Covert action failure and fiasco construction: William Hague’s 2011 Libyan venture

ABSTRACT In 2011 William Hague, then British Foreign Secretary, authorized Special Forces to enter Libya and contact rebels opposed to Muammar Gaddafi in the unfolding civil war. However, its members were detained by the rebels and ejected from the country. This article puts the literature on public policy failures into dialogue with that on covert action as a tool of foreign policy. It asks: why did this not develop into a fully fledged policy fiasco when journalists and politicians judged it to have been a major error of judgement on Hague’s part? Using narrative analysis of the contemporary reporting of this incident, we argue that the government – possessing the advantage of information asymmetry accruing from operational secrecy – was able to win the battle of narratives in a frame contestation process. The article reflects on how the study of information asymmetry can enhance the recently revivified research into foreign policy failures.


Introduction
Covert action is one of the most controversial means of executing government policy, a high-risk option fraught with inherent dangers. Given this, it might be assumed that whenever covert actions turn sour, which many do, a full-blown fiasco will be the result. However, even if a covert action is judged to have failed, secrecy may well prevent this apparent failure developing into fiasco. This creates an interesting paradox, wherein some of the most controversial foreign policy ventures that go wrong do not develop into fiascos, when less controversial domestic public policies that are alleged to have failed may well do. The goal of this article is to investigate the narrative construction of covert action fiascos, a fiasco 'type' that has been sorely neglected in the emerging research programme on public policy fiascos and foreign policy mistakes.
Our case study is Britain's attempt to insert a secret team into Libya in March 2011, at the height of the Arab Spring. We investigate the ways in which participants in accountability forums in politics and the media attempted to construct a fiasco, but ultimately failed to do so. Our argument is that Foreign Secretary William Haguethe character in the drama deemed most blameworthy by most commentatorswas narrowly able to avoid the perceived failure becoming a fiasco by successfully engaging in a frame contestation process. The government ran an effective message management campaign that downplayed the salience of the operation while stressing the imperative of a bold and united British response to the Arab Spring. We suggest that the most significant advantage for Hague was, though, that he could rely on a near monopoly over the information that would have been required for the successful elaboration of a fiasco narrative. Information asymmetry prevented critics from mobilizing behind an unassailable fiasco narrative built on firm empirical foundations that might have gained traction in the public mind. Helpfully for Hague, the politics were also on his side. The government's wider Libyan strategy did not at this point prove politically contentious, so there was less inclination for the government's critics to dig around for more dirt on Hague and Libya. Ultimately, therefore, the Foreign Secretary survived in office, was backed by the Prime Minister and Cabinet colleagues not known to be naturally sympathetic to him and, arguably most importantly, he managed to win the battle of 'frames' by having his version of events come to be accepted as the dominant one. By tracing in a systematic fashion the links between information and fiasco narratives, the article helps us appreciate the challenges of studying foreign and domestic policy fiascos more widely.
The article proceeds in three parts. First, we review the two literatures we bring together: the public policy evaluation literature and that on covert action, with a view to explaining the benefits to be garnered from linking the two. Second, we define key terms such as 'fiasco' and show empirically the opportunities and challenges of studying covert action using this conceptual toolkit. The third part presents our method and discusses the main evidence from the Libya case study. The conclusion reflects on the most prominent themes to emerge from our analysis: information as the pre-requisite to successful fiasco construction; secrecy as an under-theorized barrier to fiasco construction and to effective lesson-learning; and the relevance of message management to the successful containment of a potential fiasco.

State of the art: public policy analysis and foreign policy failures
To our knowledge, this is the first article of its kind to link the scholarship on public policy and covert action. Since the publication of Mark Bovens and Paul 't Hart's seminal work (1996), public policy evaluation literature has been dominated by discussion of domestic policy fiascos, such as the UK Poll Tax or, later, the Millennium Dome project. This is a very diverse array of cases, often strongly differentiated in terms of nature, magnitude and impact (see Howlett 2012;McConnell 2010McConnell , 2016. Meanwhile, the historiography of covert action has been dominated by linear narratives, drawn from archival research, and country-specific case studies. Accompanying this, there are many works considering specific covert actions deemed to have failed, notably the Bay of Pigs (Rasenberger 2012) and the Iran-Contra affair (Byrne 2014). These are all insightful studies. However, they lack a comprehensive theoretical framework resting on a sophisticated analysis of causation and evaluation of the material from a political science perspective. Particularly marked in this literature is a tendency to objectivize the failure as a given, without attention to the constructed and politicized nature of the events in question.
