Foreign Policy Analysis and the Making of Plan Colombia

This article, through a Foreign Policy Analysis (FPA) perspective, analyses the only long-term foreign policy decision ever made in Colombia. Since FPA portrays a theory of human political choice to analyse foreign policy behaviour, this analysis will specifically focus on Plan Colombia’s decision-makers as a case study using empirical examples. The purpose is to understand the specificity of this foreign policy decision-making process from an unexplored perspective, namely Groupthink theory. Although Janis has asserted that the process would negatively affect decision-making quality, this article contradicts this assumption based on both the boundaries and opportunities encountered when applying mainstream FPA to a non-US case study. As such, a major challenge remains when it comes to judging quality and, correspondingly, expecting certain outcomes. This article demonstrates that group cohesiveness and concurrence-seeking tendencies may be useful for explaining successful foreign policy decision outcomes.


Introduction
How was Plan Colombia created and implemented? Who sat at the decision-making table and why? How were the key decision-making dynamics? Analysing foreign policy decision-making requires an examination of a particular decision unit and, by doing so, provides useful generalisations on the nature of decisions. Although we recognise the importance of previous work on group decision-making, we have yet to identify whether the same dynamics are applicable to non-US cases.
In this decision-making analysis, we decided to apply mainstream Foreign Policy Analysis (FPA) that "focuses primarily on the role of the executive (the POTUS and his advisers)", to examine a non-US foreign policy case. 1 The objective was to investigate the role of decision-makers in drafting Plan Colombia, focusing mainly on narrowing the discussion to one specific group of Colombian decision-makers, i.e. President Pastrana and his advisers, as the ultimate decision unit. 2 We seek to understand how Colombia makes its decisions involving the US, and how this actually works in practice. explaining foreign policy behaviour are considered highly influential. 8 Additionally, we reflected on Schafer and Crichlow's work on the effects of the personality of the leader on the quality of decision outcomes. 9 In order to examine this historical case, we developed a qualitative analysis based on historical records, academic literature, press releases and interviews. With this study, we intend to add value from our empirical research in the sense that interviewing decision-makers directly involved in the decision process contributes to strengthening FPA as a human political enterprise, as stated above.
From this approach, we highlight the boundedness of Groupthink theory when applied to a non-US foreign policy decision-making process. Concerning Plan Colombia, the breakthrough this article proposes suggests that Groupthink does not necessarily lead to flawed or defective decisions. As Schafer and Crichlow caution, "process does not always determine outcome" and, although process does matter, as they assert, "it is not determinative". 10 Their contribution on the matter rests on their findings that prove that achieving higher-quality outcomes, following high-quality decision-making, is a matter of probability and that it will all depend on how "good or bad" the factors involved in decisions are. 11 In this article, we intend to contribute to the existing literature by stating that personal diplomacy, be it led by the president or his closest advisors, is an essential factor that will positively affect decision outcomes. In fact, a better understanding of the history of the group, involving the members' friendships with the president and among themselves, and most importantly the history and friendship of these members with the foreign counterpart, which will be either "favorably or unfavorably disposed to the preferences of the decision makers", will provide relevant insights into how successful the result will be. 12 This article asserts that, in general, a well-defined personal diplomacy strategy will ensure that governments succeed in their foreign policy decision-making.
The structure of the article is as follows. First, we provide a description of the FPA status as applied to the analysis of Plan Colombia. This will contribute to an overall review of FPA among Colombian scholars, as a complement to Giacalone's 2015 assessment, which investigates a broad interpretation of FPA based on the debate on autonomy, realism and dependence as the main analytical and theoretical "tools". 13  and public opinion, they have left aside such macro analyses as perceptions, ideas and identity. 14 Secondly, having presented the state of the art in the analysis of Plan Colombia decision-making, we will outline and discuss the theoretical approach. We argue that FPA is applied to understanding Plan Colombia's decision-making process from a small group dynamics perspective, specifically looking into the effects of group cohesiveness. Additionally, we will take into account the president's leadership and role as a subsidiary theoretical perspective for this study.
Thirdly, we portray the weaknesses and challenges that non-US case-specific analysts face when analysing foreign policy. We argue that small group dynamics followed by a strong presidential role in the creation of this particular foreign policy reveals the high relevance of Janis's Groupthink theory.

