Terrorist Decision Making in the Context of Risk, Attack Planning, and Attack Commission

ABSTRACT Terrorists from a wide array of ideological influences and organizational structures consider security and risk on a continuous and rational basis. The rationality of terrorism has been long noted of course but studies tended to focus on organizational reasoning behind the strategic turn toward violence. A more recent shift within the literature has examined rational behaviors that underpin the actual tactical commission of a terrorist offense. This article is interested in answering the following questions: What does the cost–benefit decision look like on a single operation? What does the planning process look like? How do terrorists choose between discrete targets? What emotions are felt during the planning and operational phases? What environmental cues are utilized in the decision-making process? Fortunately, much insight is available from the wider criminological literature where studies often provide offender-oriented accounts of the crime commission process. We hypothesize similar factors take place in terrorist decision making and search for evidence within a body of terrorist autobiographies.

The rationality of terrorism has been long noted. 4 Previous research has shown how terrorism can be an effective political strategy, 5 more optimal than other forms of military engagement, 6 can produce effective gains, 7 and can undermine confidence in the functioning and authority of the state. 8 Whereas these studies traditionally focused upon the rational adoption of terrorism as a strategy or a tactic, a nascent empirical literature examines the rational behaviors underpinning the actual commission of a terrorist offense. Such approaches focus on the situational qualities of terrorist behavior (e.g., what terrorists do and how they do it) and are largely informed by developments in the area of environmental criminology and situational crime prevention. These studies cover issues like target choice, 9 weapon choice, 10 the spatio-temporal clustering of offenses, 11 the distances traveled to commit a terrorist attack, 12 victimology, 13 and the displacement of incidents. 14 Studies such as these are, according to Freilich and Lafree, "rare exceptions" that "apply systematic empirical methods to test [criminological] theories with data on terrorism." 15 It is no surprise therefore that the "criminological study of terrorism has lagged far behind many other specialized branches of criminology." 16 However, there is currently no first-hand understanding within the terrorism literature on offender decision making in the context of risk. 17 For example, we have no answers for the following questions: What does the cost-benefit decision look like on a single operation? What does the planning process look like? How do terrorists choose between discrete targets? What emotions are felt during the planning and operational phases? What environmental cues are utilized in the decision-making process? Fortunately, much insight is available from the wider criminological literature where studies often provide offender-oriented accounts of the crime commission process. We hypothesize similar factors take place in terrorist decision making and search for evidence within a body of terrorist autobiographies. Altier et al. 18 and Shapiro 19 previously demonstrated the utility of analyzing terrorist autobiographies. Terrorist autobiographies offer insightful, granular-level data points for the study of decision making in the context of high-risk offending. This is particularly important given that the covert nature of terrorist activity can make systematic data difficult to obtain. For Altier et al., "the careful qualitative analysis of these texts provides valuable insight into the ways terrorists perceive their social world and the context, which shapes their attitudes and behavior." 20 In total, we analyzed ninety-nine terrorist autobiographies. Upon a thorough examination of these autobiographies, forty-nine offered detailed information regarding the planning for, and/or commission of, at least one terrorist event. The others largely focused upon other aspects of their developmental years, grievance formation, radicalization, nonviolent life in a terrorist group, prison life and/or ultimate disengagement. The ninety-seven autobiographies covered a range of ideologies including ethno-nationalism (62.6 percent), left-wing (20.1 percent), right-wing (8.01 percent), religious (6.01 percent), and single-issue (3.1 percent).
This exploratory study aims to consequently contribute toward the currently limited body of literature exploring terrorist decision making and attempts to define what future research may look like. This form of analysis also holds promise with regard to effectively managing and controlling the extent of terrorist behavior via situational crime prevention approaches, crime prevention through environmental design, and intelligence-led and community policing. We suggest that by identifying the key determinants of terrorist decision making, it may be possible to design interventions that heighten the risks and necessary effort for effective terrorist activity, as well as lower its rewards.