It is, therefore, a welcome development that public policy and foreign policy scholars have recently begun working collaboratively to reignite the study of foreign policy failures (see the 2016 Special Issue of the Journal of European Public Policy; Kruck et al. 2017). This work has sought to move forward both sub-disciplines via critical reflection on the nature of policy failure (Bovens and 't Hart 2016) and, on the empirical side, by 'bringing public and foreign policy together' (Oppermann and Spencer 2016a). However, much of this literature still ignores, misunderstands or neglects policies involving secret intelligence. Bovens and 't Hart offer a cursory mention of intelligence when discussing typologies of failure, arguing that it falls in the 'tragedy' category in so far as 'accomplishments can often not be publicized, [and] may suffer from a similar lack of community and political appreciation ' (2016: 658). This is fair, but overly simplistic in that it assumes programmatic success and does not account for failure. Similarly, Marsh and McConnell (2010: 575) briefly discuss the contested success or failure of extraordinary rendition, but consider neither secrecy nor how the operations came to light in the first place. Nonetheless, this new agenda provides a very welcome antidote to the benign neglect that previously characterized the relationship between scholars in the two fields. It was also a cue to our article, which adds covert action, with its inherent secrecy and controversy, into the mix.
Bringing covert action, as part of broader foreign policy activity, into dialogue with the public policy literature is valuable in three ways. First, scholars of International Relations and Foreign Policy Analysis have long been interested in why things can go wrong in foreign policy and crucially how policy-makers learn lessons from these episodes (see Allison and Zelikow 1999;Beasley 2016;Daddow 2011). However, and this is especially the case for those studying covert action, scholars have not yet traveled very far down the road of conceptualizing and categorizing analytically all the different types of mistake/ blunder/failure/fiasco. Such categorization is necessary to provide fine-grained understandings of how failure is identified, labeled and constructed through various forms of political and discursive contestation. By extension, this will generate more nuanced lessons, which are especially required in an era when lesson-learning is the main way in which publics engage with foreign policy debates (see for instance The Report of the Iraq Inquiry 2016). Moreover, covert action, and the inherent secrecy therein, problematizes the lesson-learning process of extracting visible information which can offer credible and generalizable lessons. The case study below is particularly relevant because it highlights how learning can occur from successful message management and fiasco containment, contributing novel information in a field dominated by learning from apparent failure.
Second, and linked to this, analysis of the domestic political settings within which foreign policy fiascos are constructed, and blame apportioned, casts much needed light on the interconnectedness between national and international politics. 1 For foreign policy fiascos, no less than domestic fiascos, are only foreign policy fiascos when deemed to be so by domestic political audiences and commentators. Studying covert action gives vital insights into the ways in which operations, when revealed, raise debates about democratic legitimacy, accountability, transparency, and government oversight of foreign policy activity.
The third value of our research agenda is that it draws attention to the methodological challenges of conducting fiasco research. A lot but not all the current research in what is an encouragingly pluralist field is, broadly, interpretivist in nature (see Bevir and Daddow 2015;Lynch 2014;Schwartz-Shea and Yanow 2012). There is an argument to be made, however, that the distinctions between positivist and post-positivist research have been overdrawn, and this is evident in the fiascos literature. For example, positivist accounts do not deny an element of interpretation when evaluating public policies. Likewise, post-positivists stress the constructed nature of fiascos. However, they do not claim that anything can be said about a supposed public policy failure (echoing Oppermann and Spencer 2016a: 649-50) or that everything is discourse (Howarth et al. 2000: 3). Valid and reliable empirical evidence is the essential bedrock to both modes of inquiry.
Fiascos research is very well placed to tap into this meta-debate in the philosophy of social science, as well as those that occur within these traditions (Bevir 2002;Bevir and Rhodes 2016). Alan McConnell explains this point very well (2016: 677), when he remarks that despite the formidable methodological challenges facing the fiasco analyst, this is not a challenge that should be shirked because they 'mirror classic methodological differences in the social scienceswhether the phenomenon being studied is a matter of "fact", interpretation or both'. Covert action, as discussed in the next section, starkly demonstrates the methodological conundrums facing all fiasco researchers.
