiveness of the UK’s nuclear deterrent’ (Beale and Rhoden-Paule, 2024). Clearly a failing test cannot reaffirm something’s effectiveness and so this is a rather obvious form of deception. This kind of lying creates an inherent tension in public facing documents such as strategic defence reviews. On the one hand, their purpose is to inform the public and the rest of government about UK defence arrangements. On the other hand, they must necessarily obfuscate some aspects of defence for security reasons.
In other instances, we see an element of self-deception at play. For example, respondents often set ambitious targets that in hindsight were unrealistic (Jessett et al., 2020: 28). Successive reviews planned for a ‘Future Force’ in a decade’s time which did not materialise, at least as envisaged (Defence Committee, 2014). Delivery dates and cost estimates were often optimistic and financial shortfalls were offset by supposed efficiency savings (which again were not achieved) (Chalmers, 2018: 8). On the higher policy level, the widespread belief that UK armed forces could do ‘more with less’ fed into a deepening disconnect between the expressed belief that Britain remained a military power of the first rank and the underlying reality of diminishing capabilities and relative decline. In most cases, officials and Ministers did hope to achieve these outcomes but their impracticality suggests they were deceiving themselves.
A third category of deception entailed intentionally manipulating budgets to convey a misleading impression they could balance. This activity was political motivated and internally directed. Rather than deceiving external adversaries or themselves, it was intended to assuage domestic political concerns over the cost of defence by making it appear more affordable than it would be in reality. It was often illustrated in the data with the example of efficiency and cost savings, that were included knowing that they would not be achievable. As one interviewee put it, ‘We were indulging in the alchemy – that efficiency was somehow going to make up all these goods, all these things we were canning and couldn’t afford. Well, you know, it didn’t. It was an illusion’ (Interviewee 9, 2024). Labelling this ‘alchemy’ and ‘illusion’ conveys a sense of falsity and misdirection but there is a lingering sense that such savings were at least theoretically possible. In terms of motivation, alchemy could be linked to a British tradition of ‘muddling through’ or ‘incrementalism’ (Cornish and Dorman, 2011: 344). Rather than outright mendacity, the logic could be an optimistic belief that ‘something will turn up’, allowing tough choices to be avoided. As a respondent noted:
there is a tendency in politics, when faced with a difficult choice, of making a review or pushing it into the long grass, or hoping you’ll be moved to another department before someone has to decide. Because there are winners and losers in any decision (Interviewee 10, 2024).
Thus far, our examples of deception have emerged from security requirements, self-deception or creative accounting for political reasons. In some cases, there are more blatant forms of deception which were self-described as dishonesty or lying. For instance, in focus group 2, a former senior Royal Navy officer noted that assessments of implementation were not honest, suggesting ‘no one tells the truth with those things, it’s always “green”’ (meaning completed) (FG2, 2024). A former senior MOD official suggested, In the centre we were given rather a rosy picture by all the front line commands, who basically said ‘we’ve got loads of money, it’s fine’, and . . . then it turned out that they didn’t have as much money as they were saying. That they’d promised it three or four different ways etc. etc. and that’s how it came apart (FG2, 2024).
This might seem like a case of optimism bias but the naval officer affirmed ‘Hence my point why no one’s incentivised to tell the truth in this system’ and the MOD official replied, ‘Absolutely’ (FG2, 2024).
The reference to the system here is telling. In any large network one might expect some individuals to engage in bad practices like lying but many respondents saw this as a feature of the system as a whole.
As one Defence Minister articulated it, To me, all of these strategic defence reviews are everyone trying to pretend they’re thinking of grand strategy; but actually the reality is everyone is lying. But everyone has decided we can lie together. So nothing is actually properly funded. You’ll come up with a couple of little announcements in order to go with the press release but Treasury realise they’re not properly funding it. MoD realise they’re not properly funding it. I think it was 2015 where they announced the Type 31e frigate. Treasury knew that at £250 million per ship you could maybe deliver a ship, but you can’t deliver anything with a defence capability on it. So you have a big announcement of ‘we’re going to do this at £250 million’ but everyone is part of a lie.