Variance in the Planning Process
This section looks at the degree to which terrorist attacks are typically planned and the factors that shape this process. Criminal planning should be viewed as a continuum rather than a planned versus unplanned dichotomy. 21 In a descriptive meta-analysis of criminal decision making, Gill et al. demonstrated that across fifteen studies, the number of "planners" ranged from 13.5 percent to 87 percent with a mean of 39.6 percent and a median of 37 percent. 22 The extent of the variance was largely due to the levels of complexity in the crime. The firsthand accounts of terrorist attack planning report similar variances. At one end of the spectrum are accounts of attacks being "more or less spontaneous" 23 and involving "no great pre-planning … done in minutes." 24 Whereas at the other end, attack plans developed for up to six months 25 and, in the case of the PIRA, an ongoing and sophisticated intelligence gathering process: The priority should be to develop and plan combat intelligence, and in particular target intelligence-that is, the information which unit commanders would need to mount successful operations against enemy personnel or to sabotage enemy installations. … I issued instructions to Intelligence Officers (IO) that they should study the daily and local newspapers carefully, and indeed read every serious magazine and periodical they could lay their hands on. … Afterwards, when the war in the North was at its height, many a Republican officer put to effective use information which his IO had got by following up an interesting news item or press photograph. Something like eighty-five per cent of the intelligence collected … comes from media and other open sources. 26 Such variance in attack planning occurs for a number of reasons other than the degree of complexity needed. First, it may be due to how the offenders themselves characterize planning. At times, the terrorists were, like in the words of Jacobs, "alert opportunists 'half-looking'" for vulnerabilities to exploit as they carried on with their day. 27 Such serendipitous arrangements can often make planning look "almost effortless." 28 Eamon Collins, for example, reported he "never stopped looking for military targets." 29 Jane Alpert's account of a Weathermen Underground attack on a United Fruit warehouse (a major player in the Latin American economy) is pertinent here also. She reports only becoming aware of the attack plans the same day it took place. Her co-offender Sam raised the idea. He had passed "it lots of times," noting the United Fruit sign "in mile-high letters across the top" during his everyday activities. Alpert perceived this sign as "proud ignorance" on the part of the business. In the days preceding the attack, Sam had "been checking it out" in more detail for the purpose of an attack, observing it "deserted after six o'clock at night." 30 Such examples of attack plans developing during the course of a terrorist's routine daily activities were very common. Similarly, the decision to conduct an attack may be related to a suddenly emerging opportunity. For example, PIRA's Brendan Hughes recalls an open-back British Army jeep driving unaccompanied into a PIRA stronghold: "At this time the British Army would never come in unless heavily armed and in armoured cars. This particular day we weren't expecting anything like this. … Here was something that just came out of the blue. … We were so confident and in such control of the area at that time that instinct took over: 'There's a target' and 'Hit it.'" 31 Variance in attack planning within a single cell of co-offenders was also repeatedly made clear within the terrorist autobiographies. Returning to Alpert's example above, Sam conducted planning for months but Alpert only became aware of the attack on the day itself.
Similarly, Martin McGartland's outline of PIRA activities typically depicts individuals specifically equipped with reconnaissance tasks who then pass information on to those who conduct the violent attack within hours. 32 Finally, the variance in attack planning may be due to how sophisticatedly the terrorists believe counterterrorism forces operate. For example, Bradley's account of life in PIRA details early attacks as "pure spur-of-the-moment … target of opportunity." 33 Increased surveillance by the security services later meant PIRA "had to plan weeks ahead, maybe have an op on the shelf waiting for the right circumstances, maybe almost do a dry run. You had to time ops to perfection: how long it took to walk a certain distance, drive a certain distance; plan the run back, make sure you could get the clothes washed or burnt, get a shower. … Just going out shooting was out of the question." 34

Determining Risk
No matter the length of the planning process, the autobiographies consistently offered accounts of terrorists weighing up various risks and benefits during the planning phase. The factors considered encompass both subjective and objective factors and, in many ways,, mirror criminological findings related to criminal cost-benefit decision making.