Covert action: the challenge for fiasco construction Covert action is a state's attempts to interfere in the affairs of another state in a plausibly deniable manner. It generally falls into one of four categories: political action, economic action, propaganda, and paramilitary action. By definition, covert action entails a controversial set of practices in international relations. When things go awry, the potential for fiasco narratives to emerge would seem to be very high.
The fiascoand broader policy evaluationliterature offers a useful framework for assessing the scope and limits of covert action as foreign policy choice. Mistakes and failures, although contested, are part of the daily life of any bureaucracy (McConnell 2016: 667). They are multifaceted, ranging in severity, cause and degree of politicization (see Gray 1998: 8;Moran 2001: 416-7;King and Crewe 2013: 4). Like success, failure is a subjective label 'applied by stakeholders and observers' (Bovens and 't Hart 2016: 654;see also McConnell 2016: 667-84). Fiasco is a type of failure, and not all failures become fiascos. A policy fiasco is 'a negative event that is perceived by a socially and politically significant group of people in the community to be at least partially caused by avoidable and blameworthy failures of public policymakers' (Bovens and 't Hart 2016: 653-4). Generic failure, by contrast, does not necessarily contain an assumption that individual actions are at fault (McConnell 2016: 668). We therefore posit a crucial distinction between fiasco and failure: although both are subject to debate and contestation, fiascos are more politicized. Fiascos arise out of political or reputational failure as well as programmatic or decisional/implementation failure (Bovens and 't Hart 2016: 653-4, 658;Marsh and McConnell 2010: 571). Through discursive construction, they are visible, significant, public, and involve the attribution of blame.
Fiascos therefore 'share a common process' or are 'marked by the existence of particular patterns of beliefs about events' (Gray 1998: 8). Four of these stand out. First, the perceived failing must be articulated by policy evaluation stakeholders to have been visible and immediate. Second, observers need to be able to identify causation. This process, as Howlett (2012: 543) has pointed out, is heightened when failures are deemed to have been predictable and avoidable. Third, observers have to be able to explain and interpret agents' behavior in bringing about the negative outcome: characterization and identification of the main players in the drama is important. According to Bovens et al. (2006: 323), 'astute players of the evaluation game will therefore attempt to produce facts and images that suit their aims.' Such frame contestation leads to a fourth component of a fiasco narrative: blame allocation. The media is a key arena in which this takes place, not least because it is by and through the media that fiasco frame contestations play out. By providing a 'pivotal forum for political sense-making,' it can be integral to explaining why some failures enter the limelight and escalate into fiascos (Bovens et al. 1998: 48).
For the covert action analyst, these already challenging difficulties are strongly magnified. Ontologically, it is difficult to ascertain (usually shifting) policy goals against which to measure outcomes. If we do not know what the goal was, how can we say that it failed, or indeed succeeded (McConnell 2016: 669)? Owing to the inherent secrecy of covert operations, it is very tricky to identify and label a covert operation a failure in the first place, let alone politicize it, allocate blame and construct a resonant fiasco narrative. Why? Because secrecy restricts the information flows required to emplot a credible story about what occurred, limiting the capacity of outside commentators to seek out the causes of any alleged failure. Even when an operationwhich stakeholders accuse of having faileddoes come to light, secrecy and deliberate obfuscation hide some or all of the details which are essential to the construction of fiasco narratives. Moreover, it can be a very long time before covert actions of any kind, successful or unsuccessful, come to light. The passing of time between failure and public knowledge has long prevented the necessary politicization for covert action failure to become fiasco.
Even confronted with these obstacles, covert action fiascos have been said to have occurred. We suggest that there is a discernible pattern in the narrative construction of each. In line with public policy literature discussed above, the process includes, first, a perceived visible, immediate, and consequential failure, with sufficient momentum to sustain a fiasco narrative. Second, there is a transfer of information, often through an accountability forum such as the media, which challenges government monopoly on information, and out of which stakeholders are keen to make political capital. Third, there is an attempt to identify the source of the alleged failure, usually spanning one or more levels of analysis: individual decision-makers, a wing of the bureaucracy, or perhaps the strategic foreign policy outlook of the government in toto. Identification of the source(s) or origin(s) of the perceived negative outcome is a vital function of the fiasco narrative that includes: causation, (un)avoidability, uncertainty about goals and so on. Each ingredient is weighed differently by participants arguing about whether and how far the operation was a fiasco. The core factor in each is access to information: without details of the operation, neither failure nor fiasco can be constructed.