A Foreign Policy Analysis of Plan Colombia
From its start, Plan Colombia has been widely studied by the country's academic community. Efforts have been focused on identifying the origin, agenda and consequences of what has been considered a sort of Marshall Plan for Colombia. 15 Additionally, studies have been focused on its interventionist nature, within the framework of the war on drugs directed from Washington and the intention to internationalise the conflict by the Colombian Foreign Policy Executives (FPE). 16 Finally, its fateful effects on the environment have also been studied. 17 It has become evident that it qualifies as a failed strategy in the war 14 on drugs and in the solution of Colombia's internal conflict, in addition to an overall impact (of the war on drugs) over several Latin American countries foreign policy and security concerns. 18 While this research agenda is timely, it is limited in respect to exploring the decision-making processes that led to the formulation of Plan Colombia. Most authors focus on the needs of President Andrés Pastrana after his failed peace process involving demilitarising a large area of 42,000 km 2 over almost three years (1999)(2000)(2001)(2002). The situation worsened with the death of three American activists -Terence Freitas, Ingrid Washinawatok and Laheenae Gaywho had been kidnapped by Fuerzas Armadas Revolucionarias de Colombia (FARC) in February 1999, as well as the resignation of the Minister of Defense, Rodrigo Lloreda, and a strong offensive by Fuerzas Armadas Revolucionarias de Colombia (FARC) at that time. 19 Focusing on the International Relations (IR) literature, we find notable contributions. Arlene Tickner describes the process of creating and negotiating Plan Colombia, including interviews with members of the American FPE such as Phil Chicola, Director of Andean Affairs of the US Department of State. 20 Additionally, Tickner analyses how the war on drugs was fought under the realistic American view, with negative effects for the country relating to human rights, and with Washington markedly intervening in Colombian domestic affairs, which did not allow for the development of more inclusive economic and social policies. 21 An important aspect pointed out by the author in this regard is that "[t]he management of bilateral Colombian-American relations was being delegated to a small number of individuals reporting directly to the President". 22 We will return to this later, as it is one of the main purposes of our work.
We found another contribution to the FPE and the decision-making process in the work by academic and diplomat Diego Cardona. He points out that Plan Colombia was designed by a reduced number of government officers, written in English by a small group of government experts, and as we learned throughout interviewing Jaime Ruiz, Head of the Departamento de Planeación Nacional: "What is essential is that it had first been a government plan that had not even gone through the Council of Ministers". 23 In our view, the analyses made by Cardona and Tickner come closer to the exclusive decision-making process that belongs with the president and his immediate circle. Cardona points out that what was important was to turn Plan Colombia 18 into a state project rather than just a government one. 24 This was favoured by the pro-American Colombian FPE, which did not allow for the promotion of debates or controversy in the country. 25 Regardless of the significant contributions made to the analysis of Plan Colombia, few studies have examined Plan Colombia using an actor-specific approach, with particular emphasis on the president and his close staff. Hence, the following section explores decision-making inside the immediate circle of President Pastrana. This will provide us with information on the dynamics that affected the decisionmaking process during the final period of his term in office, when Plan Colombia was proposed. We will contrast the approaches seen, and see how feasible it is to apply FPA to foreign policy phenomena in the Global South.

Small Group Dynamics: Groupthink Theory
According to Valerie Hudson, "no matter how influential or mercenary, a single leader cannot make and implement foreign policy by himself or herself". 26 Therefore, a foreign policy analyst must focus on the dynamics of a group of people, sometimes called experts, responsible for "making magic happen". While this argument may seem obvious, explaining what goes on inside this core is not necessarily an easy task.
Considering that FPA is the theory of human political choice, we first assumed that Groupthink theory could explain how Plan Colombia was created for two main reasons. First, because Colombian foreign policy-making is characterised by a presidential and personal trait in Colombian diplomacy. 27 Thus, one could assume that group cohesiveness will take place. Second, because Colombian foreign policy is regarded as improvising, lacking coordination, and based on no or limited information. 28 With such characterisation, the players would most likely rely on informal paths rather than following formal rules of behaviour when it comes to decision-making.

Definition
Groupthink is a representation of a conscious state of mind that leads individuals to adopt synchronised patterns of decision-making, which will have an effect on the quality of group decision-making. Conscious individuals are aware of their reality: "your mental state is conscious if you know that you are in it". 29 Therefore, Groupthink affects individuals involved in policy decision-making, even the most rational ones. As Janis points out, Groupthink is a "mode of thinking that persons engage in 24 when concurrence-seeking becomes so dominant in a cohesive inner circle that it tends to override realistic appraisal of alternative courses of action". 30 In such Groupthink scenarios, Janis infers that individuals may be considered "victims", for they have consciously overridden "realistic appraisals of alternative courses of action". 31 In Janis's words, "stupidity is certainly not the explanation", but it is nevertheless a concern that arises when defective policy decisions take place, as a consequence of Groupthink. 32 Janis refers to "symptoms" when describing how individuals belonging to a small group behave, and how this behaviour will necessarily affect decisionmaking. His thesis suggests that Groupthink behaviour (process) will lead to defective decisions (outcome): "poor process is likely to increase the probability of a poor outcome, and vice versa". 33 Groupthink or "groupy" behaviour will be considered faulty and defective. 34 Janis judged individuals who would engage in social conformity, shared illusions and observable weakness towards social and psychological pressures and irrationality, which would all sum up into a "cozy feeling atmosphere" represented through strong group cohesion. 35 Rather than being beneficial, this cohesion reflects, according to Janis, a "deterioration of mental efficiency, reality testing, and moral judgement". 36 In other words, what he has referred to as decision "fiascoes" as a result of concurrence-seeking behaviour among members of the group: "the more cohesive a group becomes, the less members will deliberately censor what they say because of fear of being socially punished for antagonizing the leader or any of their fellow members". 37 Furthermore, Janis warns that if critical thinking is replaced by Groupthink, the more likely it will be for members of the group to adopt "irrational and dehumanizing actions directed against out-groups" (see Table 1). 38 Groupthink is considered highly influential when it comes to explaining how groups make decisions and to establishing the quality of these decisions, which may range between low quality (Groupthink behaviour) and high quality (non-Groupthink behaviour). 39 Steve Yetiv's examination of Groupthink is applicable to understanding why and how the United States went to war by using the Persian Gulf case. This served as a valuable new contribution on how Groupthink sheds new light outside Janis's original analyses. 40 By carefully observing that the president "faced little if any opposition" from the members of his group, Yetiv evidenced symptoms of Groupthink and, as a result, defective decision-making. 41 30. Irving L. Janis, "Groupthink:

Groupthink and Bonds of Friendship
Because Groupthink results in group cohesiveness, this analysis presents friendship as an essential element for understanding "groupy thinking" theory. For instance, concurrence-seeking, shared beliefs, respect for others' opinions, feelings of solidarity and strong levels of amiability are nothing but compatible elements of friendship. Moreover, friends have a preference for themselves, as the most observable pattern this analysis can rely on. After all, "friends exhibit similarities across personality traits, values, interests, attitudes and physical appearance" and, as a result, they unveil lower levels of conflict and rivalry. 42 According to Janis, friendship in decision-making scenarios results in members becoming "less competitive and beginning to trust each other to tolerate disagreements". 43 Therefore, friendship is probably the main cause of Groupthink, because friends will always display the tendency to synchronise. It is clear that members who have worked together will reveal a stronger tendency towards cohesiveness than strangers will. 44 The element of friendship is also present in Yetiv's analysis. When referring to group cohesiveness, he assures us that "it is normal for Presidents to choose friends to join their cabinets", as the president's own strategy is to foster "groupy" thinking behaviour from which he himself, by preserving his original ideas, will benefit. 45 Friendship is also present in his reference to group insulation, where he establishes the importance of social and ideological homogeneity of group members. 46 This particularity will enhance shared values and opinions and, as a result, bolster different symptoms of Groupthink. As may seem obvious, friendship bonds will be stronger in cases where members of the group share past experiences. If this were the case, observable evidence of Groupthink would be easier to identify, as well as other symptoms.
Several symptoms can be identified in Yetiv's assumption of friendship, such as the illusion of invulnerability, rationalisation, moralisation, stereotyping and unanimity. In evidenced camaraderie, the members will probably feel over-optimistic for their shared values, and morality will strengthen their self-esteem; friends will also rationalise similarly and share their stereotype constructions; friends will protect themselves against outsiders, even if it involves sacrificing essential and useful information. 47 Moreover, in his analysis of group insulation, Yetiv refers to the group's exclusive tendency to isolate itself from outsiders, evidence of the mind-guard symptom described in Table 1. 48 But friends alone are not just essential for the president; they are also close friends. Hence, the similarities between Groupthink and friendship are perceptible; cohesion, concurrence-seeking and close-mindedness will intensify. It will almost be as if the group itself became one, a "un pour tous, tous pour un" assertion. Yetiv identified how some leaders rely on influential advisors when matters of national security are involved, adding an essential element of friendship and trust, and creating a more exclusive and smaller inner group inside the existing one. 49 Group cohesion is also present in the concept of "group efficacy". Group efficacy is, according to Hermann, "a group's belief in its capacity to perform a task objective". 50 Although the latter resembles the symptom of illusion of invulnerability where members' over-optimism allows for taking risks, Hermann asserts that the higher the sense of efficacy within a group, the more the group will strive to "persevere in the pursuit of their goal", which may lead to positive decision outcomes. 51 However, at this point there is no evidence that group efficacy will result in highquality decision-making, but the chances are high, as we will attempt to explain in Section 4.
For group efficacy to materialise, the leader, in this case the president, will play a key role in determining the group's own sense of efficacy by assigning each member a task, defining the goal and establishing the desired results. 52 Contrary to Janis's pessimism on the victimising effects of Groupthink, in Hermann's work, group cohesion is thought of as a source of collective efficacy, founded on the knowledge that "good working relationships among group members (group cohesion) increase group efficacy". 53  considered a defective source in decision-making and thus affect the quality of the outcome. Above all, accomplishing tasks is essential for the group. According to Hermann, for group efficacy to take place, the following elements must be examined: the group's past performance, vicariously perceived performance of others, support and encouragement from others, group cohesion and similarity between individual group members' beliefs, all summed up in order to "continue policy commitment facing adverse feedback". 54 Hermann concludes by reflecting upon leadership style or level of personal efficacy as an influence on the group, while the leader's beliefs may be "contagious among his advisors". 55 Nevertheless, Hermann does not value the effects of leaders who exercise a strong hierarchical control (where Groupthink takes place) versus leaders who are more open to critical assessments (notorious absence of Groupthink) among his advisors. Instead, he raises new questions on potential sources of group efficacy.
Based on Hermann's conclusion, establishing a connection between leadership style and the quality of group decision-making is essential for this study. For instance, Schafer and Crichlow have recently (2010) outlined how leadership will increase or decrease the likelihood of Groupthink. They explain how leadership is key in balancing the problem of what "too much or too little" inner circle cohesion signifies. 56 Schafer and Crichlow describe the mechanisms used by the leadership to determine the behaviours a group should seek in order to achieve its goals, in the sense that the leader will be aware of "which procedures and rules will work best for any specific group". 57 Similarly, they argue that: a leader who has a high control orientation might be less open to contradictory comments made by an advisor during a decision episode. A leader who is low on conceptual complexity might set up decision structures that limit the amount of information coming into the process. 58 And, according to the Groupthink logic, a decision process that is closed to opinions, discussions and contradictory comments will be regarded as low, or flawed quality, resulting in low-quality decision outcomes. The same will occur with a process that limits the flow of information.
Additional factors that may help establish a connection between leadership style and group decision-making quality are: "information processing, leader dominance, biases, organization, communication, deliberative and analytical skills, and judgment". 59 Above all, "it makes intuitive sense that who sits at the pinnacle of the government will affect not only what gets done in terms of policy but also how things get done in terms of process". 60 54. Ibid