In this reading, the entire process of the strategic review is implicated in a system of dishonesty, with individual decisions leading to budgetary shortfalls seen as a natural consequence. Some public policy analysts might describe this as ‘strategic misrepresentation’: the process of deliberately and systematically distorting information for strategic purposes (Flyvbjerg et al., 2002). Otherwise referred to as the ‘Machiavelli factor’, this is frequently associated with ‘cooking’ forecasts to make projects appear more viable on paper, ensuring they are initiated and sufficiently underway when the true scale of the costs become apparent such that there is no obvious benefit to abandoning the project. Thus, Flyvbjerg (2021: 534) cited an engineer arguing, ‘if we gave the true expected outcome costs nothing would be built’. In isolation, this is an attractive explanation; however, it does not fit the overall picture of the data. Rather than a single factor, failure emerges from combinations of self-deception, optimism bias, obfuscation for political reasons, as well as a systemic pattern of lying.
Dishonesty carries significant implications for any assessment of policymaking (Brunsson, 2002). If participants in reviews are lying, that corrupts essential assumptions about trust and rationality that are supposed to underpin bureaucratic systems (Weber, 1978). It also makes accountability impossible since commitments are not made in good faith and all parts are complicit in this deception. If the whole enterprise is false, then the notion of the state as a rational actor is brought into question. In addition, the bureaucratic rationality of organisations is also suspect. In the example of the frigate, it is not in the Treasury’s interest to create a funding black hole; nor is it in the Navy’s to create a capability that is not viable as a defence asset. While it might serve an individual Minister’s interest to have an announcement to make, that would not explain why ‘everyone’ in the system would be prepared to go along with it.
Modelling UK defence and security policy
Thus far, this article has highlighted some important themes from the data which contribute to the perception of policy failure. Identifying these pathologies represents a significant contribution to understanding in its own right. Three key insights stand out. First, the fragmented nature of the policy process, with a lack of vertical and horizontal coordination across Whitehall, results in decision-making clusters making choices without seemingly being aware either of their own self-interest or the guiding rationale for policy. Second, the lack of attention to implementation, even where structures are in place to execute strategy, means that reviews quickly dissipate. Third, there is widespread acknowledgement of deception and dishonesty among participants. Taken together, they paint a picture of a highly dysfunctional, incoherent and irrational policy process. While this adds to our empirical knowledge of the process of making and implementing UK defence policy, it also has ramifications for theoretical understanding.
As noted, commentary on UK defence policy is largely atheoretical. Consequently, scholars tend to fall back on US models (Grattan et al., 2015). The problem is these assume a level of rationality and coherence which are missing or marginal in the data for this project. For instance, Graham Allison’s (1971) influential tripartite analysis of US policy (Allison and Zelikow, 1999; Jones, 2017) posited three models of decision-making: a rational actor model, in which the state was a unitary actor and decision-making was cohered around the President, who made rational decisions on the basis of the national interest; an organisational process model, whereby an organisation’s standard operating procedures provided coherence and procedural rationality in policy implementation 8 ; and a bureaucratic politics (later labelled governmental politics) model, whereby policy was being shaped by domestic political pressures which affected the President’s scope and choice of response and brought a wider group of actors into the frame as influential.
Although the second and third of these models were associated with ‘irrational’ outcomes – whereby the general interest of the state as a whole was subordinated to political or bureaucratic interests – in reality, rationality was still in play. Organisations and political groups were acting in their rational self-interest, even if this contradicted national goals. In each case, an internal logic would be in operation, whether that be the national interest, sectoral interests or bureaucratic interests (Freedman, 1976). Rationality would be understood in terms of actors ‘maximising utility’, that is, using resources efficiently and appropriately to achieve their goals (Hill, 2003; McCormick, 1997). In addition, underpinning each model is normally a cognitive process of rationalisation – actors have to interpret the world around them and then explain their actions to themselves and others (Hollis and Sugden, 1993).