To put some empirical flesh on these patterns, we can look to the 1961 Bay of Pigs fiasco, when the CIA supported Cuban exiles in an attempted revolution, or the 1997 Khaled Mashal fiasco, when Israel's secret service, Mossad, botched the assassination of a Hamas leader (Kahana 2007: 65). Both were visible and immediate: the sponsoring state was quickly identified and the operation sparked much public discussion. Both were successfully articulated as being sufficiently consequential in terms of impact on policy and prestige to warrant the resources necessary to detailed investigative journalism (on America see Haefele 2001: 77-8; on Israel see Cowell 1997). Both cases also involved a transfer of information to the accountability forum of the press, which challenged the government's monopoly over information and allowed stakeholders to politicize failure and construct the fiasco. Following the perceived failure at the Bay of Pigs, senior administration officials, including the President and the director of the CIA, gave numerous unattributable briefings, which involved blame avoidance (Barrett 2016: 6-23). Similarly, Mossad briefed reporters with off-the-record details, whilst the Israeli Prime Minister attempted to make the intelligence community a scapegoat (Pateman, 2003: 154). This permitted characterization and the identification of agency, both of which are integral to fiasco emplotment, but which, usually, remain obscured regarding covert action.
The US's failed Iranian hostage rescue in 1980 and the 1986 Iran-Contra affair further the most important ingredients in a successful fiasco construction. 2 In both instances, the government lost its monopoly over information, opening the way for the emergence of credible fiasco narratives via politicization of the issue, public discussions about magnitude and consequence, and the allocation of blame (on Iran-Contra see Brody and Shapiro 1989: 353-69; on the rescue attempt see Smith 1985: 117-23). In all the above cases, the transfer of information was key. When the transfer took place, the government suddenly found itself at a disadvantage regarding covert action compared to overt action: issues of secrecy and operational security, asset protection and potential diplomatic fallout can constrain the government from constructing a compelling counter-narrative. The case study data presented in the next section shows vividly how and why these are the essential ingredients in a successful fiasco construction.
Information asymmetry and the battle of narratives over Libya 3 Fiasco authorsjournalists, public policy analysts and historianstell fiasco stories in narrative form, through which they give the events meaning for readers. Consensus around certain core 'facts of the matter' results in the repetition of keywords and phrases that offer themselves up as empirical patterns identifiable in the data (Klotz and Lynch 2007: 45-54). These patterns delimit a narrative 'field' that helps structurebut does not determinewhat can and cannot be said on a given issue (see Howarth et al. 2000: 15;Wetherell et al. 2009: 7-8). The more basic facts that can be established, and the greater the certainty about those facts, the more likely it is that a particular interpretation will establish credibility amidst the 'noise' of counter-narratives and rival emplotments.
The pertinent feature of fiasco narratives is that they are constructed very clearly as morality plays. This should not be a surprise because as Hayden White remarks (1975: 14): it 'seems possible to conclude that every historical narrative has as its latent or manifest purpose the desire to moralize the events of which it treats' (see also Daddow 2006). Fiasco narratives thus contain two interconnected features that make them fertile sources of evidence about the politics of fiasco construction (drawing on Oppermann and Spencer 2016b: 691). First, they emplot the reasons why a failure occurred. Second, they pinpoint who or what was to blame for the failure, and debate what could have been done differently to avoid the incidents or decisions that led to the (usually negative) outcome.
The method selected for the analysis of the coverage of Hague's Libyan venture was, therefore, a close study of media narratives, unpacked using narrative analysis of the relevant news reporting and opinion pieces. The research questions that guided our extraction of the data were all geared to investigating the setting, characterization and emplotment of the media reporting: (i) who were the main actors in the Libyan episode, where were they operating, and what were they doing? (ii) what was the magnitude of the failureif anyand who was to blame for any alleged failings? (iii) what were the resistances to the fiasco narratives and how successful were they? We studied the coverage from the day the story broke on 6 March 2011 and tracked how it unfolded for the next 10 days.
When the story broke it was evident that, lacking an official account of what had happened, journalists had a desperately difficult job establishing facts, especially who was involved. Phase One of the media coverage was thus driven by uncertainty regarding both the cast in the drama and the equipment they had on them at the time of capture. The Sunday Times, which broke the story, put the figure at 'soldiersup to eight men' captured escorting a 'junior diplomat' through rebel-held territory in the east of the country (Jaber 2011, see online appendix). Others followed suit but there was little consensus about the numbers involved or their identities, with speculation covering soldiers, SAS, diplomats, and MI6 officers. There was only marginally more agreement on the equipment the seized men were carrying, and the reporting again spoke to the James Bond-style nature of the mission. Sensationalized speculation stretched to fake passports, maps, communications equipment, and weapons.