Presidential Role
For small group analysis and FPA, it is quite evident that the president, as commander-in-chief, will serve as the natural leader in the decision process. He will hardly decide by himself, hence the importance of the advisory system, which we have referred to above as the inner circle members. Indeed, the presidency itself has been seen as an organisation in which "the President's staff extends his capabilities by increasing his 'available attention, knowledge, and expertise'". 61 Personal features and styles of presidents are considered relevant for understanding how the inner circle originated in the first place, or the kinds of structures and processes within the advisory system, considering that "what the President is like can influence what the advisers are like". 62 Hence, different studies on presidential leadership help generate assumptions on this issue. Of all the different variants and categorisations of presidential leadership, for this analysis focused on the theory of Groupthink we selected Preston's findings on the three main characteristics that shape presidential leadership style: power, complexity and prior policy. 63 Besides Preston, the analysis also takes into account Schafer and Crichlow's work based on Lloyd Etheredge's (1975) typology on trait analysis: levels of distrust, desire for power and self-esteem. 64 Recognising how the president appoints his advisors is the starting point for the probable occurrence of Groupthink, which will help explain how and why the president will peacefully lead a non-confrontational group composed, in most cases, of a small group of loyal friends. Yetiv, for example, explains how members of the inner circle, assuming they are loyal, will tend to accept the leader's ideas or decisions uncritically "as though they were equivalent to a group norm". 65 According to Janis, "members think of loyalty to the group as the highest form of morality. That loyalty requires each member to avoid raising controversial issues, questioning weak arguments or calling a halt to soft-headed thinking". 66 Therefore, loyalty is a matter of the utmost importance when it comes to deciding who will accompany the president at the decision-making table. As a consequence, more amiability and esprit de corps will emerge among group members, thus increasing the probability of Groupthink. Table 2 describes the connection between leadership style derived from psychological traits and the probability of Groupthink occurrence, which will later be useful to envisage the quality of the decision-making outcome. According to Schafer and Crichlow's work, psychological traits "may have an effect on the quality of the decision-making process under the direction of the leader, and they may also affect the direction or intensity of policy, thus affecting the outcome of a case". 67 In addition, "the psychological characteristics of the leader will affect such things as organizational structures, precedents in decision making, and who the leader chooses for advisors", which according to the 61  The president will choose loyal staff, as he is proactive, 5 but will also need experts to accomplish his tasks. The president will also seek to do business personally 6 because he feels involved Deliberate decision process, less dependence on the use of stereotypes and analogies Medium probability

Medium quality
Distrustinner circle bias Information will be biased; president will go for preferred expectations because he is strongly attached to his group 7

Hierarchical
The president will previously know which opinions he will consider Closed-off decisionmaking system, high propensity towards conflict 8 Medium probability Hierarchical Firmer leadership. The president will see his advisors as "subordinates" Discussions will become stifled, the president will be inflexible, but he will also push his subordinates at high levels and make the most of situations. 10 However, there may also be conflict 11 Medium probability Low quality  3 The expected outcome originates from speculation based on Groupthink theory which correlates low-quality process with low-quality outcomes and vice versa. 4 Herman and Preston, op. cit., p. 79. 5 Ibid., p. 79. 6 Ibid. 7 Ibid., p. 60. 8 Ibid., p. 54. 9 Schafer and Crichlow, op. cit., p. 50. 10 Ibid., p. 60. 11 Ibid. 12 Hermann (1999) cited in Schafer and Crichlow, op. cit., p. 60. discussion above, will reveal the president's tendency to give preference to their closest friends. 68 Regardless of the different styles, when it comes to making a decision, the president will be motivated, according to Hermann and Preston, 69 either by a cause or problem he wants to solve, or by the wish to increase his power or status. Alternatively, he may be more interested in gaining approval, support or popularity. 70 The president will organise a team that will help him out and which he will feel comfortable leading. The advisory system that he will consider creating will depend on his perception of comfort, be it loyal and non-confrontational, loyal and highly experienced, or a close group of friends, with or without experience. This study will focus on a particular decision group conformation in order to identify Groupthink and, in this eventuality, to evaluate the existence of a defective decisionmaking process as an outcome, as Janis predicts.