Subjective Factors
This section looks at internal states and emotions during the criminal decision-making process with a particular focus on what is felt and how these feelings are managed and regulated. Within the general crime literature, depictions of how fear and nerves negatively impact the decision-making process has been noted for street robbers, 35 first time sex offenders, 36 shoplifters, 37 (auto-)thieves, 38 burglars, 39 card fraudsters, 40 muggers, 41 and (armed) robbers. 42 Such feelings were regularly noted in terrorist accounts also. For example, various accounts self-reported feeling: "very solumn," 43 "stress … hand shaking … nerve-racking … agonizing," 44 "terribly nervous," 45 "nerves would burn," 46 "very tense," 47 "could not sleep … flashes of doubt," 48 "increasing frustration … paranoia," 49 "anxiety … fear," 50 "unnerving," 51 "frayed nerves … growing premonition of disaster … diaphragm tightened … heart thumped like a drum … excitement and fear," 52 "anxious and a little scared," 53 "apprehensive," 54 and "ravenously hungry … caused by nervous tension." 55 These feelings were typically elicited by conscious knowledge of the risks of arrest or death, 56 or in another case something as frivolous as outside noises. 57 These subjective feelings appeared at their most intense during the commission of an attack. Accounts self-reported feeling: "panicked," 58 "racing pulse," 59 "heart was beating fast," 60 "knees going weak … hair standing on end," 61 "legs … like jelly," 62 "mind was racing," 63 "heart was beating in my throat," 64 "adrenalin rush," 65 and "heart thumped loudly and painfully … panicky thoughts … shaking … incredible pulse and trembling." 66 These feelings were brought on by experiences such as hearing an unexpected noise, 67 an alarm sounding, 68 and the conscious awareness of security measures in place at the scene.
For some, these feelings continued after the attack. "Some forty-five minutes later we pulled up on the side street near our apartment. Still I felt panicky and my stomach churned uncontrollably." 69 For others, it was over "when it was clear that no one had followed us" 70 or "once they were on the road [to the attack] the unit quickly forgot about the nerves and concentrated on the job in hand." 71 For others, the nerves dissipated during the attack itself but usually this was the result of habituation from previous similar activities: "I felt curiously calm the whole time as if everything was happening in slow motion." 72 Sean O'Callaghan's account of the murder of Special Branch Officer Peter Flanagan highlights how the mood within a single attack cell was one of joint nervousness on the night prior to the attack. 73 However, once the murder was conducted, the three-team cell reacted very differently. While O'Callaghan willed himself "to be calm, to act as normally as possible," Lulu (the getaway driver) was full of nerves and accelerated the car before O'Callaghan had got into his seat. She also drove against the traffic on a one-way street. The third offender, Paul, was "giggling like the teenager he was … in a state of near-euphoric excitement." The heightened emotional state continued for weeks for Eric Rudolph: "My days were spent looking over my shoulder. My imagination began to play tricks on me. I started seeing unmarked cars in my rearview mirror and strange men shadowing me in the grocery store. I checked the wheel walls and undercarriage of my truck for tracking devices before going out. The mere mention of Centennial Olympic Park on the radio sent my blood pressure soaring. … Paranoia grew on my psyche like lichen moss. … Worrying about the feds weighed on me like a lead albatross around my neck." 74 Criminals self-report substance abuse during the crime commission in order to help overcome fear and nervousness. 75 Upon being apprehended, Breivik announced that he had taken a combination of ephedrine, caffeine, and aspirin (Breivik referred to this combination as an "ECA-Stack") in order to enhance performance. His manifesto outlines that the ECA-Stack "will significantly increase your strength, agility and focus … up to 30-50% for 1 to 2 hours after taking one capsule. This enhancer, in combination with a steroid cycle, will increase your physical and mental abilities by up to 100%, transforming you into an unstoppable one-man army when used in combination with proper training and a full range of body armor and weapons." 76 In the words of the psychiatric report, "he had taken the substance in order to achieve as much as possible during the operation." 77 Breivik tested positive for a higher than normal concentration of ephedrine. According to the psychiatric report, "higher doses and higher blood concentrations may give intoxication symptoms where increased confidence, increased risk-taking and loss of critical skills may occur." 78 The doctors who examined Breivik stated further that he "may have been under the influence of caffeine to such an extent that a moderate intoxication effect cannot be ruled out." 79 Breivik also claims to have taken two forms of anabolic steroids from the spring of 2011 until the day of the attacks, and a mixture of ephedrine, caffeine, and aspirin for nine weeks prior. Breivik refers to the use of these drugs as a "military strategy." 80

Objective Factors
The first-hand accounts highlight the degree to which security measures are salient in the planning phase. For example, terrorists reported avoiding "any place with a camera," 81 searching for signs of "human presence," 82 and considering the impact of poor weather on formal and informal surveillance. 83 It should be no surprise that the types of targets attracting terrorist attention typically possess security features. This section focuses on the how these factors impact decision making. There are six main conclusions and each are elaborated on below. First, terrorists often keep several potential targets in mind and choose the one with the relatively fewest risks. Second, cost-benefit analyses differ across terrorist groups and terrorists because risk preferences differ. Third, previous successful experiences decrease averseness to risk. Fourth, the weighing of security features necessitates hostile reconnaissance which itself offers risk to the terrorist in terms of detection. Fifth, what matters are perceptions of how effectively deployed security is. Sixth, third-parties often play key roles in gaining intelligence for an attack.