After a couple of days (some overlap aside) the information-gathering phase gave way to Phase Two: blame attribution. Attempts at fiasco construction were built on shaky empirical foundations because characterization had been so weak in Phase One. There was, however, early and widespread agreement that something had gone wrong with Hague's Libyan operation, the failure of which was deemed visible and controversial. In line with the aforementioned ingredients identified for a successful fiasco construction, someone had to be to blame. Coding for the adjectives used to describe the mission revealed around 20 negative descriptors of varying degrees of strength, including failure, bungle, and, indeed, fiasco. Managing the consequences of this heavily negative verdict became the government's main task over the next few days. The success of its resistance narratives and message management would determine the future of Hague himself and the credibility of the Cameron government's Libyan and wider foreign policy.
The consequences of any perceived failure were most grave for Hague himself, with the Prime Minister and government as a whole also implicated. Only a tiny portion of the flak was taken by then head of MI6, John Sawers (Sherwood 2011, see online appendix). This was very much a story centering on the 'Westminster bubble'. Three facets of the media coverage are noteworthy in this respect: the information vacuum; the sense that the British lacked control over what information was coming out of Libya; and the debate about Hague's role and responsibility.
First, central government personnel, constrained by secrecy and operational security, initially came forward with very limited information. One national and some regional newspapers picked up on comments from Defence Secretary Liam Fox. He defended the operation on the grounds that it was intended to 'build a picture' of what was going on in Libya 'so we are able to get a clearer idea of what we are able to do in terms of helping the people of Libya' (Chulov 2011a; Coventry Evening Telegraph 2011; Daily Post 2011; The Western Mail 2011, see online appendix). The Independent reported that International Development Secretary Andrew Mitchell had taken a similar line in a radio interview. There was no official comment, again because of the classified nature of the operation, from the Foreign Office or Ministry of Defence (Tapsfield 2011, see online appendix). However, unofficial remarks indicate that government ministers were spinning the operation as intelligence-gathering (legitimate, tactical and routine) rather than a more sensitive event-shaping activity. The Prime Minister's Office was approached by journalists but also declined to reveal details, merely adding that Hague had approved the operation 'in the normal way' (Wintour 2011, see online appendix). The government's line was that this was 'business as usual'.
Second, the lack of official information on the British side left the way clear for journalists to fill their copy with information gleaned from the Libyan side. Crucially, this emerged very early, adding to the sense that something was amiss. Top billing was given to a leaked telephone conversation between Richard Northern, the UK Ambassador to Libya, and a Libyan rebel leader, which took place in the immediate aftermath of the incident. In the short exchange, Northern apologized for the 'misunderstanding' and the leader chastised the 'mistake' of sending the mission team in by helicopter at night ( Covert action can thus work both ways for a government when an operation is deemed to have gone wrong. On the one hand, information asymmetry can work in a government's favor to stymie potential fiasco construction. On the other hand, it can also constrain a government's ability to construct a compelling counter-narrative, allowing journalists to speculate and build new foundations on which to construct fiasco narratives. The third noteworthy pattern in the media coverage was the attribution of blame squarely on the shoulders of the Foreign Secretary, as the originator of the operation. Consistent with public policy fiascos, the press (from across the political spectrum) highlighted the avoidable nature of the failure by emphasizing poor planning, judgement, and scrutiny (see for example Coughlin 2011b; Hughes 2011b; Wintour 2011, see online appendix). As with previous covert actions which surfaced as perceived failures, Libya had created political space between the government's intentions and the outcome, which was ruthlessly exploited by Hague's opponents. Political leaders need to make meaning and to 'fill this space with their own interpretation' of the narrative (McConnell et al. 2008: 602). Hague was trapped between the need for secrecy and the press's traditional support for 'brave' and 'heroic' Special Forces at the expense of bumbling politicians. To Hague's credit, he made a statement to the House of Commons and answered questions about the Libya incident on the afternoon of 7 March. Although secrecy impeded Hague from giving full details, he afforded himself enough latitude to ensure that this transfer of information was carefully framed to maintain government dominance of the narrative. In taking this step, Hague entered into a 'frame contest', which arises in the aftermath of crises and which affects beliefs and the relative importance individuals attach to those beliefs (Boin et al. 2009: 82;Gershkoff and Kushner 2005: 526).