Plan Colombia Decision-Making
Structural elements: the "failed state" concern. In this particular case, the issue at stake was to convince another government -President Bill Clinton'sto approve an extraordinarily large aid package with which the Colombian government would restructure the torn legacy of the previous government. Politically, the stakes were high, the Colombian government having promised to bring peace through the peace talks with the FARC guerrilla organisation. In fact, this factor explains how President Pastrana came into power in the first place.

The Negotiating Table
Small group decision-making. President Pastrana was well known for working with his friends: "As soon as Pastrana became President, he was accused of ruling with his friends", for he ruled with a "trustworthy team". 71 One particularity of this negotiating table is that all the members belonged to President Pastrana's inner circle, that is, he appointed only his closest friends to assist him in launching Plan Colombia. A description of the decision-making table is as follows.
Based on both evidence and observation, we argue that the following two explanatory factors help us to understand the motivations leading to Groupthink. We call the first one the sense of a historic legacy, with cognitive elements such as the idea of Pastrana being remembered as the president who brought peace by virtue of his Plan Colombia initiative. Second is the element of group cohesiveness, as a guarantee for Pastrana that he would not encounter opposition or dissent among the members who, being his personal non-political friends, would help him pursue his foreign policy objectives without threatening his role as president. Sense of historic legacy. The goal of achieving peace would certainly place Pastrana in a high position, for people would remember him as the president who was able to terminate a conflict afflicting the country for nearly 50 years: "Pastrana desires to be remembered by history". 72 At a president-led negotiating table such as the one described, the president's own programme, power, prestige and place in history will determine his decisions, 73 revealing both the leader's motivation and interests. Motivation entails factors that predispose a government and its people to move towards a certain decisional path concerning external affairs, and it will always be accompanied by interests, either national or personal political interests. 74 It was 1999 and the stakes were high. So was national pressure, for President Pastrana had promised nothing less than peace (this "peace discourse" had become the most popular presidential campaign promise in Colombia, especially since 1981 when negotiating the nation's peace became a common tendency). 75 President Pastrana's campaign slogan promised change, "El Cambio es Ahora", and the solution for change was peace: "I promise change, and I want you to become involved in changing Colombia … , I want Colombia to live in peace". This was followed by: "Presidents want to do what national interests demand". 76 He had been elected thanks to the "peace mandate" he proclaimed and, as a result, he had "plunged himself into the rugged terrain of peace", a subject he claimed to have expertise on after he was kidnapped in 1988 by the Medellin cartel. 77 According to Guillermo Fernández de Soto, Pastrana's Minister of Foreign Affairs, achieving peace was the government's main objective, and as such, Colombia's foreign policy was to be focused on "achieving realist perspectives on the internal armed conflict". 78 This is why Plan Colombia was highly promoted worldwide in order to achieve financial and technical assistance. 79 However, Pastrana knew that his most important ally was the US, and its involvement would demonstrate historical support for the "Diplomacy for Peace" effort. 80 The only thing left was to organise a coherent and productive team that would help the president deliver on his promises. According to President Pastrana, It was urgent to recover the country's image, insert the country into the international community's agenda, seek international cooperation and guarantee support for peace negotiations, as well as to restructure the country's defence apparatus. 81 One antecedent condition for this "sense of historic legacy" is the fact that Andrés Pastrana is the son of a former president. Moreover, this fact places President Pastrana in the third characteristic and style described in Table 2: a president with prior experience and expertise. No doubt, Pastrana had prior knowledge of policy-making from his father's government. Pastrana is therefore catalogued as a dolphin, a term used to refer to the offspring of political figures who may symbolically "carry power throughout their veins", an example of hereditary democracy. 82 Whether people agree or not, Pastrana, a natural dolphin, is thought of as an experienced and knowledgeable political man.
In terms of foreign policy, Pastrana revealed a sense of experience. He is recalled as the president who travelled the most: "in only nine months of his government, Pastrana has travelled 18 times". 83 Unfortunately, public opinion did not approve of Pastrana's excessive travelling: "Colombians, in general, hate presidential tours … they perceive that a President who travels is a lousy President". 84 Nevertheless, Pastrana's foreign policy, considering that he would launch Plan Colombia, required him to exercise personal diplomacy as a strategy to convince Clinton's government and the US Congress to assist him. According to Pastrana, I directly participated in the task of convincing the American Congress of Plan Colombia. Another thing that helped was the team I appointed to handle international affairs, starting with the Minister of Foreign Affairs Guillermo Fernández de Soto, and our Ambassador, Luis Alberto Moreno. 85 The question on "how to convince the US" required all necessary diplomatic efforts. In Pastrana's words, "Plan Colombia was successful thanks to the team I created". 86 Group cohesiveness. Pastrana's group of eight was highly cohesive due to the bonds of friendship that linked the members, and, specifically, in their relations with the president. Friendship causes groups to become cohesive, and expected behaviours will follow a concurrence-seeking tendency. This case of cohesive policy-making demonstrates the existence of a group "whose members have worked together closely for years and face a common fate", as this case reveals:

Member's profile
Relation with the President Task Jaime Ruiz Civil engineer A businessman in politics, he is smart and well-studied. 1 Perceived as the "power behind the throne". 2 They went to school together: "we went to middle and high school together and since then we became friends". 3 They worked together during Pastrana's term as Mayor of Bogotá, and later in Pastrana's political campaign. 4 He was appointed as DNP Director. "The president likes to listen to Jaime Ruiz … he is the president's head counselor". 5 To design Plan Colombia. "He is the most important man in Pastrana's government". 6 Considered the "brain" of Plan Colombia and, in general, of Pastrana's political agenda. 7 Victor G. Ricardo "A low profile man who speaks with caution, as if he was taking care of the protagonist role that Pastrana gave him". 8 He won the president's trust after convincing the FARC guerrilla organisation of Pastrana's team peace policy. He was the first one to actually meet with the FARC chief commanders. 9 Referred to as "the man of peace", Ricardo became the first High Commissioner for Peace, in charge of developing the peace negotiations agenda with the FARC's top commanders. 10 He initiated the discussions on Plan Colombia during the peace negotiation phase and searched for international recognition in support of Plan Colombia. 11

Camilo Gómez
Lawyer "A very organized and loyal man". 12 Andrés Pastrana's most loyal aide, a "pure blood Pastranist", who has worked with Pastrana since Alvaro Gómez's conservative presidential campaign in 1986. 13 In 1998, Gómez was appointed Pastrana's private secretary but, after Victor G. Ricardo's resignation in 2000, he became the government's new High Commissioner for Peace. 14 Guillermo Fernández de Soto Lawyer and economist "A serious, quiet, studious man. An international affairs expert". 15 He was Pastrana's campaign manager, and a close friend. 16 He has known Pastrana since his first political campaign (1991). 17 One of Plan Colombia's main ideologists, he was involved in the writing of Plan Colombia. 18 As Minister of Foreign Affairs, he was responsible for restructuring Colombia-US bilateral relations, as well as seeking international cooperation 19 in favour of Plan Colombia.

Mauricio Cárdenas
Economist (PhD) A renowned economic analyst. 20 Conservative, a follower of Pastrana's political line. 21 According to Pastrana, Cárdenas "is a very close friend of my family", for his father, Jorge Cárdenas Gutierrez, was a childhood friend of former President Misael Pastrana 22 (Andrés's father), and a "spiritual counselor of Andrés Pastrana, his dear friend's son". 23 To help improve the country's infrastructure and competitiveness, 24 as part as Plan Colombia's social component. Table 3. Continued.

Member's profile Relation with the President Task
Gustavo Bell Lawyer and economist (PhD) "Lucky Andrés who can count on this good and generous man's loyalty". 25 Being a liberal, Bell was certainly a strategic appointment of President Pastrana not only to win liberal votes, but also prestige, for Bell is a well-known intellectual with moral authority. 26 Gustavo Bell, Vice-President, concurrently served as Minister of Defense and High Commissioner for Human Rights after Luis Fernando Ramírez's resignation. 27 "Gustavo Bell was key for the human rights issue", 28 "especially in addressing our human rights policy towards the European Union". 29 He also worked in drafting Plan Colombia with Jaime Ruiz. 30