While there is evidence of the rationality of each of these policymaking models in our dataset, it is striking how often that logic is undercut by actual practice. For example, if we’re looking for a rational actor, the former National Security Adviser’s account above, in which they claim to be seeking to ‘organise ourselves’ and link levers with objectives, and the review drafter’s goal of creating a ‘coherent story’ would seem to fit the bill. Yet, as we saw, the actual level of central coordination in many of these reviews was patchy, with respondents surprised by how little direction they received from senior policymakers. Narratives of a rational process of central direction and heuristics leading to prioritisation and allocation of resources are followed by admissions that this did not play out in reality. Policymaking was far more decentralised and fragmented than traditional models anticipate (on the error of assuming rationality, see Heuser, 2025: 10–20).
Although the account of departments withholding information and bargaining for budgets seems to fit the bureaucratic politics model, the depiction of choices being made by clusters within and across departments undermines the sense of monolithic agencies pursuing organisational self-interest. The narrative of how decisions were arrived at also undermines the sense of standard operating procedures creating path dependency. Rather, reviews appear to be constructed on the hoof, with drafters often uncertain of their authority (self-describing as ‘naïve’), uncertain of the norms that should drive their choices, or their position in the ecosystem of decision-making, and lacking the ability to ensure implementation. Organisations did not behave monolithically and participants lacked the broader awareness of relative power and status required to advance sub-group goals.
The nearest account of policymaking allowing for the possibility that choices and actions do not follow a coherent internal logic is that of governance theorists such as Mark Bevir and R.A.W Rhodes, who chart the decentring of the British state and increasing salience of networks in British policymaking (Bevir and Rhodes, 2007). Bevir and Rhodes observed that actors seek to rationalise their beliefs and modes of behaviour with reference to traditions, but those traditions have no existence beyond that process. As a result, there is no ‘structure’ or logic as such; rather there is only the immediate responses of policy practitioners and the narratives they construct about what they are doing and why they are doing it (Bevir, 2005; Bevir and Rhodes, 2012).
As such, security and defence reviews appear to align with ‘decentred governance’ theory (Bevir and Rhodes, 2007). Given the above pathologies, efforts to construct models that assume rationality, logic and coherence as features of UK defence and security policymaking do not fit the data. Rather, the evidence suggests a more diffuse and ‘free associative’ mode of making and practising policy is in operation. Instead of constructing policy within a hierarchical political structure, with an anchoring philosophy or strategic text, participants in reviews respond to immediate stimuli via changing configurations of decision-making. Their awareness of their own interests and those of their department, and how these relate to other actors in the organisational ecosystem, is severely limited, meaning choices are often made reactively, in compressed time periods, with rationales obscure, even to those making decisions.
Underpinning traditional models of policymaking are metaphors, such as the unitary state, rational actor, the bureaucratic machine, the bargaining agent. As Lakoff and Johnson (1981: 97) noted, ‘a metaphor works when it satisfies a purpose, namely, understanding an aspect of the concept’. Since these metaphors appear inadequate in describing how strategic reviews operate in the United Kingdom, we suggest new metaphors are required. In particular, we advance the jellyfish as a better metaphor for how the UK defence and security policy system functions. Jellyfish have no central nervous system. Instead, they contain ‘nerve nets’ which capture limited information and translate that into action. Thus, if a jellyfish encounters a rock, each nerve net responds to being immediately confronted with this obstacle and responds by avoiding it. As each net does likewise, the organism as a whole appears to move as one but it is only doing so due to the stimulus responses of its individual components.