The Foreign Secretary stated: 'I authorised the dispatch of a small British diplomatic team to eastern Libya in uncertain circumstances, which we judged required protection, to build on these initial contacts and to assess the scope for closer diplomatic dialogue' (Hague 2011: cm643). Hague unsurprisingly offered no admission that the diplomatic team consisted of MI6 and SAS personnel. After all, political actors are risk-averse and will not offer extra detail unless it suits them, or they are forced to. Hague continued by explaining that the team 'was withdrawn yesterday after a serious misunderstanding about its role, leading to its temporary detention.' These deliberately bland statements normalized the incident by framing it as part of everyday, rational and legitimate Foreign Office activity. He even attempted to sell the operation as having been partially successful: 'this situation was resolved and [the team] was able to meet council president, Mr Abdul-Jalil.' (Hague 2011: cm643).
Contrary to Hague's anodyne rendering of the operation, opposition politicians sought to frame the events somewhat differently. The public policy literature argues that 'fiascos and prominent failures attract those interests that seek to capitalize on those failures' (McConnell et al. 2008: 601). Despite the national security implications, observers clearly sought to make political capital out of the Libya operation. Douglas Alexander, Labour's Shadow Foreign Secretary, accused Hague of mishandling the affair: 'does the Foreign Secretary accept that if some new neighbors moved into his street, the British public would be entitled to wonder whether he would introduce himself by ringing the doorbell, or instead choose to climb over the fence in the middle of the night?' (Alexander 2011: cm646). A good joke always plays well with journalists and political sketch-writers, and this was widely reported.
Journalists were encouraged in their attacks by accusations of 'bungling' and incompetence not only from opposition ministers, but some of Hague's fellow Conservatives and members of the Coalition government. In the debate he asked: 'Is it not clear that this mission was ill conceived, poorly planned and embarrassingly executed?' (Campbell 2011: cm649). Events were framed first as failure and second as avoidable, all overlaid with an element of farce. Hague, however, defended himself by taking 'full ministerial responsibility' (supporting the Downing Street line) and drew on Ambassador Northern's defense that this was a 'misunderstanding' but nothing more sinister.
The following days saw further attempts by Hague's opponents to bolster the traction of the fiasco narrative and undermine the Foreign Secretary. In this endeavor, parliamentary routine was on their side. Just two days after Hague's statement and initial interrogation, Prime Minister's Questions on 9 March provided a high-profile opportunity for the opposition leader Ed Miliband to land some telling punches on what he called the 'fiasco' of the government's Libyan policy (Kirkup 2011, see online appendix). There was, he said, 'a question of competence at the heart of this government' (Churcher 2011, see online appendix). However, Cameron stood by Hague, calling him an 'excellent' minister and taking 'full responsibility for everything that my government does' (Kirkup 2011; Leicester Mercury 2011a; Metro 2011, see online appendix). Hague, it seemed, was not to be isolated or picked off, despite the fevered speculation (for instance Richards 2011, see online appendix). To help the embattled Foreign Secretary, Downing Street also briefed journalists that Hague had Cameron's full support, along with the backing of fellow Cabinet ministers (Bentley 2011; Kirkup 2011, see online appendix). Hague emerged from the episode bloodied and bowed, but was not beaten into a resignation.
As the fallout continued in the media over the ensuing days the Daily Telegraph was a vocal but almost lone supporter of the Foreign Secretary. An article by Peter Oborne on 11 March claimed Hague was the victim of a 'smear campaign' by rival politicians involved in a 'battle for power'. Oborne further charged that SAS operational planning was to blame, and that it had been Hague's constitutional duty to take responsibility for errors caused by the bungling of others. Oborne also questioned the idea that this was part of a series of failures in the government's Libyan policy, especially regarding the evacuation of British nationals at the outbreak of hostilities (all in Oborne 2011, see online appendix). On 13 March Hague himself gave an interview to what was evidently his favorite newspaper, the Conservative-supporting Sunday Telegraph, in which he tried to end speculation about his future and Cameron's support for him (see Bentley 2011, see online appendix). This was extensively covered as Hague saying 'I've still got my mojo', a direct response to Liberal Democrat criticisms (Sun 2011, see online appendix), and that he would continue in office 'for an extensive period of time' (Heaven 2011, see online appendix).