1. Terrorists often keep several potential targets in mind and choose the one with the relatively fewest risks. Evidence from some autobiographies suggest that plotters keep several potential targets in mind and choose the one with the relatively fewest risks. For example, Wilkerson's account of decision making in the Weathermen noted: "When the proposal was floating about [targeting] Fort Dix, no one argued against it, but the tension in the air seemed to crystallize into a fine mist. … As yet, however, we knew nothing concrete about the base, or exactly what we were talking about or whether it would be possible. We agreed to investigate other targets as well. … One team went to each of the possible sites to do reconnaissance … [once completed]. … The conversation focused on which of the targets we had investigated were feasible. Then we discussed the logistical details required for each action." 84 Anders Breivik considered several targets including assassinating a member of government. He opted out of this particular plan because he thought that it would be too difficult, because such individuals are often well protected. Breivik stated that it was necessary to go further down the list of acceptable targets for this attack. From an operational perspective, he saw Utøya Island as ideal because it was "isolated" and "police would have problems" accessing the site. Eric Rudolph considered several abortion clinics to target with a remotely detonated improvised explosive device. He decided against the first clinic: "I didn't like the lay of the land. I'd be using a remote control system, so I focused on finding a place to put the bomb, and a place where I could detonate it from. Like a sniper, I needed line-of-sight to the target." He decided against the second clinic also: "High bushes surrounded the building, and escorts guarded the only entrance through the wall of bushes. The place had no vantage point from which I could look down into the parking lot.. 85 Michael Stone's first assassination target was the Sinn Fein politician Owen Carron. Initially Stone surveilled Carron's home address: "I knew he had two dogs. … I knew that all over the house and garden he had the best security and surveillance equipment money could buy. He had cameras and sensors. He even had tin cans tied to a tripwire strung across the field at the back of his house to alert him to the security forces that watched his every move. … I ruled out attacking him at his home because he had too much security and I could not get close enough to kill him without being spotted or killed myself. My best option was his constituency advice clinic … [it] was the weakest link in his daily routine." 86 2. The cost-benefit analyses differ across terrorist groups and terrorists because risk preferences differ. For those groups where consistent attacks are the modus operandi, evading detection, arrest, and death is key. Cost-benefit analyses also differ across individuals within the same movement. For example, in March 1988, Loyalist Michael Stone single-handedly attacked an Irish Republican funeral in Belfast with grenades and firearms. Stone hoped to "take out the Sinn Fein and IRA leadership at the graveside." Faced with thousands of mourners as well as policing and army units nearby, this was undoubtedly a highly risky attack. "Most of the time it was 50:50. I figured [this attack] would be at least 60:40 against me, but could even be less." Stone, however, felt the benefits were too great to pass up. "I believed it was worth a risk if it meant the leadership of the Republican movement was wiped out." 87 Other Loyalist terrorists, however, self-report an inability to face such risks. Johnny Adair, for example, said: "What Stone did astounded me. There was no way I would have attempted anything like that. I wouldn't have had the nerve." 88 3. Prior successful experiences decrease averseness to risk The criminological literature highlights that experience of not being caught for previous crimes downplays immediate situational risks. 89 On a related note, experience of previous crimes bolsters self-efficacy and self-belief. 90 Evidence of this within the terrorist autobiographies was also common: A steady diet of small illegal activities had boosted my confidence in our abilities to get away with things. I no longer imagined a cop hiding behind every obstacle and actually found myself feeling quite relaxed out on a mission. Still a certain level of fear is a good thing in these matters-it keeps a healthy flow of adrenalin coursing through the bloodstream, which tends to heighten awareness. I think I had finally reached this healthy medium. 91 The Hansen quote here is particularly illuminating. It shows that experience coupled with increased self-efficacy and self-belief helped her interpret the adrenalin rush as a positive factor as opposed to it being a negative factor associated with fear, nerves, and anxiety.