Luis Alberto Moreno
Business administrator Cerebral and smart, Moreno has been Andrés's lifetime alter ego. 31 Pastrana's national campaign coordinator, considered "the brain" of Pastrana's political campaigns. 32 Another very close friend of the president and true blood conservative Pastranist, "his lifelong teammate". 33 Moreno is also a high school friend of Pastrana's: "one of my best friends, Luis Alberto Moreno, TV Hoy news station manager and former school classmate of mine". 34 According to Ruiz, "He wasn't such a close friend of Pastrana's during high school as I was, but when I returned to Colombia to work in his campaign for Mayor, Luis Alberto had become his closest friend". 35    Member's profile Relation with the President Task  Table 3. Continued.
Member's profile Relation with the President Task "[On Pastrana] one can imagine his tribulations in a palace of close friends who plotted a Caucasian circle only to preserve his image". 87 In search of evidence, we investigated the relations of the group of eight with President Pastrana and the tasks they were assigned for the creation of Plan Colombia (see Table 3).
Pastrana shaped a team of advisors that would guarantee loyalty and expertise. He required a group of "geniuses" and intellects that would bolster his image and, more importantly, a team of friends with no political ambitions that would pose a threat to Pastrana's interests. There is clear evidence of group efficacy, in the sense that, according to Hermann, the leader will play a key role in determining the group's own sense of efficacy by assigning each member a task, defining the goal and establishing the desired results. 88 As discussed earlier, Pastrana best fitted in the "leader with prior experience" style. Moreover, he "will choose a loyal staff because he is proactive", but "will also need experts to accomplish his tasks"; "the President will also seek to do business personally" because "he feels involved". 89 As a result, friendship demonstrates shared morality as the main symptom of Groupthink, in the sense that "members consider loyalty to the group the highest form of morality". 90 As Camilo Gómez recalls, Yes, all this team, Guillermo Fernández, Jaime Ruiz, Luis Alberto Moreno, etc., we may say that we share a history in politics and in our personal relations. In all organisations, every bullfighter chooses his own quadrille.
When it comes to politics, there will always exist a close circle that will surround the president. This was a very close circle. 91 What really counts, no matter how odd this may seem, is friendship. Because of this, problem-solving, in the event of conflict situations that usually emerge during the decision-making process, was not a concern. After all, "trust is the key concept in building a good relationship". 92 It is very important for a president to have friends that are not afraid to warn and advise him. Even if I am the one who makes the final decision. When a president is not surrounded by friends there is a different environment. They may be respectful, but they will never be blunt with you. With friends, one can discuss and share, and decide. When there was a problem, they spoke directly with the president, and that is a very effective problemsolving strategy. 93 Regarding problem-solving and conflict resolution, for this empirical study we asked President Pastrana about the occurrence of conflicting situations. One preliminary conclusion from this analysis of friendship and policy-making is that Groupthink is demonstrated by the "tendency to agree with each other, sometimes without carefully considering the costs and benefits of various alternatives". This can result in symptoms such as illusions of invulnerability, rationalisation, moralisation, stereotyping and unanimity. 94 The important thing was that the president was there, everyone depended on the president. As I told you, they were all friends. We were all friends. It is very different when you have to appoint someone because of a political favour (logrolling). Jaime and I, for example, went to school together! We always ended up building consensus. 95 Consensus-building as mentioned by Pastrana may be a clear sign of Groupthink, but also of group efficacy. According to Hermann, group cohesion is thought of as a source of collective efficacy, founded in the knowledge that "good working relationships among group members (group cohesion) increase group efficacy", and this is what the president and his group of advisors thought. 96 Of all the members of the inner circle, Jaime Ruiz is perhaps the player who most embraces the characteristics shown above, a non-political man, therefore a nonambitious intellectual.
The person in charge of designing Plan Colombia was Jaime Ruiz. That is the reason why I decided he should resign from his position as Director of National Planning, so I told him, "I will give you an office just next to mine so we can work closely together in the strategy on how to move the government forward". 97 Ruiz is the enigmatic type, but this does not worry him. Lacking real political ambition, he is able to defend unpopular views and not to worry too much about their impact on opinion. 98 If one could propose a hierarchy, after Jaime Ruiz the closest player to President Pastrana was certainly Luis Alberto Moreno.
One thing is appointing an ordinary ambassador, and another thing is appointing a friend. With a friend, in this case Luis Alberto, we were able to work very closely, "pulling all the strings" together. 99 President Pastrana emphasised the importance of sharing information; this is one of the roles an advisor has to perform. With his "friends", as he mentioned, there was no doubt he would receive first-hand information. Additionally, he was capable of "fast thinking", in the sense that instant information produces automatic responses when designing strategies. This demonstrates that the members served as mind-guards, in which the expected behaviour will be a mechanism of selfdefence against imagined threats, another Groupthink symptom.
I also worked with Luis Alberto Moreno, and carefully observed people in the US Congress. I was always telling him "I don't like the position of this one, or this one, so work with them and have them come visit us in Colombia". There was a direct contact with Moreno and Jaime, and the Minister of Foreign Affairs on how to "sell" our diplomacy for strengthening peace. 100 In the end, it will all depend on the president's approach, whether he or she will listen and accept suggestions, or just listen; whether he or she will tolerate conflict and seek conciliation and consensus-building, or whether he or she will not tolerate discussions out of order; whether he or she will appoint loyal players, or experts; people with different goals and interests, or people that follow presidential concerns. 101 We have discussed the fact that "what the president is can influence what the advisers are". In President Pastrana's case, he clearly demonstrated his willingness to generate consensus among all the players. 102 He repeatedly displayed his pride in the team he created. Appointing friends might be seen as a strategy to guarantee not only the achievement of his goals, but also that he would receive the information he demanded from the bureaucracy as a whole. He views his advisers as a team, bonded by ties of friendship; "if the leader wants his preferences to prevail, he will probably seek out advisers who have a similar philosophy, are loyal and predisposed to please him". 103 Pastrana's idea of teamwork when it comes to policy-making is the main cause of Groupthink.