Our analysis of media narratives reveals that, consistent with fiasco construction processes observed in the extant literature, socially and politically significant groups, having identified failure, attempted to highlight avoidable mistakes. Some in The Telegraph, for example, argued that 'it is not the SAS that should hang its head in shame over this fiasco: it is the idiots who authorized it in the first place' (Coughlin 2011a, see online appendix). Similarly, The Daily Mirror quoted an ex-SAS member who accused the mission of being 'badly planned and under-resourced' (Hughes 2011a, see online appendix) and, by implication, avoidable. This is not necessarily fair, as it would have been the Director of Special Forces who advised Hague on feasibility. Yet the narrative ties in with a 'lions led by donkeys' frame, emphasizing the romance and mythologies surrounding Special Forces and MI6, who are generally held in higher regard than supposedly amateurish, blundering politicians (for a typically hagiographic example shortly after the failure see Harding 2011).
Libya, however, did not become a full-blown fiascoin the technical sense seemingly for three reasons. First, journalists interpreted the operation as a visible and immediate failure, but did not interpret it as sufficiently consequential to dig for inflammatory details over a prolonged period. The narrative that came to dominate was that the Libyan venture lacked the visible consequences associated with decisions as the invasion of Iraq in 2003 (see Ricks 2006;Bovens and 't Hart 2016: 655 and 660-1). Had a covert operation turned out similarly in Iraq as opposed to Libya, one suspects the fallout would have been much more severe, because a prior narrative of policy failure would already have been in place. Unlike Iraq, the Libyan venture was not deemed serious enoughand the overall policy programme not contentious enoughto warrant an inquiry, which might have reignited discussion and led to post-hoc fiasco construction. Hague's narrative in fact came to be accepted as the official line, an important indicator of success in this kind of narrative contestation (Klotz and Lynch 2007: 55).
For example, the following year's Intelligence and Security Committee annual report limited formal criticism to a paragraph which blamed the botched operation on time pressures and a lack of planning (Intelligence and Security Committee 2012: 19). Likewise, the Cabinet Office's lessonlearning retrospective reduced the incident to a 'serious misunderstanding' which was quickly resolved. In doing so it almost repeated Hague's words verbatim (National Security Adviser 2011: 9-10). By the time Hague stepped down as Foreign Secretary, Britain's Libya policy was hotly contested, but his March 2011 venture had seemingly been forgotten, even by close observers.
Second, the short-lived attempt by journalists to create a fiasco narrative is in part explained by the structuring effects of shifting media agendas. Soon after the story broke, an earthquake and tsunami struck Japan, killing around 20,000 people and causing a nuclear accident. The next eight weeks also included a Royal Wedding in the UK and the death of Osama bin Laden. Media attention can only last so long: the agenda was soon overtaken by otherequally dramaticevents, ones with obvious global political consequences that made for better news stories. It would require a separate article to theorize the association between commercial media decisions and fiasco construction. However, the evidence from our case study suggests that whilst the saliency of Hague's Libyan venture was very high for a brief period, more sense of intrigue or consequencenot to mention a whiff of scandal, cover-up or other wrongdoingwas required for stakeholders to continue to invest resources in chasing down elusive facts.
With or without the agglomeration of more newsworthy stories, this fiasco construction might not have taken off for the third reason: the requirements of operational secrecy hindered the media's ability to establish the facts of the matter. The case did involve the transfer of information via parliament, which challenged the executive's usual monopoly over such matters. However, official information was limited and framed blandly, thus taking the sting out of the fiasco narratives. Observers lacked even basic data on the operation to be able to explain agents' behavior and to apportion blame. The Guardian, for instance, ended up turning its attention to the 'arrogant' government silence (Norton-Taylor 2011). Moreover, the operation came amidst a fastmoving civil war, part of the Arab Spring, which was ushering in a period of intense global uncertainty reminiscent of the collapse of Communism in Eastern Europe. At the grand strategic level, the government had secured bipartisan support for its actions in Libya, including the need to protect civilians in the civil war. Negotiations were already underway regarding no-fly zones, and just two weeks later MPs overwhelmingly voted in favor of military action to protect civilians. The press supported these key decisions. This strategic consensus clearly sapped the willingness even of Hague's many opponents to continue to try to land political punches on him.