The willingness to accept risk is sometimes due to the high levels of planning previously conducted, in particular the amount of human and financial effort spent increases the cost of not conducting the attack. Hansen outlines one particular reconnaissance job. Months prior, one reconnaissance team noted the factory was empty over the weekends and was ideal for an attack that would cause no fatalities. A second reconnaissance mission noted a large number of employees present on a Friday night. The factory had: added a night shift. … It was obvious that we would have to case the plant to determine whether there was a night when no one was working, but what if there wasn't? Would we go ahead anyway, and take whatever precautions necessary to make sure no one got hurt? Even though that little voice inside me said we shouldn't go ahead and take the risk, my sense of reason told me that we had come too far and invested too much to stop now. I knew Brent wouldn't turn around and go back and neither would I. 92 4. The weighing of security features necessitates hostile reconnaissance which itself offers risk to the terrorist in terms of detection. The conscious awareness of these objective security factors often leads to doubts, irregular behavior, and an almost paranoid state where the terrorists often over-exaggerate the degree to which they are being watched and the number of security measures. For example: I had spent a week earlier that month casing [the target]. … There was no hideout near the target, no place of concealment where I could sit and scope it out, so I did my scouting on the move. After parking a mile away, I made three passes … each day. I strolled past it at different times of the day, approaching from a different angle and wearing a slightly different disguise on each pass. After each pass I'd find some dead space-behind a grocery store, apartment complex, or hedge-and change wigs and a couple articles of clothing. Then, I'd swing back … from a different direction. Gathering new information from each pass. I gradually pieced together a plan. 93 As I was about to open the door, a woman came out of an office around the corner. I answered her gaze with a sheepish smile and turned back toward the elevator. The woman kept going and disappeared into the next office. I could not now return. … If the woman saw me loitering there again, she might call a guard … [she then places the bomb in a different part of the building]. … I rode the elevator down to the lobby alone. … A guard stared at me as I went out the glass doors. Had he noticed that I came in with a pocketbook and left without one? I turned my head to avoid giving him a good look at my face. 94 With my concentration focused on the security guards patrolling the walk paths, I barely noticed a young man kneeling down in front of me. When I looked up, he snapped a picture. The flash stunned me. Like a deer caught in the headlights of an oncoming automobile, I just sat there dumbfounded, watching him walk away. He had photographed the fountain behind me, but captured my face in the frame. "I cannot go through with it," I thought. 95

What matters are perceptions of how effectively deployed security is
Criminological studies generally highlight that offenders' perceptions of how security is deployed as opposed to their presence is what matters in their risk calculus. 96 This was evident in Gerry Bradley's account of his PIRA activities. In particular, the use of helicopters and sangers in surveillance: "The chopper destroyed us. If the chopper was up, you weren't allowed to move out of a house. … Ops were cancelled regularly because of it. They could read newspaper over your shoulder from the chopper." "There was a sangar on Templar House. That one did so much damage. … Dozens of ops had to be called off because of it. They could watch the whole district. … Between the chopper and the towers, maybe 80 percent of ops had to be called off, maybe more." 97 Evidence also suggests that terrorists actively look for poor deployment of security. Eamon Collins, for example, outlines surveilling an alleged off-duty member of the Ulster Defence Regiment: "I waited and waited, but there was no sign. … I was just about to go when my patience was rewarded. A slightly-built man in his early thirties came out of the pub and walked up to the car. … He opened the car door and got in. This casual action surprised me: why had he not checked under his car for a bomb? … I took down his registration number, but I decided not to follow him that night. I had made a good start. I had a definitive description-a face to put to the information-and the guy seemed careless about his own security." 98 Eric Rudolph's reconnaissance of the Atlanta Olympics Park similarly noted: "Hundreds of security guards and cops patrolled the park. They eyeballed me going through the entrances. But there were no metal detectors, and bags were searched selectively. After sundown the crowds grew enormous. … Security at the park became overwhelmed. They stopped searching bags altogether, and the entrances flew wide open. I knew then that I could smuggle in a bomb." 99 6. Third-parties often play key roles in gaining intelligence for an attack Several criminological studies show that third parties not involved in the actual crime can spot opportunities and pass this information to the eventual offender. This has been noted for captive takers, (commercial) burglars, (armed/street) robbers, and drug dealers. 100 Such third-party intelligence was reported in first-hand accounts from both PIRA members and Loyalist terrorists. Both drew on local service and trade workers, or listened in to police scanners. 101 On occasion, terrorist groups specifically plant members in particular occupations to collect key information. 102 Paul Hill's plans crystallized with a tip-off from a fellow antiabortion activist: "I continued to secretly consider shooting an abortionist, half hoping it would not appear as plausible after I had given it more thought. The next morning, Friday, as was my practice, I went to the abortion clinic (the Ladies Center). I arrived at about eight o'clock, the time that many of the mothers began arriving. I was usually the first protester there, but that day another activist had arrived first. … I learned that he had been there when the abortionist arrived: about 7:30. More importantly, I discovered that the abortionist had arrived a few minutes prior to the police security guard. This information was like a bright green light, signalling me on."