Conclusion
This study has evidenced that decision-making may follow more informal procedures even when the players represent the bureaucracies. Moreover, that informality does not necessarily symbolise disorder or lack of coordination. After all, human decision-makers, either American or non-American, follow their own perceptions, motivations and instincts, create their own shortcuts and strategies. How to achieve the necessary information in order to solve the entire puzzle will remain a challenge. However, scholars need to be constantly challenged; they must work hard to find all the clues required for "digging in" and, as a result, discovering the mystery of human behaviour.
FPA based on a small group dynamics approach has been useful to some extent in understanding how a specific foreign policy case was formulated. Regardless of the benefits of the FPA enterprise, from our analysis we submit the following considerations for those who may wish to pursue an analysis of Latin American foreign policy: . Since traditional FPA has centred around the White House, this study is perceived as an opportunity to explore how small group dynamics, specifically Groupthink, help provide effective understanding of how foreign policy decisions are made. For this case to prove a valuable generalisation, there are three elements to consider: first, following the structural elements presented in this case, the decision was one of national security priority; second, as a consequence of the latter, the decision was made under stressful conditions; and third, the foreign policy decision formulated was originated in a powerless country with the objective of affecting a powerful one, thus exhibiting a condition of asymmetry in the process, a noticeable sign of overlay. In this particular case, the US overlay was not only evident, but constant during the decisionmaking process as a whole. Pastrana's advisors acted in the presence of influential US players. One could almost sense a feeling of subordination, particularly if the goal was to approve a millionaire aid package. . There is no such thing as "harmful group dynamics". This study has demonstrated a positive high-quality decision-making dynamic, originating in what Janis's Groupthink theory would regard as a flawed, defective process. Even though Janis predicted flawed decision-making as a result of group cohesionand concurrence-seeking tendencies provoked by a "groupy" behaviour, this case demonstrated that successful decision-making, when the projected goal was effectively accomplished, can also arise from Groupthink behaviour. Presidents will probably prefer their closest friends, and it is highly probable that the result will be group insulation, group cohesiveness, gatekeeping, and distrust towards anyone who is not inside the circle of trust. This would give rise to Groupthink. However, as has been demonstrated, this does not necessarily result in low-quality outcomes. . According to Hermann, group cohesiveness will increase group efficacy and, for this to take place, the following elements must be taken into account: the group's past performance, vicariously perceived performance of others, support and encouragement from others, group cohesion and similarity between individual group members' beliefs. 104 For example, according to Pastrana, We succeeded. We became the third country after Israel and Egypt, in receiving the greatest amount of US foreign assistance. We succeeded with the US Congress. We achieved consent, with 99% of the Congress vote. We obtained maximal approval. Plan Colombia has been a successful state policy. We succeeded in making US foreign policy-makers and Congressmen change their view of Colombia. We also succeeded in creating the concept of shared responsibility. 105 . Leaders matter: "The quality of the process is something over which decision makers can have great influence". 106 This study demonstrated that presidential coordination may create strong, cohesively working groups (not necessarily involving experts). "Groupy" behaviour may indeed be a result of the guidance of a self-confident leader who seeks to pursue effective decisions in a short time. Probably, to overcome stressful situations, the leader will be inclined to adopt shortcuts such as "group insulation, failing to value disagreement, suppressing dissent, and gatekeeping" which, according to Schafer and Crichlow, would be regarded as "problematic in decision-making". 107 In reality, when it comes to foreign policy decision-making, there is no such thing as problematic behaviours, especially in the face of a strong presidential influence. It all combines into one solid decision-making strategy that leaders carefully adopt to maximise their interest. . Powerless countries striving for foreign policy recognition, in the presence of overlay, will probably require groups to be coherent, concurrent and loyal to the president's perception (who "knows best" when it comes to foreign policymaking), even if this implies neglecting the virtues of critical thinking in a decision-making context. 108 . Additional conditions for providing generalisations are weak foreign policy bureaucracies and disorganised institutions as traits of non-US and non-European countries. For this reason, mainstream FPA will not be applicable in the same form. Nevertheless, we argue that these beliefs represent stereotypes that have traditionally identified institutional performance in the Global South, without considering institutional building and improvements in some countries. . When applying FPA to a non-US case study, one would normally think that the availability of information (academic journals) and language would be considered research barriers. However, an increasing number of universities have sufficient information and provide access to most international journals.
Additionally, more and more IR scholars are bilingual, so language should not be considered a problem. As this final discussion shows, the limits to FPA reveal a persistent unawareness of FPA, specifically of its epistemological foundations. . Finally, we assert that our analysis complements existing literature on Latin American foreign policy decision-making. On the one hand, our evidence confirms the existence of a solid presidential figure and, on the other hand, we have justified the significance of the role of the president's advisors within a decision-making process. 109 Hence, we propose a window of FPA applied to Latin American cases that will allow traditional research to reach broader spectra of analysis beyond traditional theory and internal political dynamics. 110 Our examination of this particular inner circle may be seen as a first approximation. We contend that this new knowledge could be extended to strengthen intra-regional case studies and, subsequently, to generate a valuable research agenda in the emerging field of Latin American FPA.

Disclosure Statement
No potential conflict of interest was reported by the authors.