Conclusions
Our investigation of Hague's Libyan venture highlights the multiple ingredients required for the successful construction of a fiasco narrative: politics, power, intent and contingency. Our case study supports a well-known finding in the field of Foreign Policy Analysis (Houghton, 2007: 25) that information is power as well as being the main currency of foreign policy narratives. The foreign policy marketplace, as Baum and Potter (2008: 42) put it, 'is driven primarily by the distribution of the key market commodity'informationamong the policymakers, media and public opinion. Our contribution allows us to add to this argument by showing precisely why it is that the emergence of an accountability forum within which information can circulate is crucial, yet also one of the hardest things to achieve for interested stakeholders. Information, especially in covert action, overwhelmingly favors the policy-makers.
Our first conclusion is, therefore, that lack of access to information is the prime reason why a perceived covert action failure, and a national security failure more broadly, will not easily result in the successful construction of a fiasco narrative. Secrecy, government monopoly over information, and the problematic ontologies of identifying policy goals impede even those operations deemed to have failed from becoming talked about as fiascos. This creates a fascinating paradox considered in this article: public policy failures are more likely to become emplotted as fiascos despite being less controversial and less risky, simply because they are more visible and stakeholders have more information ready to hand. Even short-lived, failed fiasco constructions such as that explored in this article offer vital information on the subjective nature of fiasco narratives.
Our second conclusion is that the role of secrecy is determinative when evaluating covert action and yet is only loosely acknowledged and accounted for in existing attempts at mid-range theoretical explanations of policy failure. This article has not sought to develop its own theory of secrecy, but has introduced the idea of secrecy into scholarly debates as a point for discussion moving forward. Our article has revealed that narratives surrounding covert action failures exhibit different emphases and some special characteristics compared to 'overt' or routine foreign policy activity. However, we are dealing with differences of magnitude not of kind, hence the urgent need for attention to the idea of secrecy in the fiascos literature. From our evidence, we suggest that secrecy can be a government's best defense against fiasco construction. Yet once commentators latch onto covert action mistakes, elites are suddenly placed in a difficult position. They cannot fully defend themselves without undermining official secrecy and/or potentially damaging delicate diplomatic relations with other states, or criticizing Special Forces and the intelligence services, which have long been romanticized in the public imagination. We argue, therefore, that attempts at fiasco construction offer insight into the power relations at play in attempts to construct and control a narrative. The study of covert action fiasco narratives is especially well placed to deliver rare insights into the most secret of policy spaces.
Our third conclusion is that secrecy seriously complicates lesson learning. Many studies, in asserting their policy relevance, argue that opening policy-making to a genuine contest of ideas is the best way to prevent fiascoand failure more broadly (Bovens and 't Hart 2016: 662). Yet covert action is the one area of policy least likely to be subject to democratic debate, even in an era of increasing official oversight and growing public interest in foreign policy. At the same time, blame games and political point scoring associated with fiascos, exaggerated when a controversial covert action is under the microscope, can obscure relevant lessons altogether (Gray 1998).
Our fourth conclusion is that, from the evidence presented above, policymakers can learn vital things about the construction of blame avoidance under the rubric of 'message management'. Narrative analysis is central to the enterprise. It is all too easy for policy-makers to assume that covert action will stay secret; they therefore tend to underestimate the political implications of disclosure. Governments communicate all sorts of messages about policies daily, but there seems little energy devoted to reflecting on the whys, wherefores, pros and cons of different types of message management under routine and crisis conditions. Studying the construction of covert action fiascos lays bare these practical challenges to policy-makers by showing the obvious potential for fiascos scholarship to impact on policy. This is no straightforward task and will require further bridges to be built between the worlds of policy and academe. Cross-fertilization of research and experiences between academics and foreign, security and intelligence policymakers, in line with the revivified research agenda of which this article is a part, will perhaps encourage governments to better consider the political ramifications of disclosure before sanctioning covert actions which, in an increasingly digital world, are ever less likely to remain secret. Notes 1. We are grateful to one of the journal reviewers for drawing our attention to the interplay between domestic and international politics as a source of novelty.
2. The hostage rescue was a secret military operation not a covert action, and it is debatable whether the Iran-Contra affair, often labelled a scandal, constitutes a fiasco (the latter hinges on interpretations of programmatic success). The differences between them lie beyond the scope of this article, but are referred to here as they are relevant to the core theme of secrecy. 3. The newspaper source data and references used in the narrative analysis are found in the online appendix.