Restrictive Deterrence
Absolute deterrence refers to an individual completely opting out of engaging in crime because of perceived risks. Restrictive deterrence, on the other hand, refers to how offenders still commit crimes but make conscious decisions to change behavior in how they commit crimes in order to minimize risk. It comes in many forms. First, offenders may choose objectively less risky crimes. Direct Action's Ann Hansen, for example, outlines the group "decided to keep the actions small and simple so that people could get involved without having to fear serious prison time as a consequence." 103 Second, groups may choose inconspicuous members to conduct particular operations. Of female PIRA members, Gerry Bradley noted: "They usually went well-dressed, like office workers. They knew how to get past checks. Women were also brilliant at intelligence work, pretending to walk kids, push prams, get into places, nose around." 104 Similarly, days before his bombing Eric Rudolph entered Centennial Park "to test the security at the entrance." On this occasion, he noted the temporary security fence and that "it was impossible to avoid the cameras. There were simply too many of them." To circumvent these security features he decided to wear a "light disguise" on the day of the attack. 105 Third, attacks may become less frequent but more spectacular because of the need for increased planning. Returning to Bradley: "Operations were becoming increasingly difficult and dangerous, with all the complications already described: forensics, surveillance, ubiquitous helicopters, the security forces' more sophisticated understanding of IRA methods, quicker reaction by police and army, fewer opportunities because of better field craft by the security forces. There, while operations became less frequent, they tended to become larger set-pieces, the planning and execution of which were lengthy and painstaking and meticulous." 106

Conclusion
This article has been interested in explaining the development of a terrorist event as opposed to the development of a terrorist. The two are very different strands. The latter is concerned with understanding psychological and social influences. Putting aside the debates and contestations within that realm of research, even if it had perfect answers, it would still be unable to explain the terrorist event process. At best, it might provide insight into areas related to formal (e.g., criminal justice system) or informal (e.g., socialization into societal norms) social control-related endeavors that seek to punish, prevent, disrupt, or rehabilitate the offender. The successful implementation of a terrorist attack, however, relies on not only the presence of a motivated offender, but also requires a lack of capable guardianship, and the presence of a suitable victim. In other words, it needs opportunity.
Terrorists make cost-benefit decisions in much the same way as ordinary criminals. The field of crime prevention evaluation is testament to the vast potential for situationally focused crime prevention approaches to reduce crime. This means focusing on the settings in which offenses take place, rather than the underlying motivation or criminal disposition of the individual. Reducing the opportunities for terrorism via environmental design broadly construed is, therefore, a valid and worthwhile pursuit. Each type of terrorist attack, be it a vehicular assault or a bombing, depends on a crystallization of multiple opportunities. In turn, each specific attack type offers its own set of particular environmental opportunities that can be manipulated with the intention of impacting the terrorist cost-benefit calculus. Such endeavors increase the effort via target hardening, controlling access to facilities, deflecting offenders, and controlling access to the necessary weapons. They also increase the risks by extending guardianship, assisting with natural surveillance, reducing anonymity, utilizing place management, and strengthening forms of formal surveillance. They may also reduce the rewards of an attack by concealing or removing potential targets.