To the authors, this captures how participants described their experience of strategic defence reviews. Weak central direction and lack of coordination between the centre and departments mean the UK defence policy sphere lacks a ‘brain’ rationalising and cohering policy. Disconnection between departments and the clustering of decisions across bureaucratic structures (in constantly changing forms) presents a picture of a more diffuse and fragmented process than any static model can convey. Any sense of rationality and collective cognition is further undermined by the disconnect between policy ambition and resources, compounded by the widespread use of deception or even outright lies. Conjuring the metaphor of a jellyfish thereby conveys this picture of an organic, neural arrangement of reactive components. We assert this is a more descriptively accurate rendering of the data than formal models which are linear, rational and impose a coherence that is absent from the thing being studied (Heuser, 2025)
There are two possible responses in light of these insights. Policymakers could seek to impose a hierarchical, linear model of control over policy formation. The caveat is that governments have been attempting this for decades with decreasing success. Even authoritarian governments are unable to establish lasting and meaningful control over policy outputs and outcomes (Monaghan, 2012).
The alternative option would be to lean into the jellyfish model and stimulate various parts of the policy system to respond to challenges by direct, adaptive and agile forms of governance. This solution is often associated with tech start-ups or the creation of the US’ Defense Advanced Research Projects Agency (DARPA) and is promoted by Dominic Cummings in the UK context (Cummings, 2017). Governance operates via smaller clusters of empowered decision-makers, rather than monolithic bureaucracies. It echoes the concept of ‘Integration by Design’, which has been pushed by reformers in recent reviews, entailing ‘cross-functional teams’ that are ‘brought together and interests aligned around the achievement of specific outcomes’. Yet, while this has been on the agenda of defence reform for some time, a supporter suggests ‘I don’t believe we are any closer to it’ (Interviewee 17, 2025). In part, this may be due to bureaucratic resistance; however, there are other issues, such as ensuring democratic accountability (especially when it comes to external advisors). In addition, it perhaps favours crisis response (to an immediate and tangible stimulus) over addressing chronic and long-term challenges.
Conclusion
Existing accounts of British security and defence reviews are largely atheoretical. When they do seek to model policymaking, they lean on American models that overstate the intentionality and rationality in the policy process. Based on our extensive original empirical data, we have identified seven key pathologies in the review process: weak central direction, a disconnect between the centre and individual departments, a disconnect between departments across Whitehall, the clustering of decision-making, a lack of alignment between budget and ambition, a lack of attention to implementation, and widespread deception. Taken together, these undercut models that assume linear policy processes carried out by rational actors operating within a hierarchical structure.
In their place, we posit the jellyfish model as an original and descriptively accurate metaphor for the fluid, organic, neural process through which these reviews are made, as implied by the data. Earlier scholarship on defence policymaking in International Relations frequently made use of analogous models to explain the policy process (Heuser, 2025; Hilsman, 1987). This approach has fallen out of fashion, but the result is that scholarship lacks a theoretical anchoring; or grounds analysis in conceptual models that assume a level of rationality and coherence that the data suggests is missing in British defence policymaking.
Our work adds to recent attempts to theorise British defence by drawing attention to organisational processes, as understood by practitioners (Nemeth and Dew, 2024; Uttley et al., 2019). The purpose of a model is to allow the analyst to understand how social phenomena operate and to guide their analysis. Our data suggests that UK defence policy is fragmented, spasmodic, and reactive to specific, immediate stimuli. It presents a convincing account of why outputs seem to lack strategic coherence and government actors struggle to coordinate their activities, even when they are consciously looking to do so (as in the fusion doctrine). Our research has shown how this process is reinforced by the culture of the British defence and security community, especially the widespread use of deception, which inhibits logical and rational assessment of defence commitments. The jellyfish model is useful for future analysts as it steers them away from overstating the level of structure and organisation in the defence policy process, fosters scepticism over apparent bureaucratic mechanisms for oversight and implementation, and encourages them to identify and analyse mutative clusters of decision-making, if they want to explain outcomes. As such, it provides a way of conceptualising policymaking that could be applied to other cases of defence and security review failure, in the United Kingdom and abroad.
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: The study was partly funded by Research England.
Informed consent
All participants have provided informed consent for their inclusion in this work.