The results also highlight the opportunity to manipulate emotions in order to disrupt terrorist decision making. Fear appears to be ubiquitous in offender decision making. This is true across the stages of a crime (pre-planning, execution, getaway). It is also true across a wide range of crimes. Fear functions to ensure a state of readiness among those contemplating crime and as a compensatory mechanism to alert individuals to the potential risks. Interventions that aim to increase such fear during situational decision making should therefore be beneficial. The sources of fear are multiple. They include objective features such as fear of detection by police/security, fear of detection by conscientious bystanders, fear of the unknown, fear of co-conspirators being deceptive, and fear of interaction with others. Fear of the "unknown" is also paramount. The source of fear also includes subjective features of the environment that the would-be offender cannot formalize but subjectively "feels." Such feelings can be multiplied if the would-be offender believes the ability of security to detect suspicious behavior is high. Interventions that therefore highlight, embellish, and evidence the ability of security, staff, and/or bystanders to detect suspicious behavior should have a positive net benefit. As hostiles are assessing a scene for security weaknesses and opportunities, they are just as likely to witness such communications if properly displayed. This helps "nudge" the would-be offender into objectively, bringing these factors into their risk/reward calculation. Interventions that also minimize the volume of available information for hostiles should also therefore increase uncertainty and either lead to full disruption or the hostiles taking ever greater risks to minimize their uncertainty and in turn, maximizing the chances of detection. Finally, the fact that fear is ubiquitous may improve the likelihood of detection. Fear may lead to suspicious behaviors, consciously "acting normally," which itself may appear suspicious, and lead to poor decision making. What suspicious and deceptive behaviors objectively look like requires a great deal of further research.
This article utilized autobiographies as a source and a couple of limitations are worth deliberating on. First, those who write autobiographies may not be a representative sample of all terrorists. Their drive and ability to write often erudite and engaging depictions of their ideological motivations, illegal activities, and self-reflections perhaps makes them atypical. The degree to which this presumed lack of representativeness impacts the generalizability of the results is specific to the research question posed. We suggest that research questions focused on motivation, engagement, and disengagement narratives, in particular, are potentially susceptible to such biases as they are quite idiosyncratic. Here, we have attempted to mitigate such biases by using insights from the wider offender decision-making literature as a foundation for our findings. That literature is informed by a wide array of criminal types, with data collected in different settings (prison and fieldwork), with differing data collection methodologies (structured interviews, semi-structured interviews, experimental designs, surveys, etc.) deployed on a wide range of demographic populations. The terrorism-specific findings depicted here in this article are simple replications of what the criminological literature has long noted about how offenders engage in cost-benefit decisions on specific criminal events.
Second, few autobiographical accounts from jihadist sources informed this article. Simply, such accounts are few and far between. Where they do exist, operational details are even rarer. Instead, aspects related to how grievances formed and ideologies were adopted dominate. It is fair to say the average jihadi has a greater proclivity for mass-scale atrocities and self-annihilation during an attack than those terrorists largely covered in this article. It would also be fair then to assume that their calculations of costs and benefits would be wholly different than what we depict here, thus making our results inapplicable to jihadis. However, other primary source documents reassert the applicability of our findings to Western-based jihadis. For example, criminal indictments, criminal complaints, and prosecutions of those inspired by ISIS in the United States highlight issues regarding risk preferences varying within a single cell, 107 fears and anxieties, 108 the success of others downplaying fears and anxieties, 109 conducting hostile reconnaissance, 110 the sense that some targets are too difficult because of security features, 111 weighing up several potential targets at once, 112 the salience of objective factors leading to strange behaviors that may lead to detection, 113 the perceived effectiveness of security features being taken into consideration, 114 third-parties informing risk decisions, 115 and various behaviors related to restrictive deterrence. 116 Jihadis therefore engage in a similar decision-making process focused upon cost and benefit as that of terrorists of other ideological shades. The risk of detection, disruption, arrest, and conviction looms large even in plots where the offenders intend to die. Getting caught prior to committing jihad promises a long-term prison sentence without any of the benefits. This causes uncertainty and doubt. The search for further information has the intention of reducing this uncertainty and making success more likely. All of these issues provide potential points for counterterrorism practice to impact on decision making.

Funding
This work was funded by the Centre for Research and Evidence on Security Threats (ESRC Award: ES/ N009614/1).